TRANSPORTATION RESEARCH BOARO

2016 EXECUTlVE COMMITTEE'

Chair: James M. erites, ExecUlivc Vice Presiden! of Operation" Marie Therese Dominguez, Administrator, Pipeline and Hazardous
Dallas-Fort Worth Inlemational Airpon, Texas Materials Safety Administration. U.S. Department ofTransportation
Vke Chair: Paul Trombino 111,Directof.lowa Depanmcm of (ex officio)
Tran~port1lion, Ames Sarah Feinberg, Administrator, Federal Railroad Administration.
Executive Director: Neil J. Pedcrsen. Transportation Re~arch Board U.S. Depanment ofTransportation (ex officio)
Carolyn F1owers. Acting Administrator, Federal Transit Administratioll,
U.S. Depanment ofTransportation (ex officio)
Victoria A. Arroyo. Execulivl' Director, Georgetown Climate Center;
LeRoy Gishi, Chief, Division ofTransportation, Bureau of Indian
Assislaol Vean. CCOters and In,titules; and Professor and Director. Affairs, U.S. Department of the Interior, Washington, D,e. (ex officio)
Environmental Law Program, Georgetown Universily Law Ccnter, John T, Gray 11, Senior Vice President. Folicy and Economics,
Washington,O.e. Associatíon of American Railroads, Washington, D.C. (ex officio)
Senlt E. Bermelt. Director, Arkansas Stal!' Highway and Transportalion Michael P, Huerta. Administrator, Federal Aviation Administrntion,
Depanment, Linte Rock U.s. Dcpartrnem ofTransportation (ex officio)
Jennifer Cohan, Secretar)', Delaware Dep;utment ofTransportation, llover
Paul N, Jaenichen, Sr., Administrator, Maritíme Administration.
Malcolrn Dougherty. Director, California Departmelll of U.S. Department ofTran,ponatíon (ex officio)
Transportation. Sacramento Bevan B, Kirley, Research Associate, University of Nonh Carolina
A. Slewart Fotheringham. Profeswr, School ofGeograpbical Sciences Highway Safety Research Center. Chapel Hill, and Chair, TRB Young
and Urban Planning, Ari1.Ona SUte University, Tempe Membcrs Council (ex officio)
John S. Halikowski, Director, Arizona Departrnent ofTmnsponation,
Gregory G. Nadeau. Administrator, Federal Highway Administralion,
Pl10cnix U.S. Depanment ofTransponalion (ex officio)
Sus.m Hanson. Distinguished University Professor Emerita. Gr,uluate Wayne Nastri. Acling Executive Officer, Somh Coast Air QualilY
School of Geography, Clark. University, Woreester, Massachusetts Management Dislrict. Diamond Bar, California (ex officio)
Sleve Heminger, Executi\'e Director, M-ctropolitan Transportatíon Mark R. Rosekind, Adminislrator. National Highway Traffic Safety
Cornrnission, Oak.land, California Administration, U.S. Depanment ofTransportation (ex officio)
Chris T. Hendriekson, HamcN:hlag Profess"r of Engineering, Camegie
Craig A. Rutland, U.S. Air Force Pavement Engineer. U.S. Air Fofl.'e
Mellon University, Pilt5burgh, Pennsylvania Civil Engineer Center, Tyndall Air Foree Base, Aorida (ex officio)
]effrey D. Holl, Managing Director, Power, Energy, and Infrastructure Reuben Sarkar, Deputy Assistant Secretary for Transportation.
Group, BMO Capital Mark.ets Corporation. New York. .

,
U.S. Department uf Energy (ex officío)
S. Jack Hu, Vice President for Re>earch and J. Reid and Polly AndeNln Richard A, While. Acling President and CEO, American Public
Pr"fessor of ManufaclUring, University of Miehigan. Ann Arbor Transportation Association, Washingtun, D.e. (ex officío)
Roger B. Hoff, President, HGLC, LLC, Farmington Hitls, Miehigan Gregory D. Wintree, Assistant Secretary for Research and Technology,
Geraldioe KoaU. Professor, Sol Price School of Public Policy, Vilerbi Office of the Secretary, U.S. Departrnent of Transportatinn (ex officio)
Sc1looI ofEngineering, University of Southern California, Los Angeles Frederick G. (Bud) Wright, Executi,'e Director. American As,ociation
Ysela L1ort. Consultant, Miami, Florida
of State Highll'ay and Transportation Officíals, Washington, D.e.
Melinda McGralh. Executive Director, Missi"ippi Department of
(ex officio)
Tramportation, Jack.son Paul F. Zukunft (Admiral, U.S. Coast Guard), Commalldant, U.S. Cuast
James P. Redeker, Commissioner, Connecticut Department of Guard. U.S. Departmenl of H"meland Security (ex officio)
Transportation, Newington
Mark L. Rosenberg. Execmive Director, The Task. Force for Global
Health, lne" Decatur, Georgia
Kumares C. Sin ha. Olson Distinguishcd Professor of Civil Engineering.
Purdue University, West Lafayette.lndiana
Daniel Sperling. Professor ofCivil Engirteering and Environmental
Seience and Poliey; Director, Ins!Ílute of Transportation Studies,
University of California, Davis
Kirk T. steudle, Director, Michigan Dcpanment ofTransportation,
Lansing (Pa.st Chair, 2014)
Gary C. Thomas, President and Executive Director, Dalias Arca Rapid
Tnmsportation Research Board publieations are a,'ailable by ordering
Transit, Dalias, Texa.s
individual publications directly from!he TRB Business Office, through
POli Thomas. Senior Vice President of Stale Governrnent Affairs, United
the Internet at www.TRB.org,orby annual sub'«.Tiplion thmugh
Pareel Service, Washington, D.e.
organizationalor individual affiliation wi!h TRE. Affilíates and Iibrary
Kalherine F. Tumbull. Executíve Associate Director and Research
subscribers are e1igib1e for substantial disenunts. For funher information,
Sdentist, Texas A&M Transportation Institote, College Station
contael me Transportation Research Board Business Office. 500 Fifth
Dean Wise, Vice President of Network Strategy, Burlíngton Nortltem
Street, NW, Washington, OC 20001 (te1ephone 202-334-3213;
Santa Fe Railway, Fort Worth, Texas
fax 202-334-2519; or e.mail TRlhales@nas.edu).

Thomas p, Bostick (Lieotenant Gcneral, U.S. Army), Chief of Engineer;

Copyright 2016 by the National Academy of Scicnces.
and Commanding Genend, U.S. Army Corps of Engineers, Washington,
AH righlS reserved.
D.e. (ex offieio)
Printed in Ihe United Stale, uf America.
James C. Card (Vice Adntira!. U.S. CoaSl Guard. retired), Maritime
Consoltant. The Woodlands. Texas, and Chair, TRB Marine Board
ISBN 978-0-309-36997-8ISlípcased set of three volumes)
(ex officio)
[SBN 978.0-309-36998-5 [Volume 1]
T. F. Scott Darling IIJ, Acting AdminisO"ator and ChiefCoun>eI, Federal
ISBN 978-0.309.36999-2 [Volurne 2]
Motor Carrier Safety Administration. U.S. Department ofTransportation
ISBN 978-0-309-37lXJO-4 [Volume 3J
(ex officio)
ISBN 978-0.309-37001-1 [Volume 4, onlíne only)
• Mombership asof JUIIe 2016,
 The National Academies of
SClENCES' ENGINEERING' MEDICINE

The National Academ:,.. of Sciences was cstablished in 1863 by an Act of Congress,

signcd by President Lineoln, as a private. nongovernmental institution 10 advise ,he
nalion 00 issues rclalcd 10 seicoce and technology. Members are elecled by their peces
foc outstanding contributions to rescarch. Dr. Ralph J. Cicerone is presidcnt.
The National Acadcm)' of En~ineering was eslablished in 1964 under the chartee of
the National Acaderny of Sciences to bring the practices of engincering 10 advising
the nation. Members are clccled by their peces foc extraordinary contributions 10
engineering. Dr. C. D. Mote, lc., is president.
The National Acaderny of Medicine (fonnerly (he Institule of Medicine) was
cstablished in 1970 under Ihe charter of Ihe National Academy of Sciences lo advise
Ihe nation on medical and health issues. Members are elecled by Iheir peers for
distinguished conlributions 10 medicine and health. Dc. Viclor J. Dzau is presiden!.

The Ihree Acadcmics work logether as the National Academies of Sciences, Engineering,
and Medicine to provide independem, objective analysis and advice to Ihe nalion and
conduct other activities lO solve complex problems and inform public policy decisions.
The Academies also encourage education and research. recognize outslanding
contribulions 10 knowledge, and increase public undcrstanding in malters of science.
engincering, and medicine.

Learn more about the National Academies of Scicnces. Engineering, and Medicine at
www.national.academ¡es.org.

The Tl'ansportation Rcsearch Board is one of seven major programs of Ihe National
Academies of Sciences, Engineering, and Medicine. Thc mission of the Transportalion
Research Board is lO increase the benefits Ihal transportation contributes lO society by
providing leadership in Iransportation innovalion and progress Ihrough research and
information exchangc. conducled within a seuing Ihal is objective. inlerdisciplinary, and
multimoda1. The Board's varied committces. task forces, and panels annually engage aboul
7,000 engineers, scientists, and olhcr transportation researchers and practitioners from Ihe
public and private sectors and academia. aH of whom contribute their expertise in the public
imeres!. The program is supported by state Iransportalion departmems, federal agencies
including the component administralions ofthe U.S. Departmenl ofTransponation, and
olher organizations and individual s imeresled in the development of transportation.

Lcarn more aboul Ihe Transportalion Research Board al www.TRR.org.

 Highway Capacity Manual: A Guide for MultimodaJ Mobility Analysis

CHAPTER 16
URBAN STREET FACILITIES

CONTENTS

1. INTRODUCTlON 16-1
Overview 16.1
Chapter Organization 16-1
Related HCM Content 16-2

2. ca NCEPTS 16-3
Mul timodal Eval uation Framework 16.3
Analysis Type 16.4
Spatial and Temporal Limits 16-5
Urban 5treet Fadlity Defined 16-6
LOS Criteria 16-6
Scope of the Methodologies 16-9
Limitations of the Methodologies 16-9

3. MOTORIZED VEHIClE METHODOlOGY 16-10

Scope oí the Methodology 16.10
Required Data and Saurces 16.13
Overview oí the Methodology 16-14
Computational 5teps 16-15

4. PEDESTRIAN METHODOLOGY 16.17

Scope of the Methodology 16-17
Required Data and Saurces 16-18
Overview oí the Methodology 16.19
Computational 5teps 16-19

5. BICYClE METHODOLOGY 16-22

Scope oí the Methodology 16-22
Required Data and Saurces 16-23
Overview of the Methodology 16-23
Computational 5teps 16.24

Chapter 16¡Urban 5treet Facilities Contents

Version 6.0 Page 16-i
Highway Capacity Manual: A Guide for Mulfimodal Mobilify Ana/ysis

6. TRANS IT METH O DO LOGY ..........................................................................• 16-26

Scope oí the Methodology 16-26
Required Data and Sources 16-26
Overview of the Methodology 16-27
ComputationaI Steps 16.27

7. APPLI CA TI O NS 16-29
Exam pIe Problems 16-29
Generalized Daily Service Volumes ]6-29
Analysis Type ] 6-3]
Use oí Al temati ve Tools 16-32

8. REFEREN CES ..............•........................................................................................ 16-33

Contents Chapter 16/Urban 5treet Facilities

Page 16-ii Version 6.0
 Highway Capacity Manual: A Guide for Mulfimodal Mobilify Analysis

LIST OF EXHIBITS

Exhibit 16.1 Integrated Multimodal Evaluation Framework...... . 16-4

Exhibit 16-2 Signal Spacing Associated with Effectively lsolated
Operation 16-5
Exhibit 16-3 LOS Criteria: Motorized Vehide Mode 16-8
Exhibit 16-4 LOS Criteria: Pedestrian Mode 16-8
Exhibit 16-5 LOS Criteria: Bicyde and Transit Modes.................................... .16-9
Exhibit 16.6 Three Altemative Study Approaches 16-11
Exhibit 16.7 Required Input Data and PotentiaI Data Sources for
Motorized Vehide Anal ysi s 16-13
Exhibit 16.8 Motorized Vehide Methodology far Urban Strcet Facilities 16-15
Exhibit 16-9 Qualitative Description of Pedestrian Space 16-18
Exhibit 16-10 Required Input Data and Potential Data Sources for
Pedestrian Anal ysis 16.18
Exhibit 16-11 Pedestrian Methodology for Urban Street Facilities 16-19
Exhibit 16-12 Required Input Data and Potential Data Sources for Bicyde
Analysis 16-23
Exhibit 16-13 Bicyc1eMethodology for Urban Street Facilities 16-23
Exhibit 16-14 Required Input Data and Potential Data Sources for Transit
Analysis 16-26
Exhibit 16-15 Transit Methodology for Urban Street Facilities 16-27
Exhibit 16-16 Generalized Daily Service Volumes for Urban Street
Facilities 16-30

O1apter 16jUrtan Street Facilities Contents

V~6.0 Page 16-iii
 Highway Capacity Manual: A Guide for Multimodal Mobilify Analysis

1. INTRODUCTION

OVERVIEW VOLUME 3: INTERRUPTED FlOW

16. Urban Streel: Facilities
This chapter describes methodologies for evaluating the operation of each of 17. Urnan Street Reliability and
ATOM
the following urban street trave! modes: motorized vehiele, pedestrian, bicyele,
18. Urban stTeet Segments
and transito Each methodology is used to evaluate the quality of service provided 19. Slgnalized lntersections
to road users traveling along an urban street fadlity. A detailed description of 20. TWSC Intersections
21. AWSC Intersections
each travel mode is provided in Chapter 2, Applications. 22. Roundabouts
23. Ramp Terminals and
Design or operational decisions that are intended to iroprove the service Alternative Intersections
provided to one mode can sometimes have an adverse impact on the service 24. Off-5treet Pedestrian and
Bicycle Facilities
provided to another mode. The challenge for the analyst is to design and operate
the urban street in such a way that all relevant travel modes are reasonably
accommodated. The methodologies described in this chapter are intended to
assist the analyst by providing a means of assessing the performance of each
travel mode that takes account of the influence of other modes.
This chapter describes methodologies for evaluating urban street facility
performance from the perspective of motorists, pedestrians, bieyclists, and transit
riders. These methodologies are referred to as the motorized vehiele
methodology, pedestrian methodology, bieyele methodology, and transit
mcthodology. COIlectively,the methodologies can be used to evaluate urban
street facility operation from a multimodal perspective.
Each methodology in this chapter is focused on the evaluation of an urban
street facility that is made up of two or more segments. A separate methodology
for evaluating the performance of individual segments is describcd in Chapter
18, Urban Street Segments. The performance measures associated with each
segment are then aggregated to the facility level with the methodology described
in this chapter.
A facility's performance is describcd by the use of one or more quantitative
measurcs that charactcrize sorne aspcct of the service provided to a spedfic road-
user group. Performance measures cited in this chapter inelude motorized
vehiele travel speed, motorized vehiele stop rate, automobile traveler perception
score, pedestrian travel speed, pedestrian space, pedestrian level-of-service
(LOS) score, bicyele travel speed, bieyele LOS score, transit vehiele travel speed,
and transit passenger LOS score.
The four methodologies describcd in this chapter are based largely on the
products of two National Cooperative Highway Research Program projects (1, 2).
Contributions to the methodology from other research are referenced in the
relevant scctions.

CHAPTER ORGANIZATlON
Section 2 of this chapter presents concepts uscd to describe urban street
facility performance from an operations perspective. A multimodal evaluation
framework is initially discussed. Guidance is then provided for establishing the
facility analysis boundaries and the analysis period duration. A discussion about
how an urban street facility is defined for the purpose of this chapter foIlows.

Chapter 16/Urban Street Facilities Introduction

Version 6.0 Page 16-1
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Finally, the service measures and LOS thresholds used in the methodology are
examined.
Section 3 presents the rnethodology for evaluating motorized vehide service
along an urhan street facility.1t indudes a description of the scope of the
rnethodology and the required input data. It coneludes with a description of the
computational steps that are followed for each application of the rnethodology.
Section 4 presents the methodology for evaluating pedestrian service along
an urban street facility. It ineludes a discussion of methodology scope, input
data, and computational steps.
Section 5 presents the methodology for evaluating bicyde service along an
urban street facility. It ineludes a discussion of methodology scope, input data,
and computational steps.
Sedion 6 presents the methodology for evaluating transit ridcr service along
an urban street facility. It ineludes a discussion of methodology scope, input
data, and computational steps.
Section 7 presents guidance on using the results of the facility evaluation. It
indudes example results from each rnethodology and a discussion of situations
where alternative evaluation tools may be appropriate.

RELATED HCM CDNTENT

Other Highway Capacity Manual (HCM) content related to this chapter
ineludes the following:
• Chapter ]7, Urban Street Reliability and ATDM, which provides a
methodology for evaluating travel time reliability and guidance for using
this methodology to evaluate alternative active traffic and demand
managernent (ATDM) strategies;
• Chapter ]8, Urban Street Segments, which describes concepts and
methodologies for the evaluation of an urban strcet segment;
• Chapter 29, Urban Street Facilities: Supplemental, which provides details
of the reliability methodology, a procedure for sustained spillback
analysis, information about the use of alternative evaluation too1s,and
example prob1ems demonstrating both the urban street facility
rnethodologies and the reliability methodology;
• Chapter 30, Urban Street Segments: Supplemental, which describes
procedures for predicting platoon fiow, spillback, and deJay due to turns
from the major street; a planning-Ievel analysis application; and example
problems demonstrating the urhan street segment rnethodologies; and
• Section K, Urban Streets, in Part 2 of the Plamlillg and Prelimillary
Engjlleering Applicatiolls Cuide fa the HCM, which describes how to
incorporate this chapter's methods and performance measures into a
planning or preliminary engineering effort.

Introduction Chapter 16¡Urban $treet Facilities

Page 16-2 Version 6.0
 Hjghway Capacity Manual: A Guide for Multimodal Mobility Analysis

2. CONCEPTS

This scction presents concepts used to describe urban street facility

performance from an operations perspective. The first subsection describes a
multimodal evaluation framework that promotes consideration of each urban
strcet travel mode and its interactions with other modes. The second assists the
analyst in determining the type of analysis to be conducted. The third provides
guidance for establishing the facility analysis boundaries and the analysis period
duration. The fourth describes how an urban street facility is defined in terms of
points, links, and segments. The fifth discusses the service measures and LOS
thresholds used in the methodology. The last identifies the scope of the collective
set of mcthodologies.

MUL TIMODAL EVALUATlON FRAMEWORK

The urban strect right-of-way is typically shared by multiple travel modes.

Travelers associated with the more common modes inelude motorists,
pedcstrians, bieyelists, and transit passengers. Thc factors that influence the
pcrception of quality of service provided to these travelers vary by mode because
each mode has a different trip purpose, length, and expcctation.
The shared street right-of-way typically requires that the modes operate in
elose proximity to each other, sometimes even sharing the same portion of the
cross section (e.g., a vehicular traffic lane). This sharing may be workable when
the modes are characterized by low demand volumes; however, acceptable
operation for moderate to high volumes typically requires the spatial separation
of thc modes along the street and temporal (Le., signal) separation at the
intersections.
The methodologies described in this chapter can be used to evaluate
simultaneously the LOS provided to each travel mode on an urban street. A
framework for this evaluation is shown in Exhibit 16-1.
Thc framework shown in Exhibit 16-1ilIustratcs the integrated multimodal
evaluation approach supported by the methodologies. Thc LOS providcd to each
travel mode is scparately evaluated. The relative importance given to each
mode's LOS should be determined by the analyst (or operating agency) and
reflect consideration of the subject street's functional class and purpose. The LOS
for each mode should IlOt be combined into one overall LOS for the street. This
restriction is based on the recognition that trip purpose, length, and expectation
for each mode are different and that their combination does not produce a
meaningful result.
Exhibit 16-1 iIIustrates how the travcl modes compete for limited right-of-
way along thc street and at the intersections. They also compete for Iimited
signal time at the intersections. For a given right-of-way, the allocation of space
to one mode often requires a reduction (or elimination) of space for other modes
and a corresponding reduction in their servicc quality.

Olapter 16jUrtan 5treet Fadlíties

Concepts
VetWn 6.0 Page 16-3
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Exhibit 16-1
Tl'llnsit Mode
Integrated Multimodal
Evaluation Framework

R~.:'!,.,
M"'"""'
~ Sóde",alk
••••
llicycIe BusIones,
18~es unouts

--
--Priofity

Adverse FIow loterllctioo'

The lower part of Exhibit 16-1 illustrates the potential adverse interaetions
between the motorized vehide mode and the other modes. As the volume or
speed of the vehicle traffie stream inereases, the LOS for the other modes may
deerease. If bicyde, pedestrian, or transit flows inerease, the LOS for the
motorized vehide traffie stream may deerease. In general, ehanges that alter
resouree alloeation or flow interaetion to improve the LOS for one mode may
affect the other modes.

ANAL YSIS TYPE

The phrase analysis type is used to describe the purpose for whieh a
methodology is used. Eaeh purpose is assodated with a different level of detail,
sinee it relates to the precision of the input data, the number of default values
used, and the desired aecuracy of the results. Three analysis types are recognized
in this ehapter:
• OperationaL
• Design, and
• Planning and preliminary engineering.
These analysis types are discussed in more detail in Chapter 2, Applieations.

Olapter 16fUrban 5treet Facilities

Concepts
Version 6.0
Page 16-4
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

SPATIAL ANO TEMPORAL L1MITS

Analysis Boundaries
The methodologies described in this chapter are typically used to evaluate an
entire facility; however, for sorne speciíic conditions, evaluation of the entire
facility may not be necessary. For thesc conditions, the appropriate segment or
intersection chapter methodology may be used alone to evaluate selected
segments or intersections. In general, the analyst determines the spatial scope of
each analysis (e.g., one intersection, one segment, two segments, or aH segments
on the facility) on the basis of analysis objectives and agency directives.
Evaluation of an individual segment or intersection may be acceptable when
the segment or intersection is considered to operate in iso/ation from upstream
signalized intersections. A segment or intersection that is effectively isolated
experiences negligible influence froro upstream signalized intersections. Flow on
an isolated segment or at an isolated intersection is effectively random over the
cyde, and no platoon pattem is discernible in the cydic prafile oí arrivals. These
characteristics are more likely to be found when (a) the nearest upstream
signalized intersection is sufficiently distant from the subject segment or
intersection and (b) the subject segment or intersection, if signalized, is not
coordinated with the upstream signal.
A segrnent or intersection is sufficiently distant from the nearest upstream
signal if an intermediate intersection uses STOPor YIELDcontrol to regulate
through traffic on the facility. lf there is no intermediate STOP-or YIELD-controlled
intersection, Exhibit 16.2 can be used to obtain an indication of whether a
segment or an intersection is sufficiently distant from an upstream signa!. lf the
distance between signals is aboye the trend line, the subject intersection or
segment is likely to operate as effectively isolated (provided that it is not
coordinated with the upstream signal).

Exhibit 16-2
Signal Spadng Assodated
___~~~o,-~~nt _
with Effectively Isolated
is elfectjvely isolated.
Operation

3,000

2,000
Irtersection or segment
is not íSOIated.
1,000

o
25 30 35 40 45 50 ss
Speed Limit (mi/h)

Study Period and Analysis Period

The study period is the time interval represented by the performance
evaluation. It consists of one or more consecutive analysis periods. An analysis
period is the time interval evaluated by a single application of the methodology.

O1apter 16¡Urtan Street Facilities

Version 6.0 """'" 16-5
Page pi>
Highway Capacity Manual: A Guide for Mulfimodal Mobility Analysis

The methodology is based on the assumption that traffic conditions are

steady during the analysis period (Le., systematic change over time is negligible).
For this reason, the durabon oi the analysis period is in the range of 0.25 to 1 h.
The longer durations in this range are sometimes used for planning analyses. In
general, the analyst should use caution with analysis periods that exceed 1 h
because traffic conditions are not typically steady for long time periods and
because the adverse impact of short peaks in trafiic demand may not be detected
in the evaluation.

URDAN STREET FACILITY DEFINED

Terminology
For the purpose of analysis, the urban street is separated into individual
elements that are physically adjacent and operate as a single entity in serving
travelers. Two elements are commonly found on an urban street system: points
and links. A point is the boundary between Iinks and is usualIy represented by an
intersection or ramp terminaL A link is a length of roadway between two points.
A link and its boundary intersections are referred to as a segmento An urban street
facility is a length of roadway that is composed of contiguous urban street
segments and is typically functionally c1assified as an urban arterial or collector
street.

Facility Length Considerations

At least one intersection (or ramp terminal) along the facility must have a
type of control that can impose a legal requirement to stop or yield on the
through movement. A significant change in one or more facility characteristics
may indicate the end of one facility and the start of a second. These
characteristics indude cross-section features (e.g., number of through lanes,
shoulder width, curb presence), annual average daily traffic volume, roadside
development density and type, and vehide speed. One or more of these
characteristics will often change significantly when the street crosscs an urban-
to-suburban area boundary or intersects a freeway interchange.
lf a facility assessment is desired for a given travel mode, the analyst will
need to evaluate all oí the scgments that make up the fadlity for a common
travel direction and aggregate the performance measures for each segment to
obtain a facility performance estimate.

LOS CRITERIA
This subsection describes the LOS criteria for the motorized vehide,
pedestrian, bicyde, and transit modes. The crHeria for the motorized vehide
mode are difierent from the criteria used for the other modes. Specifically, the
criteria for the motorized vehide mode are bascd on performance measures that
are field-measurable and perceivable by travelers. With one exception, the
criteria for the pedestrian and bicyde modes are based on scores reported by
travelers indicating their perception of service quality. The exception is the
pedestrian space measure (used with the pedestrian mode), which is field-
measurable and perceivable by pedestrians. The criteria for the transit mode are
based on measured changes in transit patronage due to changes in service quality.

Olapter 16jUrban $treet Facilities

Con<,pts
VersiofJ 6.0
Page 16-6
 Highway Capadty Manual: A Guide for Multimodal Mobility Analysis

Motorized Vehicle Mode

AII uses DI the word "voIume"
Through-vehicle travel speed is used to characterize vehicular LOS for a or the phrase "vofu~to-
given direction of travel aJong an urban street facility. This speed reflects the capacity ratio" in this chapter
refer to demand vofume or
factors that influenee running time along eaeh link and the delay incurred by
demand.vofu~to-eapacity
through vehicles at eaeh boundary intersection. This performance measure ratio.
indica tes the degree of mobility provided by the facility. The "tollowing
paragraphs characterize eaeh service leve!.
LOS A describes primarily free-flow operation. Vehides are completely ''Free-fIow speed" is the
average running speed ot
unimpeded in their ability to maneuver within the traffie stream. Control del ay through vehides traveilng
at the boundary interseetions is minimal. The travel speed exeeeds 80% of the a/ong a segment under Iow-
vofume conditions and not
base free-flow speed, and the volume-to-eapacity ratio is no greater than 1.0. defayed by traffie controf
devices or other vehides.
LOS B describes reasonably unimpeded operation. The ability to maneuvcr
The ''base tree-f!ow spee(i" is
within the traffic stream is only slightly restricted, and control deJay at the define(} to be the tree-flow
boundary intersections is not significant. The travel speed is between 67% and speed on fonger segments.
80% of the base free-flow speed, and the volume-to-capacity ratio is no greater
than 1.0.

LOS e describes stable operation. The ability to maneuver and ehange lanes
at midsegment locations may be more restricted than at LOS B. Longer queues at
the boundary interseetions may contribute to lower travel speeds. The traveJ
spl->edis between 50% and 67% of the base free-flow speed, and the volume-to-
eapacity ratio is no greater than 1.0.
LOS O indica tes a less stable condition in which small ¡ncreases in flow may
cause substantial increases in delay and decreases in travel spced. This operation
may be due to adverse signal progrcssion, high volume, or inappropriate signal
timing at the boundary intersections. The travel speed is betwccn 40% and 50%
of the base free-flow speed, and the volume-to-capacity ratio is no greater than 1.0.
LOS E is characterized by unstable operation and significant delay. Such
operations may be due to sorne combination of adverse progression, high
volume, and inappropriatc signa] timing at the boundary intersections. The
traveJ speed is between 30% and 40% of the base free-flol'.' speed, and the
volumc-to-eapacity ratio is no greater than 1.0.
LOS F is characterized by flol'.' at extremely lol'.' speed. Congestion is Jikely
occurring at the boundary interseetions, as indicated by high delay and extensive
qucuing. The travel speed is 30% or Jess of the base free-flow speed, or the
voJume-to-capacity ratio is greater than 1.0.
Exhibit 16-3 lists the LOS thresholds established for the motorized vehicle
mode on urban strects. The threshold value is interpolated l'.'hen the base free-
flow speed is betl'.'cen the values shown in the column headings of this exhibit.
For example, the LOS A thrcshold for a segment with a base free-flow speed of
42 milh ;, 34 milh [, (42 - 40)/(45 - 40) , (36 - 32) + 321.

O1apter 16¡Urban 5treet Facilities Concepts

Version 6.0 Page 16-7
Highway Capacify Manual: A Guide for Mulfimodal Mobility Analysis

Exhibit 16.3 frave! Soeed Ihresho!d by Base Free-F!QW Speed {mi/hl Volume-to-
LOS Qíteria: Motorized LOS 55 50 45 40 35 30 25 Ca aci Ratio.
VehicleMode A >44 >40 >36 >32 >28 >24 >20 ~ LO
B >37 >34 >30 >27 >23 >20 >17
e >28 >25 >23 >20 >18 >15 >13
D >22 >20 >18 >16 >14 >12 >10
E >17 >15 >14 >12 >11 >9 >8
F ~17 :5:15 ~14 ~12 ~11 ~9 ~8
F An >1.0
No'" 'The critical vo/ume-to-eapacity ratio is based on ronsideratioo ot the through movement volume-to-
capacity ratio at each boundary Intersection in the subject dlrection of travel. The critical volume-to-
capacity ratio is the Iilrge5t ratio of!:hose considered.

Nonmotorized Vehicle Modes

Historically, this manual has used a single performance measure as the basis
foe defining LOS. However, rescaech documentcd in Chapter 5, Quality and
Level-of-Serviee Concepts, indicates that travelers consider a wide vanety of
factors in assessing the quahty of service provided to them. Sorne oí these factors
can be described as performance measures (e.g., spccd), and others can be
considered basic deseriptors of the urban strcet eharacter (e.g., sidewalk width).
The methodologies in Chapter 18, Urban Strcet Segments, and Chaptee 19,
Signalized Intersections, mathcmatically combine thesc factors into a scoee for
the segment or intersection, respeetively. This seore is then used in this chapter to
determine the LOS that is provided for a given direetion of travel along a facility.
Exhibit 16-4lists the range of seores associated with eaeh LOS foe the
pedestrian travel mode. The LOS for this mode is determined by consideration of
both the LOS seore and the average pedestrian space 00 the sidewalk. The
applieable LOS for an evaluation is determincd from the table by finding the
intersection of the row coeresponding to thc eomputed seore value and the
eolumn corresponding to the eomputed space value.

Exhibit 16-4 Pedestrian LOSby AveragePedestrjanSoace(tr/p)

i
LOS Oitefia:PedestrianMode LOSSCore >60 >40-60 >24--40 >15-24 >8.0-15i S8.0
~2.00 A B e o E F
>2.00-2.75 B B e o E F
>2.75-3.50 e e e o E F
>3.50-4.25 o o o o E F
>4.25-5.00 E E E E E F
>5.00 F F F F F F
Note: ' In Cn)SS.fIow situalioos, the LOS E/F threshokl iS 13 ft"/p. Chapter 4 describes the concept of ¥cross fIow"
aOOsituations where it shoulá be considere<!.

The assodation between LOS seoee and LOS is based on traveler pereeption
rescarch. Travelers wcre asked to rate the quality of scrvice assodated with a
specific trip along an urban street. The letter A was used to repeesent the best
quality of service, and the letter F was used to reprcsent thc worst quality of
service. "Best" and "worsf' were leh undefined, allowing respondcnts to
idcntify the best and worst conditions on the basis of their traveling experience
and peeeeption of serviee quality.
Exhibit 16-5 lists the range of seores that are associatcd with each LOS for the
bicyele and transit modes. This exhibit is also applicable for determining
pedestrian LOS when a sidewalk is not available.

Concepts Chaprer 16/Urban Street Fadlities

Version 6.0
Page 16-8
 Highway Capacity Manual: A Guide for Multimodal Mobmty Analysis

LOS lOS SCore Exhibit 16-5

A :S2.00 lOS Criteria: Bicycle and
B >2.00-2.75 Transit Modes
e >2.75-3.50
o >3.50-4.25
E >4.25-5.00
F >5.00

SCOPE OF THE METHODOLOGIES

This section identifies thc conditions for which each methodology is
applicable.
• BOlllldary intersectiolls. AH methodologies can be used to evaluate facility
performance with signalizcd or two-way STOP-controlledboundary
intersections. In the laUcr case, the cross street is STOI'-controlled.The
motorized vchicle methodology can also be used to evaluate performance
with all-way 5TOP-or YIElD-control1cd(e.g., roundabout) boundary
intcrsections.
• Street types. The four methodologics were developed with a focus on
arterial and collector street conditions. Jf a methodology is used to
evaluate a local street, the performance estima tes should be carefully
reviewcd for accuracy.
• Flow cOflditiOflS. The four methodologics are based on the analysis of
steady traffic conditions and are not well suited to the evaluation of
unsteady conditions (e.g.,congestion, cydic spillback, signal prcemption).
• Target road IIsers. Collectively, the four methodologies were developed to
estimate the LOS perceived by motorized vchicle drivers, pedestrians,
bicydists, and transit passengers. Thcy were not developed to provide an
estimate of the LOS perceived by other road users (e.g., commercial
vehicle drivers, automobile passengers, delivery truck drivers, or
recreational vehicle drivers). However, the perceptions of these other road
users are likely to be reasonably well represcnted by the road users for
whom the methodologies were devcloped.
• lllftllellces in fhe right-of-way. A road user's perception of quality of scrvice
is influenced by many factors inside and outside of the urban strcet right.
of.way. However, the methodologies in this chapter were specifically
constructed to cxclude factors that are outside of the right-of-way (e.g.,
buildings, parking lots, scenery, or landscaped yards) that might
influence a travcler's perspective. This approach was followcd because
factors outside of the right-of way are not under the direct control of the
4

agency operating the street.

L1MITATIONS OF THE METHODOLOGIES

The urban street facility methodology uses the performance measures
estimated by the segment and intersection methodologies in Chapters 18 to 23.
Thus, it incorporates the Iimitations of these methodologies (which are identified
in the respective segment or interscction chapter).

O1apter 16/Uroan 5treet Facilities Concepts

Version 6.0 Page 16-9
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

3. MOTORIZED VEHIClE METHODOLOGY

This section describes the methodology for evaluating the quality of service
provided to motorized vehicles on an urban street facility. Extensions to this
methodology for evaIuating more complex urban street operationaI elements are
described in Chapter 29, Urban Street Facilities: SupplementaI.

SCOPE OF THE METHODOLOGY

The overall scope of the four methodoIogies was provided in Section 2. This
section identifies the additionaI conditions for which the motorized vehicle
methodology is applicable.
• Target traveI modes. The motorized vehicle methodology addresses mixed
automobile, motorcyclc, truck, and transit traffic streams in which the
automobile represents the largest percentage of aH vehicles. The
methodology is not designed to evaluate the performance of other typcs
of vehicles (e.g., golf carts, motorized bicycles).
• MobiIity focus. The motorized vehicle methodology is intended to facilitate
the evaIuation of mobility. Accessibility to adjaccnt properties by way of
motorized vehicIe is not directly evaIuated with this methodoIogy.
Regardless, a street's accessibility shouId also be considered when its
performance is evaluated, especialIy if the street is intended to provide
such access. Often, factors that favor mobility reflect minimal Ievels of
access and vice versa.

Spatial and Temporal Limits

Analysis Boundaries
A facility evaluation considers An analysis of onIy one travel direction (when the street serves two-way
both diredionS of lTavel (when
the street serves two-way traffic) does not adequateIy recognize the interactions between vehicles at the
lTaffic). boundary intersections and their influence on segment operation. For this reason,
both travel directions on a two-way street shouId be evaIuated.
The analysis boundary for each boundary intersection is defined by the
operational influence area of the intersection. It should include the most distant
extent of any intersection-reiatcd queue expected to occur during the study
periodo For thcse reasoos, the influence area for a signalized intersection is Iikely
to extend at Ieast 250 ft back from the stop line on each intersection leg.

5tudy Perlad and Analysis Pen"od

Thc concepts of study periad and allalysis period are defined in Section 2 in
general terms. They are defined more precisely in this subsection as they relate to
the motorized vehicle methodology.
Exhibit 16-6 demonstrates three alternative approaches an analyst might use
for a given evaluation. Other altematives exist, and the study period can exceed
1 h. Approach A has traditionally been used and, unIcss otherwise justified, is
the approach that is recommended.

Motorized Vehicle Methodology Chapter 16jUrban 5treet Facilities

Page 16-10 VersiOn 6.0
 Highway Capacity Manual: A Guide tor MulfimodaJ Mobility Analysis

Aoproach A Approach B Approach e Exhibit 16-6

Three Altemative 5tudy
Study Period '" 1.0 h Study Period = 1.0 h Study Period = 1.0 h Approaches

Single analysis period Single aoalysis periOd Muttiple analysis periods

T=0.2Sh T=1.0h T=0.2Sh

Time

~ • aoalysis period

Approach A is based on evaluation of the peak lS-min period during the The use of peak 15-min trafflC
multipfied by tour íSprefen-ed
study periodo The analysis period T is 0.25 h. The equivalent hourly flow rate in for exíSting conditions when
vehicles per hour (veh/h) used for the analysis is based on either (a) a peak 15- traffic counts are available. The
use of a l-h demand voIume
min traffic count multiplied by four or (b) a l-h demand volume divided by the divided by a peak hour factor
peak hour factor. The former option is preferred for existing conditions whcn is preferred when voIumes are
projected or when hourly
traffic counts are available; the latter option is preferred when hourly volumes projected vofumes have been
are projected or when hourly projected volumes are added to existing volumes. added lo exlsting voIumes.

Additional discussion on use of the peak hour factor is provided in the

subsection titled Required Data and Sources.
Approach B is based on the evaluation of one 1-h analysis period that is
coinddent with the study periodo The analysis period, T, is 1.0 h. The flow rate
used is equivalent to the 1-h demand volume (Le., the peak hour factor is not
used). This approach implidtly assumes that the arrival rate of vehicles is
constant throughout the period of study. Therefore, the effects of peaking within
the hour may not be identified, and the analyst risks underestimating the delay
actually incurred.
Approach e uses a 1-h study period and divides it into four 0.25.h analysis
periods. This approach accounts for systematic flow rate variation among
analysis periods and for queues that carry over to the next analysis periodo It
produces a more accurate representation of delay. It is called "multiple time
period analysis" and is described in the next subsection.
Regardless of analysis period duration, a single-period analysis (Le.,
Approach A or B) is typical for planning applications.

Multiple Time Pedod Ana/ysis

lf the analysis period's demand volume exceeds capacity, a multiple time
period analysis should be undertaken in which the study period ineludes an
initial analysis period with no initial queue and a final analysis period with no
residual queue. On a movement.by-movement and intersection-by-intersection
basis, the initial queue for the second and subsequent periods is equal to the

Chapter 16/Urban Street: Facilities Motorized Vehicle Methodology

Version 6.0 Page 16-11
Highway Capacity Manual: A Guide for Multimodal Mobility Analys;s

residual queue from the previous periodo This approach provides a more
accurate estimate of the delay associated with the congestiono
If evaluation of multiple analysis periods is determined to be important, the
performance estimates for each period should be separately reported. In this
situation, reporting an average performance for the study period is not
encouraged because it may obscure extreme values and suggest acceptable
operation when in reality sorne analysis periods have unacceptahle operation.

Facility Length Considerations

Urban arterial and collcctor streets are designed to accommodate longer trips
than local strects. Thcy also have a significant rnobility function and support the
hierarchy of movement by connecting to streets of higher and lower functional
c1ass. An urban strcct facility with thcse attributes typically has a length of 1 mi
or more in downtown areas and 2 mi or more in other areas. The methodology
described in this chapter is focused on the evaluation of mobility on streets with
these charactcristics. Streets with shorter length may be evaluated with this
methodology; however, the primary function of these streets is likely to be access
to adjacent properties (as opposed to mobility).

Segment Length Considerations

The motorizcd vehicle methodology described in this section is not
appropriate for the analysis of "short" segments that are bounded by signalized
interscctions. In contrast, the methodology described in Chapter 23, Ramp
Terminals and Aitemative Intersections, is appropriate for the analysis of short
segmcnts at signalized interchangcs.
A short segment can expcrience "cyclic spillhack." This spillback occurs
when a queue extends back from one intersection into the other intersection (Le.,
spills back) during a portion of each signal cycle and then subsides. Specific
conditions under which a segment bounded by signalized intersections should
be considered "short" are difficult to define. General guidance in this regard is
provided in a similarly titled section in Chapter 18, Urban Strcct Segments.
The methodology described in this section is applicable to facilities made up
of segmcnts with each segment being 2 mi or less in length. This restriction is
based on the fact that STOP-,YIELD-,or signal-controlled intersections are Iikely to
have negligible effects on urban street operation when segment length exceeds 2
mi. Therefore, if a segment exceeds 2 mi in length, the analyst should evaluale the
scgment as an uninterrupted-flow highway segment with isolated intersections.

Performance Measures
Performance measures applicable to the motorized vehicle travel mode
includc travel speed and stop rate. LOS is also considered a performance
measure. It is useful for describing street performance to elected officials, palicy
makers, administrators, oc the publico LOS is based on travel speed and the
critical volume-to-capacity ratio.

Motorized Vehicle Methodology Chapter 16/Urban Street Fadlities

Page 16.12 Version 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

REQUIRED DATA ANO SOURCES

This subsection describes the input data needed for the motorized vehide
methodology. The data are Iisted in Exhibit 16-7 and are identified as "input data
elements." They must be separately spccified for each segment and for the
through-movement group at each boundary intersection.

D'la Potential Data Exhibit 16.7

category Location Input Data Element Basis Source(s) Required Input Data and
Potential Data Sources for
Geom"'" 5egment 5egment length 5egment
Fíeld data, aeria!
Motorized Vehicle Analysis
design photo
00"" 5egment Analysis penod duration Facility 5et by ana!yst
Boundary Through-movement
Volume-to-capacity ratio HCM method output
PerfOfTTlance intersection group
measures Base free-flow speed 5egment HCM method output
5egment
Travel speed 5egment HCM method output
Notes: Through-movemeot group " one value for the segment ttlrough movement at t:hedownstream boundary
intel"iection (inclusive of any tum movel11el1tsin a shared tane).
Segment " one value or condition far each segment and direction of travel on the facility.
Facility " one value or condition fOf the facility.

The next-to-last column in Exhibit 16-7 indicates whether the input data are
needed for a movement group at a boundary intersection, the segmcnt, or thc
facilit)'. The input data needed to evaluate the segment are idcntified in Chapter
18, Urban Street Scgments. Similarly, the input data nccded to evaluate the
boundary interscctions are identified in thc appropriate chapter (Le., Chapters 19
to 23).
í Segment Length
Segment length is the distance between the boundary intersections that
define the segmento The point of measurement at each intersection is the stop
line, the yield Hne, or the functional equivalent in the subjcct direction of travel.
This length is measured along the centerline uf thc strect. If it differs in the h •...
o
travel directiuns, an average length is used. Dne length is needed for each
segment on the facility.

Analysis Period Duration

The analysis period is thc time interval considered for the performance
evaluation. Its duration is in the range of 15 min to 1 h, with longer durations in
this range sometimes used for planning analyses. In general, the analyst should
use caution in interpreting the results from an analysis period uf 1 h or more
because the adverse impact of short peaks in traffie demand may not be deteeted.
Any 15-min period of interest can be evaluated with the methodolog)';
huwever, a complete evaluation shouId alwa)'s indude an analysis of conditions
during the 15-min period that experiences the highest traffic demand during a
24-h periodo
OperatiO/wl alla/ysis. A 15-min anal)'sis period should be used foc operational
analyses. This duration will accurately capture the adverse effects of demand
peaks.
Planlling ana/ysis. A 15-min anal)'sis period is used for most planning
analyses. However, a 1.h analysis period can be used, if appropriate.

Chapter 16/Urban Street Fadlities Motorized Vehicle Methodology

VersiQn 6.0 Page 16-13
Highway Capacity Manual: A Guide tor Multimodal Mobility Analysis

stream that reflects the running speed along the street foc through vehicles and
any deJay they may incuc at the boundary intersections. Thc tcavel speed for
through vehicles is determined for each segment by using the procedures
described in Chapter 17. Thc travel speed for the facility is calculated by using
Eguation 16-3:

Equation 16.]

where Su is the travel speed for the facility (rni/h), and ST.~.jis thc travel speed
of through vehicles for segment i(mi/h).

Step 3: Determine Spatial Stop Rate

The spatial stop rate for the facility is thc ratio of stop count to facility length.
It relates the number of full stops incurred by the average through vehicle to the
distance traveled. The spatial stop rate for through vehicles is deterrnined foc
each segment by using the procedures dcscribed in Chapter 17. The spatial stop
rate for the facility is calculated by using Eguation 16-4:

Equation 16-4
_ L~l HSeg•i L¡
H,- m
Li=l L¡
where HF is thc spatial stop rate for the facility (stops/mi), and H~; is thc spatial
stop rale for segrnent i (stops/mi).
The spatial stop rate from Equation 16-4can be used to estimate an
automobile traveler perception scoce for thc facility if desired. The eguations in
Step 10, Section 3, of Chapter 18 are used for this purposc. Thc value of HF would
be substituted for H~ in each equation. Similarly, the proportion of interscctions
with a left-tum lane Pm would be calculated for the entire facility and this one
value used in each eguation.

Step 4: Determine Motorized Vehicle LOS

LOS is determined for both directions of traveJ along the facility. Exhibit 16.3
lists the LOS thrcsholds established foc this purpose. As indicated in this exhibit,
LOS is defined by travel speed, where the LOS travel spced thresholds vary by
base free.flow speed. The base free-flow speed is computed in Step 1 and the
travel speed is computed in Step 2.
The footnote to Exhibit 16-3 indicales that volume-to-capacity ratio foc the
through movernent at the downstrcam boundary interscctions is also relevant to
the determination of facility LOS. This footnote indicates that LOS F is assigned
to the subject direction of travel if a volumc-to-capacity ratio greater than 1.0
cxists for the through movemcnt at one or more boundary intersections.
Facility LOS must be interpreted with caution. It can suggest acceptable
operation of the facility when in reality certain scgments are operating at an
unacceptable LOS. For each traveJ direction, the analyst should always verify
that each segment is providing acceptable operation and consider reporting the
LOS for the poorcst-perfonning segment as a means of providing context for the
interpretation of facility LOS.

Mototized Vehide Met:hodology Chapter 16/Urban Street Faálities

Page 16-16 V~6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

4. PEDESTRIAN METHODOLOGY

This section describes the methodology for evaluating the quality of service
provided to pedestrians traveling along an urban strcet.

SCOPE OF THE METHODOLOGY

Thc overall scope of the four methodologies was provided in Section 2. This
section identifies the additional conditions for which the pedestrian
methodology is applicable.
• Target lraveI modes. The pedestrian methodology addresses travel by
walking in the urban street right-of-way. It is not designed to evaluate the
performance of other travel means (e.g., Scg\',ra)',roller skatcs).
• "Typical pedestria,," /OCIIS. The pedestrian methodology is not designed to
retlect the perceptions of an)' particular pedestrian subgroup, such as
pedcstrians with disabilities. The performance measures obtained from
the mcthodology are not intended to be indicators of a sidewalk's
compliance with v.s. Access Board guidelines related to the Americans
with Disabilities Act requiremcnts. For this rcason, they should not be
considered as a substitute for a formal compliance assessment of a
pedestrian facility.

Spatial Limits
Side of Street to Be Eva/uated
Urban street performance from a pedestrian perspective is separately
cvaluatcd for each side of the strcct. Unless otherwise stated, all variables idelltijied
in this seclion are specijic lo fhe sllbject side o/lhe $freet. If a sidewalk is not available
for the subject side of the street, pcdestrians are assumed to walk in the street on
the subject side (even if there is a sidewalk on the other side).

Segment-8ased Eva/uation
The pedestrian methodology aggregates the performance of the segments
that make up the facility. In this regard, it considers the performance of each link
and boundary intersection. The methodologics for cvaluating the link and
boundary intersection are described in Chapters 18and 19, respectively.
The methodology is focused on the analysis of facilities with either signal-
controlled or two-way STOP.controlledboundary intersections. This edition of the
HCM does not inelude a procedure for evaluating a facility's performance when
a boundary intersection is an all.way sroP-controlled intersection, a roundabout,
or a signalized intcrchange ramp terminal.

Peñormance Measures
Performance measurcs applicablc to the pcdestrian travel mode inelude
pedestrian travcl specd, pedcstrian spacc, and pedestrian LOS score. The LOS
score is an indication of the typical pcdcstrian's pcrception of the overall facility
travel experience.

Qlapter 16/Urban Street Facilities Pedestrian Methodology

Version 6.0 Page 16-17
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

LOS is also eonsidered a performance measure. It is useful for describing

facility performance to elceted officials, policy makers, administrators, or the
publie. LOS is based on pedestrian space and pedestrian LOS seore.
"Pedestrian space" is the average amount of sidewalk area available to eaeh
pedestrian walking along the facility. A larger area is more desirable from the
pedestrian perspective. Exhibit 16-9 provides a qualitative description of
pedestrian space that can be used to evaluate sidewalk performance from a
circulation-area perspective.

Exhibit 16-9
QualitativeDesaiptien ef
Pedestrian Space
=~P~
.• ~.~'t~n~.
Random
I p)
a~n~s~p~a~,~.~(~ft'~
Platoon
Flow Flow Description
>60 >530 Abilityte move in desired path, no need te alter movements
>~ >90-530 Dccasionalneed lO adjust path to avoid conflicts
>24-40 >40-90 Frequent need to adjust path to avoid conflicts
>15-24 >23-40 Speed and abilityto pass slower pedestrians restricted
>8-15 >11-23 Speed restricted, very Iimitedabilityto pass slower pedestrians
:S8 <11 Speec! severely restricted, frequent contact with other users

The first two columns in Exhibit 16-9 indieate a sensitivity to flow condition.
Random pedestrian flow is typical of most facilities. Platoon flow is appropriate
for facilities made up of shorter segments (e.g., in downtown arcas) with
signalized boundary intersections.

REQUIRED DATA ANO SOURCES

This subsection describes the input data needed for the pedestrian
methodology. The data are listed in Exhibit 16-10 and are identified as "input
data elements." They must be separately specified for eaeh segment and
direction of travel on the facility. Segment length is defined in the subseetion
titled Required Data and Sources in $eetion 3.

Exhibit 16-10
Required Input Data and
=-----
--;o::a ••
Potential Data Source(s)
Category Location Input Data Element
Potential Data Sources fer Geometric 5egment 5egment length Reld data, aerial photo
Pedestrian Analysis design Presence of a sidewalk Reld data, aerial photo
Performance 5egment Pedestrian space HCMmethod output
measures Pedestrian travel speed HCMmethod output
Pedestrian lOS score for segment HCM method output

Presence of a Sidewalk
A sidewalk is a paved walkway that is provided at the side of the roadway.
Pedestrians are assumed to walk in the street if a sidewalk is not presento An
indication of sidewalk presenee is needed for each side of interest for each
segment on the facility.

Pedestrian Space
Pedestrian spaee is a performance measure that describes the average
circulation arca available to eaeh pedestrian traveling along the sidewalk. A
procedure is described in Chapter 18 for estimating this quantity for a given
sidewalk. One value is needed for each sidewalk of interest associated with eaeh
segment on the facility.

Pedestrian Methodology Chapter 16/Urban5tIeet Facilities

Page 16-18 VersiorI 6. O
 Highway Capacify Manual: A Guide for Multimodol Mobility Anolysis

Pedestrian Travel Speed

Pedestrian travel speed is the ratio of segmcnt length to pedestrian travel
time. Travel time is computed as the sum of segment walking time and control
detay at the downstream boundary intersection. A proccdurc for computing this
travel speed is dcscribed in Chapter 18. One speed is needed for each sidewalk of
interest associated with each segment on the facility.

Pedestrian LOS Score for Segment

Thc pedestrian LOS seore foc the segment is used in the pcdcstrian
methodology to determine facility LOS. It is obtained from the pedestrian
methodology described in Chapter 18. One score is needed for each direction of
travei of interest for each segment on the facility.

OVERVIEW OF THE METHODOLOGY

This subsection describes the methodology for evaluating the performance of
an urban street facility in terms of its scrvicc to pcdestrians. The methodology is
applicd through a series of four steps that culminate in the determination of the
facility LOS. These steps are iIlustrated in Exhibit 16-11.

Exhibit 16-11
5tep 1: Determine Pedestrian Space PedestIian Methodology for
Urban 5treet Facilities

Step 2: Determine Pedestrian Travel Speed

Step 3: Determine Pedestrian LOS $eore

Step 4: Determine Pedestrian LOS

A methodology for evaluating off-street pedestrian facilities is provided in

Chapter 24, Off.Street Pedestrian and Bicycle Facilities. Off.street facilities
inelude those for which the characteristics of motor vehiele traffic do not playa
strong role in determining quality of service from the perspective of pedestrians.

COMPUTATIONAL STEPS
Step 1: Determine Pedestrian Space
Pedestrians are sensitive to the amount of space separating them from other
pedestrians and obstades as they walk along a sidewalk. Average pedestrian
space is an indicator of facility performance for travcl in a sidewalk. This step is
applicablc only when the sidewalk exists on the subject side of the street.
The pedestrian space is determined fOfcach segmcnt by using thc
procedures dcscribed in Chapter 18, Drban Street Scgmcnts. The pcdcstrian
space for the facility is calculated by using Equation 16-5:

Chapter 16/Urban Street Facilities Pedestrian Methodology

Version 6.0 Page 16-19
Hjghway Capacity Manual: A Gu;de for Multimodal Mobility Analysjs

Equation 16-5

where
Ap.F pedestrian space for the facility (ft2/p),
L¡ length of segment i (ft),
m number of segments on the facility, and
Ap.; pedestrian space for segment i (ft2jp).
The pedestrian space for the facility reflects the space provided on the
sidewalk along the segmento It does not consider the comer circulation area or
the crosswalk circulation area at the intersections. The analyst should verify that
the intersection comers and crosswalks adequately aeeommodate pedestrians by
using the methodology in Section 5 of Chapter 19.

Step 2: Determine Pedestrian Travel Speed

The travel speed for the facility is the ratio of facility length to facility trave!
time. It represents an equivalent average speed of pedestrians that refleets their
walking speed along the sidewalk and any delay they may incur at the boundary
intersections. The travel speed for pedestrians is determined for eaeh segment by
using the procedures deseribed in Chapter 18. The pedestrian travel speed far the
faeility is ealculated by using Equation 16.6:

Equation 16-6 ¿f;1 Li

STP.F = L.
~m !
Li=lS .
Tp.seg.!

where Srp,F is the travel speed of through pedestrians for the facility (ft/s), and
STp,~,; is the travel speed of through pcdestrians for segment i (ft/s).
In general, a travel spced of 4.0 ft/s or more is eonsidered desirable, and a
speed of 2.0 fl/s or less is eonsidered undesirable.

Step 3: Determine Pedestrian LOS Score

The pedestrian LOS seore for the facility is eomputed in this step. It is a
travel-time-weighted average of the pedestrian LOS seores for the individual
Iinks and intersections that make up the facility. The LOS seore for eaeh segment
(determined by using the procedures described in Chapter 18) is used to compute
the total weighted seore for the facility. The total travel time far the facility is
eomputed by using the trave! speed from Step 2. The LOS seore for the facility is
computed by using the ratio of the total weighted score and the total travel time.
This seore is ealculated by using Equation 16-7:
,
rt;l WTTp,1
Equation 16-7 Ip.F = 0.75 L~ L.
r (s;~/) f + 0.125

Pedestrian Methodology Chapter 16/Urban Street Facilities

Page 16-20 Version 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

with

WTTpi
_ ( L¡ ) (I",g .•- 0.125)3
- --- Equation 16-8
'STp,seg.i 0.75

pedestrian LOS score for the facility,

travel-time-weightcd average pedestrian LOS score for segment i, and
J p.«'g.l . pedestrian LOS score for segment i,

Step 4: Determine Pedestrian lOS

The pedestrian LOS for the facility is determined by using the pedestrian
LOS seore from Step 3 and the average pedestrian spaee from Step 1. These two
performance measures are eompared with their respective thresholds in Exhibit
16-4 to determine the LOS for the specified direction of travel along the subjeet
facility. If the sidewalk does not exist and pcdestrians are relegated to walking in
the street, LOS is determincd by using Exhibit 16-5 because the pedestrian space
concept does not apply.
Facility LOS must be interpreted with caution. It can suggest aceeptable
operation of the facility when in reality certain segments are operating at an
unaeceptable LOS. For caeh travel direction, the analyst should always verify
that eaeh segment is providing acccptable operation and eonsider reporting the
LOS for the poorest-performing segment as a means of providing eontext for the
interpretation of facility LOS.

Chapter 16/Urban 5treet Facilities Pedestrian Methodology

Version 6.0 Page 16-21
Highway Capacify Manual: A Guide for Multimodal Mobility Analysis

5. BICYCLE METHODOLOGY

This section describes the methodology for evaluating the quality of service
provided to bicydists traveling along an urban street.

SCOPE OF THE METHODOLOGY

The overall scope of the four methodologies was provided in Section 2. This
section identifies the additional conditions for which the bieyde methodology is
applicable.
• Target travcl modes. The bicyde methodology addresses travel by bicyde in
the urban street right-of~way. It is not designed to evaluate the
performance of other travcl means (e.g., motorized bicyde, rickshaw).
• Shared or exclusive lalles. The bicyde methodology can be used to evaluate
thc service provided to bicydists when they share a lane with motorized
vehicles or when they travel in an exelusive bicyele lane.

Spatial Limits
Trave/ Directions to Be Eva/uated
Urban street facility performance from a bicyclist perspective is separately
evaluated for cach travel direction along the street. Un/ess otherwise stated, alf
variables identified in this sectio" are specifie to the subject direetioll o/ trave/. The
bicyde is assumed to travel in the strcct (possibly in a bicyele lane) and in the
same direction as adjacent motorized vehides.

Segment-aased Eva/uation
The bicycle methodology aggregates the performance of the segments that
make up the facility. In this regard, it considers the performance of each link and
boundary intersection. The methodologies for evaluating the link and boundary
interscction are described in Chapters 18 and 19, respectively.
The methodology is focused on the analysis of a facility with either signal-
controlled or two-way sroP-controlled boundary intcrsections. This edition of the
HCM does not inelude a procedure for evaluating a facility's perfonnance when
a boundary interscction is an all-way STOP-controlled intersection, a roundabout,
or a signalized interchange ramp terminal.

Performance Measures
Performance measures applicable to the bicyde travel mode indude bicyde
travel speed and bicyele LOS seo re. The LOS score is an indication of the typical
bicyclist's perception of the overall segment travel expcrience.
LOS is al50 considercd a performance measure. It is useful for describing
segment performance to elected officials, policy makers, administrators, or the
public. LOS is bascd on the bicyclist LOS score.

Bicycle Methodology Chapter 16/Urban 5treet Facilities

Page 16-22 Version 6.0
 Highway Capacity Manual: A Gujde for Multimodal Mobility Analysis

limitations of the Methodology

This subsection identifies a known limitation oi the bicycle methodology.
Specifically, thc mcthodology is not applicable whcn the bicycle lanes occur
intermittently along the facility. If this condition is present, the analyst should
eonsider using alternative methods or tools for the evaluation.

REQUIRED DATA AND SOURCES

This subseetion describes the input data nceded for the bicycle methodology.
The data are listed in Exhibit 16-12and are identified as "input data elements."
They must be separately specified for each segment and direction oi travel on the
iacHity. Segment Icngth is defined in the subsection titled Required Data and
Sources in Seetion 3.

Data category Location Input Data Element Potential Data Source(s) Exhibit 16-12
Geometric design 5egment 5egment length Field data, aerial photo Required Input Data and
Performance 5egment Bicycie travel speed HCM methex:! output Potential Data Sources for
Bicycie Analysis
measures Bisycle lOS score for segment HCM method Otltput

Bicycle Travel Speed

Bicycle travel specd is the ratio of segment length to bicycle travel time.
Travel time is computed as the sum oi segment running time and control delay
at the downstream boundary intersection. This speed is computed only when a
bicycle lane is present on the segmcnt. A proeedure for computing this trave!
speed is described in Chapter 18. One speed is needed ior each direetion oi Iravel
of interest for eaeh segmcnt on the iacility.

Bicycle LOSScore for Segment

The bicycle LOS score for the segment is used in the bicycle methodology to
estimate facility LOS. It is obtained irom the bicycle methodology in Chapter 18.
One score is needed ior each direction oi travel oi interest ior eaeh segment on
the facility.

OVERVIEW OF THE METHODOLOGY

This subsection describes the methodology for evaluating the performance of
an urban street facility in terms oi its service to bicyclists. Thc methodology is
applied through a series of three steps that culminate in the determination oi the
facility LOS. These steps are illustrated in Exhibit 16-13.

Exhibit 16.13
Step 1: Determine Bicycle Travel Speed Bicycie Methodology for Urban
Street Facilities

Step 2: Determine Bícycle LOS Score

Step 3: Determine Bicycle LOS

Chapter 16jUrban Street Facilities Bicycle Methodology

VersiOn 6.0 Page 16-23
Highway Capacity Manual: A Guide for Mulfimodal Mobility Analysis

A methodology for evaluating off-strcet bicyde facilities is provided in

Chapter 24, Off-Street Pedestrian and Bicycle Facilities. Off-street facilities
include those for which the charaderistics of motor vehicle traffic do not playa
strong role in determining quality of service from the perspective of bicydists.

COMPUTATIONAL STEPS
Step 1: Determine Bicycle ,!ravel Speed
lOe travel speed foc the faeility is the ratio of faeility length to facility travel
time. It represents an equivalent average speed of bicydes that reflects their
running speed along the strcet and any delay they may incur at the boundary
interseetions, lOe travel spced for bicydes is determined for each segment by
using the proeedures deseribed in Chapter 18. The bicyde travel speed foc the
facility is ealculated by using Equation 16-9;

Equation 16-9
,E~1
L¡
STb.F = L.
,Em !
i=l STb,seg,i

wheee
5TH travel speed of through bicycles for the facility (mi/h),

L, length of segment i (ft),

m number of segments on the faeility, and
travel speed of through bicycles for segment i (mi/h).

Step 2: Determine Bicycle LOSScore

The bicycle LOS seore foc the fadlity is eomputed in this step.1t is a travel-
time-weighted average of the bicycle LOS seores for the individuallinks and
interseetions that make up the facility. The LOS seoee foe eaeh segmcnt
(determined by using the procedures described in Chaptee 18) is uscd to compute
the total weighted seore for the faeility, lOe total travel time for the fadlity is
eomputed by using the travel speed from Step 1. The LOS seore foe the faeility is
eomputed by using the ratio of the total weighted score and the total travel time.
This seore is ealculated by using Equation 16-10:
1

~ wrr b.i
Equation 16-10

with
¡
b,F
= 075
•

r
,E1=l
(,E~1L¡)
STb,F r + O 125
o

Equation 16-11 WTTbi _- ( --- L, ) (lb_e

------
Bo' - 00125)'
'STb.seg,¡ 0.75

where h.r is the bicycle LOS scoce foc the facility, \¥ITb.i is the teavel-time-
weighted average bicycle LOS scoce for segment i, and lb.~.iis the bicycle LOS
scoce for segment i.

Bicycle Methodology O1apter 16/Urban 5treet Fadlities

Page 16-24 Version 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Step 3: Determine Bicycle LOS

The bicycle LOS for the facility is determined by using the bicycle LOS score
from Step 2. This performance measure is compared with the thresholds in
Exhibit 16.5 to determine the LOS for the specified direction of travel along the
subject facility.
Facility LOS must be interpreted with caution. lt can suggest acceptable
operation of the facility when in reality certain segments are operating at an
unacceptable LOS. For each traveI direction, the analyst should always verify
that each segment is providing acceptable operation and consider reporting the
LOS for the poorest-perfonning segment as a means of providing context for the
interpretation of facility LOS.

Chapter 16{Urban 5treet Facilities Bicyde Methodology

Versk!n 6.0 Page 16-25
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

6. TRANSIT METHOOOLOGY

This section describes the methodology for evaluating the quality of service
provided to transit passengers travcling along an urban strect.

SCOPE OF THE METHOOOLOGY

The overall scope of the four methodologies was provided in Section 2. This
section identifies the additional conditions for which the transit mcthodology is
applicable. Spccifically, the transit methodology is limitcd to the evaluation of
public transit vehicles operating in mixed or exclusive traffic lanes and stopping
along the street. It is not designed to evaluate the performance of other travel
means (e.g., grade-separated rai! transit).

Spatial Limits
Trave! Directions to Be Eva/uated
Urban street facility performance from a transit passenger perspective is
scparately evaluated for cach travel direction along the strcet. Unless ofherwise
sfated, all variables identijied in this section are specijic to the sllbject direction o/ traveI.

Route-Based Eva/uation
The methodology is used to evaluate a single transit route on the fac¡lity. If
multiple routes exist on the facility, each route is evaluated by using a separate
application oí the methodology.

Peñormance Measures
Performance measurcs applicable to the transit travel mode include transit
vehicle travel speed and transit LOS score. The LOS score is an indication of the
typical transit rider's perception of the overall travel cxperience.
LOS is also considered a performance measure. It is useful for describing
segment performance to elected officials, policy makers, administrators, or the
publico LOS is based on the transit LOS score.

REQUIREO OATA ANO SOURCES

This subsection describes the input data needed for the transit mcthodology.
The data are Usted in Exhibit 16-14and are identified as "input data clements."
They must be separately specified for eaeh segment and direction of travel on the
facility. Segment Iength is defined in thc Required Data and Sourees subsection
in Scction 3.

Exhibit 16-14 Data category Location Input Data Element Potential Data Source(s)
Required Input Data and Geometric design 5egment 5egment length Field data, aerial photo
Potential Data Sources for Performance 5egment Transit travel speed HCM method output
Transit Analysis measures Transit lOS score for segment HCM method output

Transit Methodology Chapter 16/Urban Street Facilities

Page 16-26 V~6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Transit Travel Speed

Transit travel speed is the ratio of segment length to transit travel time.
Travel time is computed as the sum oE segment running time and control deiay
at the downstream boundary intersection. A procedure Eorcomputing this travel
speed is described in Chaptcr 18. One speed is nccded far each direction of travel
oE interest Eoreach segment on the facility.

Transit LOS Score for Segment

The transit LOS score for the segment is used in the transit methodology to
estimate fadlity LOS. lt is obtaincd from the transit methodology in Chapter 18.
One score is needed for cach direction of travel of interest far each segment on
the Eacility.

OVERVIEW OF THE METHODOLOGY

This subsection describes the methodology far evaluating the performance of
an urban strcet facility in terms of its sen.ice to transit passengers. The
methodology is applied through a series of thrt.. 'C steps that culminate in thc
determination of facility LOS. These stcps are illustrated in Exhibit 16-15.

Exhibit 16-15
Step 1: DetermineTransit Travel Speed Transit Methodology for Urban
5treet Facilities

Step 2: DetermineTransit LOSScore

Step 3: DetermineTransit LOS

Procedures for estimating transit vehicle performance on grade-separated or

non-public-street rights-of-way, along with procedures for estimating origin-
destination service quality, are provided in the Transit Capacity and Qualíty o/
SeroiceManllal (3).

COMPUTATIONAL STEPS

Step 1: Determine Transit Travel Speed

The travel speed for the facility is the ratio of facility length tu facility travel
time. It represents an equivalent average spced uf transit vehicles that reflects
their running speed along the street and any delay they may incur at the
boundary intersection. The travel spt..'Cdfor a transit vehicle is determined for
each segment by using the procedures describcd in Chapter 18. The transit travel
speed for the facility is calculated by using Equation 16-12:

Chapter 16¡Urban 5treet Facilities Transit Methodology

V~6.0 Page 16-27
Highway Capacity Manual: A Guide for Mulfimodal Mobility Analysis

where
5nf travel speed of transit vehicles for the facility (mi/h),

L; length of scgment i (ft),

m number of segments on the fadlity, and

travel speed of transit vehicles for segment i (mi/h).

Step 2: Determine Transit LOS Score

The transit LOS score for the fadlity is computed in this step. lt is a length-
weighted average of the transit LOS seore for the individual segments that make
up the fadlity. The segment seores are determined by using the proeedures
described in Chapter 18. The seore for the fadlity is ealculated by using Equation
16-130
L~llc.seg,i L¡
Equation 16-13 It,F = '\'m L
L..í=l 1

where I,.f is the transit LOS seore for the fadlity, and I',~i is the transit LOS score
for segment i.

Step 3: Determine Transit LOS

The transit LOS for the facility is determined by using the transit LOS seore
froro Step 2. This performance measure is compared with the thresholds in
Exhibit 16-5 to determine the LOS for the specified direction of travel along the
subjeet facility.
Fadlity LOS must be interpreted with caution. lt ean suggest aeeeptable
operation of the facility when in rcality eertain segments are operating at an
unacceptable LOS. For each travel direetion, the analyst should always verify
that each segment is providing aeceptable operation and eonsider reporting the
LOS for the poorest.perfonning segment as a means of providing context for the
interpretation of facility LOS.

Transit MethodoIogy Chapter 16{Urban 5treet Facilities

Page 16-28 VefS'Of¡ 6.0
 Highway Capacity Manuol: A Guide for Multimodal Mobility Analysis

7. APPLICATIONS

EXAMPLE PROBLEMS
Chapter 29, Urban Street Facilities: Supplemental, describes the application
of each of the four methodologies through the use of example problems. The
focus of the examples is on iIlustrating the multimodal facility evaluanon
process. An operational aoalysis level is used for all examples.

GENERALlZED DAILY SERVICE VOLUMES

Generalized daily service volume tables provide a means of assessing a large
number of urban streets in a region or jurisdiction quickly to determine which
facilities need to be assessed more carefully (by using operational analysis) to
ameliorate existing or pending problems.
To build a generalized daily service volume table for urban street facilities, a
number of simplifying assumptions must be made. The assumptions made here
include the foUowing:
• AH segments of the facility have the same number of through lanes (one,
two, or three) in each direction.
• Only traffic signal control is used along the facility (Le., no roundabouts
or all-way STOP-controlledintersections exist).
• The traffic signals are coordinated and semiactuated, the arrival type is 4,
the traffic signal cycle time C is 120 s, and the weighted average green-to-
cycle-Iength (g/C) ratio for through movements (defined below) is 0.45.
• Exclusive left-tum lanes with protected left-turn phasing and adequate
queue storage are provided at each signalized intersection, and no
exclusive right-tum lanes are provided.
• At each traffic signaJ, 10%of the traffic 00 the urban street facility turns
left and 10% turns right.
• The peak hour factor is 0.92.
• The facility length is 2 mi, and no restrictive medians exist along the
facility.
• The base saturation flow rate So is 1,900passenger cars per hour per lane
(pc/hnn).
The weighted average g/C ratio of an urban street is the average of the critical
intersection through glC ratio and the average of all the other glC ratios for the
urban street. For example, if there are four signals with a through g/C ratio of
0.50 and one signal with a through glC ratio of 0.40, the weighted average glC
ratio for the urban street is 0.45. The weighted glC ratio takes into account the
adverse effect of the critical intersection and the overall quality of flow for the
urban street.
Generalized daily service volumes are provided in Exhibit 16-16 far urban
street facilities with posted speeds of 30 and 45 mi/h; two, four, or six lanes (both

Chapter 16/Urban 5treet Facilities Applications

Vers.oo 6.0 Page 16-29
Highwoy Copodty Monual: A Guide tor Multimodal MobiJjty Analysis

dircctions); and six combinations of the K.factor and D-factor. To use this table,
analysts must sclect a combination of K and D appropriate for their locality.
The 30-mi/h values further assume an average traffic signal spacing of
1,050 ft and 20 access points/mi, while the 45-mi/h values assume an average
traffic signal spacing of 1,500 ft and 10 access points/mi.

Exhibit 16.16 ;1 ; lu v h
GeneralizedDailyService K. O-
Two.Lane Streets Four-Lane Streets Six.lane Streets
Volumesfor Urban Street Facto Facto
~B~C~D~SE~SB~SC~D~SE~SBWSC~D~SE
Facilities
-< _ 30 mi 'h
0.55 NA 1.7 11.8 17.8 NA 2.2 24.7 35.8 NA 2.6 38.7 54.0
0.09
0.60 NA 1.6 10.8 16.4 NA 2.0 22.7 32.8 NA 2.' 35.6 49.5
0.55 NA 1.6 10.7 16.1 NA 2.0 22.3 32.2 NA 2.4 34.' 48.6
0.10
0.60 NA 1.4 '.8 14.7 NA 1.8 20.4 29.5 NA 2.2 32.0 44.5
0.55 NA 1.4 '.7 14.6 NA 1.8 20.3 29.3 NA 2.1 31.7 44.1
0.11 NA 1.3 13.4 NA 1.7 18.6 26.9 NA 2.0 29.1 40.5
0.60 8.'
-< - 45m''h
0.55 NA 7.7 15.9 18.3 NA 16.5 33.6 36.8 NA 25.4 51.7 55.3
0.09
0.60 NA 7.1 14.5 16.8 NA 15.1 30.8 33.7 NA 23.4 47.4 50.7
0.55 NA 7.0 14.3 16.5 NA 14.9 30.2 33.1 NA 23.0 46.5 49.7
0.10
0.60 NA 6.' 13.1 15.1 NA 13.6 27.7 30.3 NA 21.0 42.7 45.6
0.55 NA 6.3 13.0 15.0 NA 13.5 27.5 30.1 NA 20.9 42.3 45.2
0.11 12.4 25.2 27.6 NA
0.60 NA 5.8 11.9 13.8 NA 19.1 38.8 41.5
Notes: NA,. notapplicable;LOScannotbeachievedwittlthe statedassumptions.
Generalassumptionsincludenoroundaboutsor an~waySTQP-O.lf1troHe(l íntersectiorls
alongthe facility;
coor(linated,
semíactuatedtrafficsignals;Arriva!
Type4; 120-scycletime;protecte<lleft-turn phases;0.45
weightedaveragewCratlo; exclusiveleft.turnlaneswithadequateqoeoestorageprovidedat trafflC
signals;noexclusiveright-turnlanesprovióed;no restrietivemeclian;2-mlfadlitylength;10%oftrafflC
turnsleftan(l10%tumsrightat eachtr'afficsignal;peal<.hourfactor= 0.92;aOObasesaturationflowrate
= 1,900pc/h/ln.
A(I(Iitional
assumptionsfor3O-mi/hfacilities:
signalspaclng= 1,050ftaOO20<HXeSS poínts,'mí.
Additionalassumptlonsfor45-mi/hfacilities:
signa!spacing= 1,500Il:aOO10accesspointslml.

Exhibit 16-16 is provided for general planning use and should nof be used to
analyze any spedfic urban street facility or to make final decisions on important
design features. A full opcrational analysis using this chapter's methodology is
required for such specific applications.
The exhibit is useful in evaluating the OYeran performance of a large number
of urban strects within a jurisdiction, as a first pass to determine wherc problems
might exist or arise, or in determining where improvements might be needed.
However, any urban strcct identified as Iikely to experience problems or need
improvement should be subjectcd to a full operational analysis before any
dedsions on implementing spccific improvements are made.
Daily service volumes are strongly affected by the K- and D-factors choscn as
typical for the analysis. The values used for the facilities under study should be
reasonable. Also, if any characteristic is significantly different from the typical
values used to develop Exhibit 16-16, particulariy the weighted average g/C ratio
and traffic signal spacing, the values taken from this exhibit will not be
representativc of the study facilities. In such cases, analysts are advised to
develop their own generalized service volume tables by using representative
local values or to proceed to a full operational analysis.

Applications Chapter 16/Urban Street Facilities

Page 16-30 Vern¥l6.0
 Highway Capacity Manual; A Guide tor Multimodal Mobility Analysis

ANAL YSIS TYPE

The four methodologies described in this chapter can each be used in three
types of analysis. Thesc analysis typcs are described as operational, design, and
planning and preliminary engineering. The selected analysis type applies to the
methodology described in this chapter and to all supporting methodologies. The
characteristics of each analysis type are deseribed in the subsequent paragraphs.

Operational Analysis
The objeetive of an opcrational analysis is to determine the LOS for current
or near-term conditions when details of traffie volumes, geometry, and traffic
control are known. AH the methodology steps are implementcd and all
calculation procedures are applied to compute a wide range of performance
measures. The operational analysis type will provide the most reliable results
because it uses no (or mini mal) dcfault values.

Design Analysis
The objective of a design analysis is to identify the alterna tives that operate
al the target level of the specified performance measures (or provide a better
Icvel of performance). The analyst may then recommend the "best" design
alternative after consideration of the full range of factors.
The nature of the design analysis type depends on whether the boundary
intersections are unsignalized or signalized. When the facility has unsignalized
boundary intersections, the analyst specifies traffic conditions and target I('vels
for a set of performance measures. The methodology is then applied by using an
iterative approach in which alternativc geometric conditions are separately
evaluated.
When the facility has signalized boundary intersections, the design analysis
type has two variations. Both require the specification of traffic conditions and
target levels for a set of performance measures. One variation requires the
additional specification of the signalization conditions. The methodology is then
applied by using an iterative approach in which altcrnative geometric conditions
are separately evaluated.
The second variation of the design analysis requires the additional
specification of geometric conditions. The mcthodology is then applicd by using
an iterative approach in which alternative signalization conditions are evaluated.

Planning and Preliminary Engineering Analysis

The objective of a planning and preliminary enginecring analysis can be (al
to determine the LOS for either a proposed facility ar an existing facility in a
future year or (b) to size the overall geometrics of a proposed facility.
The level of precision inherent in planning and preliminary engineering
analyses is typically lower than for opcrational analyscs because default values
are often substituted for ficld-measured values of many of the input variables.
Recommended default values for this purpose are provided in Chapters 18 to 23.

Chapter 16/Urban 5treet Facilities Applications

Version 6.0 Page 16-31
Highway Capacity Manual: A Guide for Mulfimodal Mobility Analysis

USE OF ALTERNATlVE TOOLS

Chapter 29, Urban Street Facilities: Supplemental, indudes a set of examples
illustrating the use oí alternative tools in addressing the stated limitations of this
chapter and Chapter 18, Urban Street Segments. Specifically, these examples are
used to iIIustrate (a) the application of detenninistic tools to optimize the signal
timing, (b) the effect of midsegment parking maneuvers on facility operation, (e)
the effect oí using a roundabout as a segment boundary, and (d) the use of
simulated vehide trajectories to evaluate the proportion of time that the back of
the queue on the minor-street approach to a two-way sToP-controlled
intersection exceeds a specified distance from the stop lineo

Applications Chapter 16/Urban 5treet Facilities

Page 16-32 VetSkln 6.0
 Highway Capacity Manual: A Guide for MultjmodaJ Mobility Analys;s

8. REFERENCES

These references can be found

1. Bonneson, J. A., M. P. Pratt, and M. A. Vandehey. Predicting the Performallceof
in the Tedmical Reference
Automobile Traffic 011 Urban Streets: Filial Rl'P0rt. National Coopcrativc Library in VoIume 4.
Highway Research Program Pruject 3-79. Texas Transportation Institute,
Texas A&M University, College Station, Jan. 2008.
2. Dowling, R. G., D. B. Reinke, A. Flannery, P. Ryus, M. Vandehey, T. A.
Petritsch, B. W. Landis, N. M. Rouphail, and J. A. Bonneson. NCHRP Report
616: Multimodal Leve! of Semice Allalysis fár Urball Streets. Transportation
Research Board of the National Academies, Washington, D.C., 2008.
3. Kittelson & Associates, Ine.; Parsons Brinckerhoff; KFH Group, lnc.; Texas
A&M Transportation Institute; and Arup. TCRP Report 165: Transit Capacity
alld Quality ofSemice Manual, 3rd ed. Transportation Rcscarch Board of the
NationaI Academies, Washington, D.C., 2013.

Chapter 16/Urban Street Facilities References

Version 6.0 Page 16-33
Highway Capacity Manual: A Guide for Mulfimodal Mobility Analysis

LIST OF EXHIBITS

Exhibit 17-1 ATDM Tactics and Measures for Urban Streets 17-5
Exhibit 17-2 High-Level Representation of the Method for Estimating the
Travel Time Distribution 17-8
Exhibit 17-3 General Data Categories Required for a Reliability
Evaluation 17-12
Exhibit 17-4 Temporal and Spatial Dimensions of Reliability 17-14
Exhibit 17-5 Dcfault Hour-of-Day Demand Ratios (ADT/AADT) 17-16
Exhibit 17-6 Default Day-of-Week Demand Ratios (ADT/AADT) 17-16
Exhibit 17-7 Default Month-of- Year Demand Ratios (ADT/AADT) 17-16
Exhibit 17-8 Default Values for Weather Events 17-17
Exhibit 17-9 Dcfault Values for Inddents 17-19
Exhibit 17-10 Default lnddent Clearance Times 17-20
Exhibit 17-11 Default lnddent Distribution with Shoulder Presence 17-22
Exhibit 17-12 Default Inddent Distribution Without Shoulder Presence 17-22
Exhibit 17-13 Rcliability Methodology Framework 17-25
Exhibit 17-14 Interrelationship Between Causes of Congestion and the
Facility 17-31
Exhibit 17-15 Example Matrix Allocating Annual Vehide Hours of Delay
by Cause 17-32
Exhibit 17-16 Example Pie Chart of Congestion Causes 17-32
Exhibit 17-17 Student's t-Statistic 17-39
Exhibit 17-18 Use Cases for Travel Time Reliability 17-40

Contents O1apter 17/Urnan Street Reliability aOO ATDM

Page 17-ji Version 6.0
 Hjghway Capadty Manual: A Guide for Multimodal Mobility Analysjs

1. INTRODUCTION

OVERVIEW VOLUME 3; INTERRUPTED FLOW

16. Urban Street Fadlities
This chapter describes a methodology for evaluating the travel time 17. Urnan 5lreet Reliability
and ATDM
reliability experienced by motorists on an urban street facility. Travel time 18. Urban Slreet Segments
reliability retlects the distribution of trip travel time over an extended periodo 19. Signalized lntersections
20. TWSC Intersections
The distribution arises from the occurrence of a number of factors that intluence 21. AWSC Intersections
travel time (e.g., weather events, incidents, work zone presence). The distribution 22. Roundabouts
23. Ramp Terminals and
describes how o/tell these factors occur and how bad operations are as a resulto Alternative Intersections
24. Off.Street Pedestrian and
The methodology's reliability performance measures can be used for a
Bicycle Facilities
variety of facility management functions. For example, they can be used as the
basis for quantifying the degree of severity of Level of Service (LOS) F
(oversaturated) conditions. They can also be used for developing agency
performance standards for oversaturated facilities. Finally, they can be used to
quantify the impacts of physical and operational treatments designed to improve
travel time reliability.
The methodology is focused on the evaluation of an urban street facility
(with consideration of the segmen!s that make it up). It is used with the
methodologies in other Highway Capacity Manual (HCM) chapters to compute the
desired performance measures. Specifically, the methodology for aggregating
segment performance measures to obtain an estimate of facility performance is
described in Chapter 16. The methodology for evaluating the individual
segments is described in Chapter 18. The methodologies in Chapters 16 and 18
are applicable to an urban street facility that typically has a length of 1 mi or
more in downtown areas and 2 mi or more in other arcas.
The methodology described in this chapter is largely based on the product of
a 5trategic Highway Research Program 2 project (1). Contributions to the
methodology from other research are referenced in the relevant sections.
An important application of the methodology is in the evaluation of active
traffic and demand management (ATDM) tactics. An ATDM tactic adapts the
facility configuration and controls to (or in anticipation oí) variations in demand,
incidents, and weather to maintain a high level of facility performance (2). Thcsc
tactics are related to temporal changes in speed and signal control, geometric
configuration, and traffic demand volume. In its current form, the methodology
is most amenable to the evaluation of geometric ronfiguration modifications
(e.g., dynamic lane assignments, reversible lanes, and dynamic tum restrictions).

CHAPTER ORGANlZATION
Section 2 presen!s travel time variability and reliability concepts, ineluding
objectives of reliability analysis and definitions of reliability terms. This section
also ineludes an overview of ATDM and the range of strategies applicable to
urban street facilities.
Section 3 presents the core methodology for evaluating urban street
reliability. It ineludes a description of the scope of the methodology and its

Chapter 17¡Urban 5treet Relíability and ATDM lntroduction

Version 6.0 Page 17-1
Highway Capacity Manual: A Guide for Multimodal Mobilify Analysis

required input data. lt concludes with an overview of the reliability evaluation

methodology.
Section 4 describes various ATDM strategies applicable to urban streets and
typical tactics for implementing each strategy.
Section 5 presents guidance on using the results of the reliability evaluation.
It includes example results that illustrate an application of the reliability methods
to an urban street facility, and it discusses typical cases for which a reliability
evaluation can provide useful information.

RELATE O HCM CDNTENT

Other HCM content related to this chapter includes the following:
• Chapter 16, Urban Street Facilities, which describes concepts and
rnethodologies for the evaluation of an urban street facility;
• Chapter 18, Urban Street Segments, which describes concepts and
melhodologies for the evaluation of an urban street segment;
• Chapter 29, Urban Street Facilities: Supplemental, which provides details
of the reliability methodology, a proccdurc for sustained spillback
analysis, inforrnation about the use of alternative evaluation tools, and
example problems dernonstrating both the urban street facility
methodologies and the reliability methodology;
• Chapter 30, Urban Street Segments: Supplemental, which describes
procedures for predicting platoon flow, spillback, and delay due to turos
from the major street; a planning-level analysis application; and example
problems demonstrating the urban street segment methodologies; and
• Chapter 37, ATDM: SupplementaL which summarizes the steps involved
in designing an ATDM programo

Introduction O1apter 17fUrban 5treet Relíability and ATDM

Page 17-2 Vet5iM 6.0
 Highway Capacify Manual: A Guide for Mulfimodal Mobilify Analysis

2. CONCEPTS

This section summarizes kcy reliability concepts. The first subsection

discusses the rca50ns why an analyst might want to evaluate a fadlity's
reHability. The second provides important dcfinitions related to reliability
evaluation. The third describes performance measures and typical values. The
fourth summarizes key ATDM concepts.

OBJECTIVES FOR RELlABILITY ANALYSIS

An important step in any analysis is defining why the analysis is being
performed. Key qucstions or issues should be defined, performance measures
that help answer those questions identified, and a basis of comparison for
interpreting the analysis results estabJished. Reliability analysis is no different.
The following are examples of potential objectives of a reliability analysis:
• Tracking the reliability of a set of facilities in a jurisdictian or region ayer
time to prioritize them for operational or physical treatments,
• Diagnosing the primary causes of the reliability problems on a given
facility so that an improvement program can be developed, and
• Evaluating the effects of a particular treatment or improvement on facility
reliability.
More broadly, travel time reliability analysis can be used to improve the Reliab/llty analysis can be used
ro improve t!Ie operation,
operation, planning. prioritization, and programming of transportation system pliJnning, priontization, and
improvement projects in the following applications: long-range transportation programming of transportation
system improvement projects.
plans, transportatian improvement programs, corridor or areawide plans, major
investment studies, congestion management, operations planning. and demand
forecasting. The Use Cases portian of Section 5, Applications, describes these
applications in greater detail.

DEFINITIONS
The following terms are used in this chapter:
• Scenario. A unique combination of traffic demand, capacity, geometry,
and traffic control conditions. It can represent one or more analysis
pcriods provided that all periods have the same combination of demand,
capacity, geometry, and control.
• Study periodo The time interval (within a day) that is represented by the
performance evaluation. It consists of one or more consccutive analysis
periods.
• Analysis periodo The time interval evaluated by a single application of an
HCM methodology.
• Study sectioo. The length of facility over which reliability is to be
computed. Since reliability is computed for through traffic ooly, the
length of the facility should not be so long that through traffic is a low
percentage of total traffic on the facility. Thc length of facility to be

Chapter l7fUman 5treet Reliabihty and ATDM Concepts

VersiOn 6.0 Pagel7-3
Highway Capacify Manual: A Guide for Mu/fimoda/ Mobility Analysis

evaluated should be less than the distance a vehiele traveHng at the

average speed can achieve in 15 mino
• Reliability reporling periodo The specific days over which reliability is to
be computed, for example, aH weekdays in ayear.
• Special event. ShorHerm events that produce intense traffic demands on
a fadlity for limited periods of time. These demands may be addresscd by
temporary changes in the facility's geometry or traffic control
characteristics, or both. Example spedal events inelude major sporting
events, concerts, and festivals.
Additional terms are defined in Chapter 9, Glossary.

ACTIVE TRAFFIC ANO OEMANO MANAGEMENT

A TDM is dynamic real-time AmM is the dynamic management, control, and influence of travel demand,
management and control of
tlle arteria! system. traffic demand, and traffic flow of transportation facilities. Through the use of
available too1s and assets, traffic flow is managed and traveler behavior is
influenced in real time to achieve operational objcctives, such as preventing or
delaying breakdown conditions, improving safety, promoting sustainable travel
modes, reducing emissions, or maximizing system efficiency (3).

Speetrum of ATDM Applications

A TDM appfications on urtJan ATDM applications on an urban street can range widely from innovative
streets CQn range from the
simpfe ro the compfex. uses of conventional control hardware to the deployment of state-of-the-art real.
time traffic control systems. A conventional traffic-actuated signa! that measures
detector occupancy in real time and uses the information to adjust phase splits
automatically (within a fixed narrow allowed range) and skip unneeded phases
is an example of a relatively simple ATDM application. An example of a more
advanced ATDM application is traffic adaptive control, under which the system
controller dynamically measures demand and adjusts phase splits and offscts in
real time to serve pIatoons of vehicIes.

ATDM Options tor Urban Streets

When an agency adopts an overall ATDM strategy for managing its urban
streets, the strategy is typically composcd of various tactical actions taken from
one or more of the following tactical groups:
• Arterial monitoring tactics,
• Signal and speed control tactics,
• Gcometric configuration tactics, and
• Demand modification tactics.
Each tactical group has a specific objective and set of measures that are
implemented to achieve the overall agency ATDM goals. Typical ATDM
measures associated with each tactical group are shown in Exhibit 17-1. These
groups and measures are reviewed in the following paragraphs.

Co"",,,, Qlapter 17¡Urban Street Reliability and ATDM

Page 17-4 Vem?n6.0
 Highway Capadty Manual; A Guide for Multimoda! MobiJjty Anolysis

Exhibit 17-1
ATDM Tactics arld Measures
Agency ATM for Urban 5treets
Strategy

I
I 1 1 1
Arterial SignaljSpeed Geometric Demand
Monitoring Control Configuration Modification
Tactics Tactics Tactics Tactics

j r .
I I
.. Oetectors
• Adaptive Dynamic lane • Traveler Info •

. (ameras
Probes
. (ontrol
Speed . Assignments
Reversible
• Route
Guidance
. (ommercial
.. Harmonilation
Sp&ial Plans . lanes
Dynamic Turo
• Dynamic
Parking
SourcE's
EMS/Transit Restrictions • Congestion
Priority Pricing

Note: ATM'" active traffic management; EMS'" emergency medical services.

SOurce: Adapted from Dowling and Elias(2).

Arterial Monitoring radies

The objective of ATDM measures within the arterial monitoring tacticaJ
group is to obtain actionable real-time information on urban strect performance.
This objective may be aehieved by several means, such as the use of closed-
circuit television cameras and vehicle detectors, communication between
connected vehicles and the signal controller, or the purchase of travel speed data
from a commercial vendor of real-time traffie data. The choice of measures to
achieve the tactical objective of monitoring becomes the ageney's monitoring
tactic for the facility.

Speed and Signa! Control rames

The objective of ATDM measures within the speed and signal control tactical
group is to adapt signal timing (and speed limits if appropriate) to maximize
production (capacity) and minimize cost (deJay and stops). Measures to achievc
this objective inelude dynamic signal control (traffic actuated, traffie responsive,
and traffic adaptive) and dynamic speed Iimits that may be communicated via
roadside signs or overhead signs or that may be communicated direcHy to
connected vehicles. Signal timing affects the capacity of a street by changing the
allocation of green time between users of the street (through movements, transit,
pedestrians, and tum movements). It can .lIso affect the speed of travel on the
facility through signaJ coordination. NCHRP SYllfhesis 403 (4) and the Sigllal
Timillg Mallllal (5) provide more information on advanced signal control options.

Geometric Configuration Tadies

Thc objective of ATDM measures within thc gcometric configuration tactical
group is to adjust lane use on the urban strcct dynamically to improve the match
with demand, thereby increasing the street's capacity. These rneasures can
inelude changing the number of lanes dcsignated fur tums, changing the vehiele

Chapter 17/Urban Street Reliability and ATDM Concepts

Version 6.0 Page 17-5
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

types allawed to use a lane, or even changing the direction of flow foc certain
Janes. Practicalities limit the ability to open and clase parking lanes on a street
dynamically (drivers must be wamed when they first park their vehicle of when
they must leave). Dynamic tum restrictions are induded in this tactical approach.

Demand Management Tactics

The objective of AIDM measures within the demand modification tactical
group is to improve the match between demand and the available capacity.ln
the context of the HCM, demand management primarily relates to traveler
information services and guidance. The travel information is provided in the
hope that a better-informed traveling public will shift to less congested facilities
and thereby better balance demand with available capacity. A more proactive
form of demand management is to provide actual dynamic route and mode of
travel guidance, making travelers aware of additional routing and modal
options. Congestion pricing can be used to reinforce the traveler guidance for the
urban street.

Con<epts Olapter 17fUrban Street Reliability and ATDM

Page 17-6 VeI$iOn 6.0
 Highway Capacify Manual: A Guide lor Multimodal Mobility Analysis

3. CORE METHODOLOGY

At its core, the reliability methodology consists of hundreds of repetitions of

the urban street fadlity methodology presented in Chapter 16. In contrast to the
Chapter 16 methodology, where the inputs represent average values for a
defined analysis period, the reliability methodology varies the demand, capacity,
geometry, and traffic control inputs to the facility methodology with each
repetition (i.e., scenario).
AH the HCM performance measures output by the fadlity methodology are
assembled for each scenario and used to describe the fadlity's performance over
the course of ayear (or other user-defined reliability reporting period).
Performance can be dcscribed on the basis of a percentile result (e.g., the 80th or
95th percentile travel time) or the probability of achieving a particular level of
service (e.g., the facility operates at LOS O during X%of weekday hours during
the year). Many other variability and reliability performance measures can be
developed from the facility's travel time distribution.
The reliability methodology is sensitive to the main sources of variability that
lead to travel time unreliability. These sources are as follows:
• Temporal variability in traffic demand-both regular variations by hour
of the day, day of the week, and month or season of the ycar and random
variations between hours and days;
• lncidents that block travellanes or that otherwise affect traffic opcrations
and thus capacity;
• Wcather events that affect capacity and possibly demand;
• Work zones that c10seor restrict travellanes, thus affecting eapacity; and
• Special events producing atypical traffic demands that may require
management by special traffic control measures.
Work zones and special events are loeation-specific parameters that must be Input rota needed for a
reliabiflty evaluation (beyond
provided by thc analyst. Location-specific data related to traffie demand those needed for an HCM
variabi!ity, incidents, and weather patteros are best provided by the analyst if faality evafuation) consist of
demand van'ation data,
they are available; however, the reliability methodology also provides default inddent rota, weather data,
values for use when local data are unavailable or the analysis does not require work zones, and speo'al
events. Tñe first t/Jree types of
that level of preosion. data Ciln be defaulted when
they are not avaifable Ioca/Iy.
Scenarios are built from combinations of conditions associated with each
source of trave! time variability. For example, one seenario could represent
dcmand volumes representative of Fridays in May, fair weather, and one lane
c10sed for 30 min because of an incident that occurs during the p.m. peak houe. A The method for estimating the
trave! time distribution
probability of oecureence is assooated with each scenario on the basis of local CiJfa¡lates the performance of
data provided by the analyst (or the method's default data) and is used to a series of scenarios
representing different
develop a travel time distribution for the reliability reporting periodo combinations of conditions that
affect a faci/ity's cJemand or
Exhibit 17-2 provides a high-lcvel representation of the mcthodology for
CiJpacity, or both.
estimating the travel time distribution. The base dataset consists of all the data
needed to evaluate the base HCM facility methodology for a single study period,
plus data that describe the variations in demand, weather, and so forth that occur

Chapter 17tUrban 5treet Reliability and ATDM Core Methodology

Verskln 6.0 Page 17-7
Highway Capadty Manual: A Guide for Multimodol MobUity Analysis

over the course of the reliability reporting pt.'riod, along with the frequency of a
particular event's occurrence. The scenario generator creates a set of scenarios in
wruch the base facility demand and capacity are adjusted to reflect the changes in
demand and capacity that occur under each combination of conditions. Each
scenario is submitted to the HCM fadlity methodology, which calculates the
facility travel time associated with the scenario. The individual facility travel
times are then compiled into the facility's travel time distribution. This
distribution can be used to develop a variety of reliability and variability
performance measures for the facility.

Exhibit 17-2 Base Dataset

High-level Representation of
the Method for Estlmating the Demand ••Do
Travel TIme Distribution Capacity "Co
free-Flow Speed" So
fraffk Control. To

Scenarios
Scenario Generator
Oemand Pattern" d - Core HCM FaciJitv Method

Incidenl ••/ - Demand: D.

~ Chapter 16
(Urban Street facilities)
Weather" W ~
Speclal [venl" s Capacity " c..
free-flow Speed: S.
fraffic Control: T. 1
Travel Time Oistributlon
Performance Measures
Planning TIme Inde~
80th Percentile TravelTIme Inde~
Rellabilíly Rating
~ --,--
1-
j'"
',.
1':•.. ~ ,..
,,--~., .,
On-TIme Performance
Semi-Standard Oevlation
elc. " ''''

Because of the hundreds (or even thousands) of scenarios that are generated,
implementation of this method is only practical through software. Software
automates the scenario generation process, performs the computations associated
with the HCM facility methodology for each scenario, and stores and processes
the output performance measures generated for each scenario. Source code
listings for the research-grade computational engine (Le.,STREETVAL) are
providcd in the Tcchnical Reference Library in HCM Volume 4.

SCOPE OF THE METHOOOLOGY

The reliability methodology can be used to evaluate the following sources of
unreliable travel time:
• Demand fluctuations,
• Weathcr,
• Traffic incidcnts,
• Work zones,
• Special events,

<:ore Methodology O1apter 17jUrtlan Street Reliability and ATDM

Page 17-8 Ver5ion 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

• lnadequate base capadty, and

• Traffic control devices on urban streets.
Demand fiuctuations are represented in the methodology in terms of
systematic and random demand variation by hour of day, day of week, and
month of year. Fluctuations due to diversion are not addressed directly by the
methodology but can be optionally provided by the analyst for work zones and
spedal events through the demand specified in an aItemative dataset.

Performance Measures
The reliability methodology generates distributions of the performance
measures produced by the HCM facility methodologies. Each distribution
represents the variation of one performance measure during the reliability
reporting periodo Performance measures applicable to urban street facilities
inelude travel time, travel specd, delay, and spatial stop rate, among others.
The distribution of a performance measure can be aggregated over the
reliability reporting period to produce an overal! total (or average). Measures of
this nature are described in the first subsection to follow. The distribution can
also be described by using percentile values to indicate measure variability or
propensity for failure. Measures of this nature are described in the second
subsection to follow.

Measures Describing Typicaf (Average) Conditions

Several traditional performance measures are used to describe the overall
average operation of a facility.ln combination with reliability measures, they
providc a complete picture of fadlity performance and form a useful basis for
aItemative evaluation. The following are useful traditional measures:
• Trave/ time (minutes). Travel time is a versatile measure, since it can be
monitored over time (for trend analysis), monetized (in calculating
benefits), and used in the calculation of other measures (e.g., delay).
Facility lengths usually remain the same over time, allowing apples-to-
applcs comparisons of travel times estimated for a facility in different
ycars or under different drcumstances.
• AmlUul de/ay (\'Chiclehours and person hours). Annual delay is the
average vehicle hours of travel or persan hours of travel occurring minus
what would occur under free-flow conditions. Delay is useful because
economic analyses have a long history of monetizing delay.

Measures Describing Reliability

~F~now speed~ ís the
Numerous performance measures are available for quantifying aspects of
average runníng speed of
facility travel time reliability. Many of them combine a specified percentile value through vehicfeS traveling
along a segment under .bw-
with a typical or ideal value to compute a dimensionless index that describes the
voIume ronditions and not
relative variability, relative propensity for failure, or both. For urban street delayed by trafflC control
facilities, the base free-flow speed is used as the ideal value on which aH speed- devices or other vehicfeS.
The "base free-fIow speed~ is
and travel-time-related reliability indices are based. defined ro be the free-ffow
A commonly used index is the travel time index (Tri). It is defined as the speed on Ionger segments.

ratio of the actual travel time on a facility to the travel time at the base free-flow

Chapter 17fUrban Street Reliability and ATDM Core Methodology

Vl'fSim 6.0 Page 17-9
Highway Capacity Manual: A Guide lar Multimadal Mability Analysis

speed. Other indices are defined in Section 2 of Chapter 4, Traffie Operations and
Capacity Concepts.
The following measures are useful for describing (a) travel time variability or
(b) the suecess or failure of individual trips in meeting a target travel time or
sp€M'
• 95th percentile ITI or planlling time index (PT!) (unitless). The ratio of the
95th percentile highest travel time to the travel time at the base free.flow
speed. This measure is useful for estimating how much extra time
travelers must budget to ensure an on-time arrival and for describing
near-worst-case eonditions on urban facilities.
• 80fh percentile ITI (unitless). The ratio of the 80th percentile highest travel
time to the travel time at the base free-flow speed. Research indica tes that
this measure is more sensitive to operational changes than the PTI (6),
which makes it useful for comparison and prioritization purposes.
• 50th percentile ITl (unitless). The ratio oE the 50th percentile highest travel
time to the travel time at the base free-flow speed. This measurc can be
used for trend analysis and to demonstrate changes in performance
resulting from an operational strategy, eapadty improvement, or change
in demando
• Mean ITl (unitless). The ratio of the average travel time to the travel time
at the base free.flow speed. This measure can be used Earthe same
purposes as the 50th percentile 111. However, the mean ITI will typieally
have somewhat higher values than the 50th percentile 111 because oC the
influence oE rare, very long travel times in the distribution.
• Failure or oll-time measures (percentage). The percent oC trips (0£ percent oC
time) with space mean speeds above (on time) or bclow (faiJure) one or
more target values (e.g., 35, 45, and 50 mi/h). These measures address
how often trips succeed or Eailin achieving a desired travel time or speed.
• Reliability rating (percentage). The percentage oE vehic1e miles traveled on
the facility associated with a 111less than 2.50. This threshold
approximates the point beyond which urban street fadlity travel times
beeorne much more variable (Le.,unreliable).
• Semi-sta"dard deviatioll (unitless). A one-sided standard deviation, with
the reference point at the base free-flow speed instead of the mean. It
provides the variability distance from free-flow eonditions.
• Standard deviatioll (unitless). The standard statistical measure.
• Misery index (unitless). This measure is useful as a descriptor of near-
wO£st.case condítions on rural facilities.
In many cases, an analyst may wish to consider several of these measures to
obtain a complete picture oE travel time reliability. However, the reliability rating
is recornmended as part oE any HCM.based reliability analysis because it is a single
measure reflecting the traveler point of view (by stating the potential for unreliable
travel). The use of the rcliability rating and other reliability rneasures is iIIustrated
in example problems in Chapter 29, Urban Street Facilities: Supplemental.

(ore Methodology Qlapter 17jUrban 5treet Reliability and ATDM

Page 17-10 Version 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Limitations of the Methodology

Bccause the reliability methodology is based on applying the urban street
methodologies multiple times, it inherits the limitations of those methodologies,
as described in Chapters 16, 18, and 19, respectively. The reliability methodology
has additionallimitations as describcd in the following paragraphs.
In general, the urban street reliability methodology can be used to evaluate
the performance of most urban street facilities. However, the methodology does
not address the following events or conditions:
• Truck pickup and delivery (double parking);
• Signal malfunction;
• Railroad crassing;
• Railroad and emergency vehide preemption;
• Signal plan transition; and
• Fog, dust storms, smoke, high winds, or sun glare.
Lane or shoulder blockage due to truck pickup-and-delivery activities in
dowotown urbao areas can be considered incident-like in terms of the
randomness of their ocrurrence and the temporal exteot of the evento The dwell
time for these activities can range fram 10 to 20 min (7).
A signal malfunction occurs when one or more elements of the signal system
are oot operating io the intended manner. These elements indude vehide
detectors, signal heads, and controller hardware. A failure of one or more of
these elements typically results in poor facility operation.
A railroad crossiog the fadlity at a midsegment location effectively blocks
traffic flow while the train is present. Traio crossing time can be lengthy (Le.,
typically 5 to 10 mio) and can result in considerable cangestion extendiog for one
or more subsequeot analysis periods.
Railroad preemptian acrurs when a train crosses a crass-street leg of a
signalized intersectian. The signal operation is disrupted to dear the tracks
safely. Signal coordination may be disrupted for several cydes after traio
dearance.
When a new timing plan is invoked, the contraller goes through a transition
fram the previous plan to the new plan. The transition periad can last several
cydes, during which traffic progression is significantly disrupted.
Sorne weather canditions that restrict driver visibility or degrade vehide
stability are not addressed by the methodology. Thcy indude fog, dust storms,
smoke, and high winds.

O1apter 17/Urban 5treet Relíability and ATDM COfe Methodology

Vt'iS'On 6. O Page 17-11
Highway Capacity Manual: A Guide fer Multjmedal Mebility Analysis

REQUIRED DATA ANO SOURCES

HCM Facility Analysis Input Data

The input data needed to evaluate an urban steeet facility for one analysis
period are also needed for a reliability evaluation. These input data are dcscribed
in Chapters 16,18, and 19 and are referred to as an HCM dataset in this chapter.
For sorne reliability evaluations, more than one HCM dataset will be
required. One HCM dataset, the base dataset, is always required and is used to
describe base conditions (particularly demand and factors influencing capacity
and free.flow speed) when work zones and special events are not present. The
base dataset can represent average dernand conditions or the dernand measured
on a spccific day. The reliability methodology indudes a procedure for factoring
the average-day or specific-day demands (on the basis of user-supplied or
defaulted demand patterns) to generate demands representative of al1 other time
periods during the reliability reporting periodo
Additional HCM datasets are used, as needed, to describe conditions when a
specific work zone is present or when a special event oecurs. They are ealled
aIternative datasets. The usee must specify any changes in base conditions (e.g.,
demand, traffic control, available lanes) associated with the work zone or speciaI
evcnt, along with the times when the altemative dataset is in effect. For example,
if a work zone exists during a given rnonth, an altemative dataset is used to
describe average conditions for the analysis period during that month.

Additional Data Required for Reliability Evaluation

Additional data (beyond those described in the previous subsection) are
required for a reliability evaluation. Exhibit 17-3 gives the general categories of
data that are required.

Exhibit 17.3 Data category Description

General Data cate90ries Functionalelass Functionaldass required when defaulted demand pattems are
Requiredfor a Reliability
Evaluation Nearest dty """.
Required when defaulted weather data are used.
Geometric; Presence of shoulder.
Time periods Analysisperiod, study period, rellabilityreporting periodo
Demalldpattems Hour-of-day(JQ, day-of-week,and month-of-yeardemand
factors relativete annual average daílytraffie. Demand ehange
due to rain and snOW.can be defaulted.
Weather Rain, snow, alld temperature data by rnonth. Pavement runoff
duration fer a soow eventocan be defaulted.
Incidents Probabílitiesef specifiecrash and lncidenttypes by location.
A1ternativety,segment and intersection crash frequencies. Crash
frequency adjustment factors. Factors influendng incident
duration. The latter twe factors can be defaulted.
WorklOnes and spedal events Olanges te base conditions(alternative dataset) and schedule.
Traffiecounts Day and time of trafflecounts used in base and altemative
datasets.

As shown in Exhibit 17~3, most reliability-specific inputs can be defaultcd.

Dcfault values are identified in the following subsections. They allow analysts in
"data poor" regions laeking detailed demand, weathee, or incident data to apply
the reliability methodology and obtain reasonable resuIts. At the same time, the

Core MethodoIogy Olapter 17/Urnan5treet Reliabilityand ATDM

Page 17-12 Vet5k¥7 6.0
 Highway Capacity Manual; A Guide for Multimodal Mobility Analysis

methodology allows analysts in "data rich" regions to provide local data for
these inputs when the most accurate results are desired.

Fundiona/ C1ass
The functional elass of the subject facility is a required input when the
analyst chooses to use the default time period adjustment factors. These factors
are used for estimating traffic volume during each of the various scenarios that
make up the reliability reporting periodo The default factors are described in the
subsequent scction titled Demand Pattem Data.
The following functional elasses are considered:
• Urban expressway,
• Urban principal arterial street, and
• Urban minar arterial street.
An urban principal arterial strect cmphasizes mobility over access. lt serves
intra-area traveL such as that bctween a central business district and outlying
residential arcas or that between a freeway and an important activity center. It is
typically used far relatively long trips within the urban area ar for through trips
that enter, Icave, or pass through the city. An urban minor arterial street
provides a balance between mobility and access. It interconnects with and
augments the urbOlnprincipal arterial street system. It is typically used for trips
of moderate length within relatively small geographic areas (8).
The methodology addresscs roadways that (a) belong to one of the
aforementioned elasscs and (b) do not have fuIl access control. If a roadway has
full access control; it is considered to be a freeway, and the analyst should use the
freeway methodology.

Nearest City
The nearest city is a rcquired input when the analyst chooses to use the
dcfault weather data. Thc analyst sclects from 284 US. cities. The default
weather data are described in a subsequent subsection titled Weather Data.

Geometrics
The indication of the presence of outside (Le., right-side) shouldcrs is a
required input when the analyst chooses to use the default incident location data.
This input is specified for the facility. The dcfault incidcnt location data are
described in a subsequent subsection litled lncident Data.
For a shoulder to be considered present, it must be wide enough to store a
disabled vehiele (so that the vehiele does not block traffie flow in the adjaccnt
traffie lane). If on-street parking is allowed, the analyst will nced to determine
whether oecupancy of the shoulder during the study period is sufficient to
preelude its use as a refuge for disabled vehieles. The proportion of on-street
parking occupied would nt..'Cdto be less than 30% to provide reasonable
assurance of the opportunity to mov(' a disabled vehiele {rom the through lanes
to an open stall.

Olapter 17/Urnan $treet Reliability and ATDM Core Methodology

Version 6.0 Page 17-13
Highway Capacity Manual: A Guide for Multjmodol Mobility Anolysjs

Time Pen"ods
The time periods that need to be specified inelude the analysis perlod, the
study period, and the reliability reporting periodo Exhibit 17-4 presents the
relationships between these periods. They are defined in the following
paragraphs.

Exhibit 17-4
Temporaland Spatial Eadl cell is one
allillysis period of
Dimensionsof Reliability an allillysis segmento

5tud {20:00
V
"'"00
14:00
"7-------
1
5tlJeIy 5ection

Source: Zegeer et al. (1).

Analysis Period
The analysis period is the time interval used for the performance evaluation.
It can range from 15 min to 1 h, with longer durations in this range sometimes
used for planning analyses. A shorter duration in this range is typically used for
operational analyses. Additional guidance for determining the analysis penod
duration is provided in Chapter 16, Urban Street Facilities.
Shofter anafysis periods allow A shorter analysis penod duration is desirable for reliability evaluations
relativefy brief inddents and
weather events ro be because H reduces the minimum event duration threshold and thereby increases
amsidered in refiability the number of incidents and weather events that are included in scenarios. In this
evaluations.
regard, the strueture of the reliability methodology is such that events that are
shorter than one-half of the analysis period duration are ignored (Le., they will
not be recognized in the scenario generation process).

" 5tudy Period

If an uroan stTeet fadlity has The study period is the time interval (within a day) that is represented by the
lWO or more bfne.of-day Signal
timing plans, a separate study performance evaluation. lt eonsists of one or more consecutive analysis periods.
period should be establiShed A typical study period is 1.0 to 6.0 h in duration and is stated to represcnt
for each pfan periodo
specific times of the day and days of the week (e.g., weekdays froro 4:00 to 6:00
p.m.). If oversaturated eonditions occur during the study period, at least the first
analysis period should be undersaturated. The maximum study period duration
is 24 h.
The geometric design elements and traffie control features of the facility
must be unchanged during the study periodo Thus, the intersection lane
assignments and signal timing plan should be the same throughout the study

COreMethodology Chapter17/Urnan5treet Reliability

anclATDM
Page 17-14 VetSbn 6.0
 Highway Capacity Manual: A Guide lor Multimodal Mobility Analysis

periodo In addition, jf the directional distribution of traffic volume ehangcs

signifieantly during the day, separate study periods should be established for
eaeh time period where the directional distribution is relatively eonstant.

Re/iability Reporting Period

The reliability reporting period represents the spedfic days over which the
travel time distribution is to be computed. A typieal reporting period for a
reliability evaluation is 6 to 12 months. The period is spedfied by start and end
dates as well as by the days of week being considered. The reliability reporting
period is used with the study period to describe the temporal representation of
the performance measure fully (e.g., average travel time on weekdays from 4:00
to 6:00 p.m. for the current year).

Demand Pattem Data

Demand pattern data are used by the reliability method to adjust the
demand volumes in the base and alternative datasets to reflect demands during
all the other time periods in the reliability reporting periodo The data inelude (a)
adjustment factors to account for demand variation by hour of day, day of week,
and month of year and (b) adjustment factors to account foc ehange in traffie
demand due to weather conditions.

nme Period Adjustment FactDrs

The methodology requires day-of-week and month-of-year faetors,
expressed as ratios of the average day-of-week and average month-of-year
demando Also required are hour-of-day faetors expressed as a pereentage of
annual average daily traffic (AADT). The specifie faetors needed are described as
follows:
• Hour-of-day factors for each hour of the study period (up to 24, but
typically six or fewer in practice),
• Day-of-week factors for eaeh day inc1uded as part of the reliability
reporting period (up to seven), and
• Month-of-year factors for each month included as part of the reliability
reporting period (up to 12).
Default hour-of-day, day-of-week, and month-of-year traffic demand
adjustment factors are given in Exhibit 17-5 through Exhibit 17-7, respeetively.
The factors should be replaeed with data £rom permanent traffie count stations
whenever available for streets that are similar to the subject fadlity and located
ncar it.

Chapter 17/Urban 5treet ReliabiHty and ATDM Core Methodology

VersiM 6.0 Page 17-15
Highway Capadty Manual: A Guide for Multimodal Mobility Analysis

:;'~H~OU~'~=~~E~X~P~"'~<W~'~Y;;=:;~P~'~;"~C~ipa~;
A~rt~"~;'~'~=~~M~;~"~O'tA~rt~"~ia~I~~
Exhibit 17-5
Default Hour-of-DayDemaOO Starting Weekday Weekend Weekday Weekend Weekday Weekend
Ratios (ADT/AADT) Midnight 0.010 0.023 0.010 0.023 0.010 0.Q28
1 a.m. 0.006 0.015 0.006 0.014 0.006 0.023
2 a.m. 0.004 0.008 0.005 0.010 0.004 0.021
3 a.m. 0.004 0.005 0.005 0.006 0.002 0.008
4 a.m. 0.007 0.005 0.009 0.006 0.002 0.005
S a.m. 0.025 0.009 0.030 0.010 0.007 0.005
6 a.m. 0.058 0.016 0.054 0.017 0.023 0.011
7 a.m. 0.077 0.023 0.071 0.024 0.067 0.018
8 a.m. 0.053 0.036 0.058 0.035 0.066 0.030
9 a.m. 0.037 0.045 0.047 0.046 0.054 0.048
10 a.m. 0.037 0.057 0.046 0.056 0.051 0.054
11 a.m. 0.042 0.066 0.050 0.054 0.056 0.057
Noon 0.045 0.076 0.053 0.071 0.071 0.074
1 p.m. 0.045 0.073 0.054 0.071 0.066 0.071
2 p.m. 0.057 0.074 0.063 0.072 0.060 0.069
3 p.m. 0.073 0.075 0.069 0.073 0.062 0.067
4 p.m. 0.087 0.075 0.072 0.073 0.063 0.071
5 p.m. 0.090 0.071 0.077 0.073 0.075 0.068
6 p.m. 0.068 0.063 0.062 0.063 0.Q70 0.067
7 p.m. 0.049 0.051 0.044 0.052 0.053 0.056
8 p.m. 0.040 0.043 0.035 0.044 0.044 0.049
9 p.m. 0.037 0.037 0.033 0.038 0.035 0.040
10 p.m. 0.029 0.032 0.026 0.033 0.033 0.035
11 p.m. 0.019 0.023 0.021 0.026 0.019 0.024

Source: Hallenbed el al. (9).

Exhlbit 17-6
Default Day-of-WeekDemaOO
:===~D~.~y¡:::======JDe¡;;m~.~"~d~
SuOOay
••~t~;O¡:::== 0.87
Ratios (ADT/AADT) Monday 0.98
Tuesday 0.98
Wednesday 1.00
Thursday 1.03
Friday U5
saturday 0.99
SOurce: Hallenbecl<. el al. (9).

Exhibit 17-7
Default Month-of.Year
DemaOORatios (ADT/AADT)
:=~M~O~"~th~====]E~x~pre~~~W¡!!'YL==Jp5n~."~C~iP~.~I~A~rt~.~,;~.[1
January
February
==!M~;~"O~'~Art~.~n~
0.802
0.874
.• 0.831
1.021
0.881
0.944
~I=
Mardl 0.936 1.030 1.016
April 0.958 0.987 0.844
May 1.026 1.012 1.025
June 1.068 1.050 1.060
July U07 0.991 1.150
Augusl: 1.142 1.054 1.110
5eptember 1.OB8 1.091 1.081
October 1.069 0.952 1.036
November 0.962 0.992 0.989
December 0.933 0.938 0.903
Source: Hallenbed el al. (9).

Demand Change Factors

The three "demand change factors" account for a change in traffic demand
due to weather conditions. One factor describes demand change during dry
weather (by definition it has a value of 1.0). A second factor de~ribes demand
change during a rain evento The third factor describes demand changc for a snow
evento During a step of the methodology, the demand volume is multiplied by
thc dcmand change factor corresponding to the weather associated with a given

Core MethodoIogy Chapter 17/Urban Street ReliabilityaOOATDM

Page 17-16 Verooo6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

analysis periodo A factor less than 1.0 corresponds to a reduction in demand

during the evento
Research indicates that urban street traffie dcmand tends to drop 15°;',to 30%
during snow events (10). Thcsc motorists likely altered the start time of their
eommute or stayed home to avoid the bad weather. In the absence of local data, a
default value of 0.80 may be used for snnw events.
The research is less clear on the effed of rain on traffie demando The effeet of
rain may vary with the trip purpose <lnd the annual freguency of rain events in
the vicinity of the subject facility. A dcfault factor value of 1.0 is reeommended
for rain eveots. These default values are summarized io Exhibit 17-8. The input
data item in the last row oi this exhibit is discussed in the next subsection.

Input Data Item Default Value Exhibit 17-8

Demandchangefactorfordryweather 1.00 DefaultValuesforWeather
DemandchangefactorfO( rainevent 1.00 Events
Demandchangefactorforsnowevent 0.80
Pavementrunoffdurationforsnowevent 0.5 h

Weather Data

Weather Event Statistics

A reliability evaluatioo rcquires the weather data identified in the iollowing
listo These data rcpresent averages by month of ycar for a reeent lO'year periodo
• Total normal prccipitation (in.),
• Total normal snowfall (in.),
• Number of days with precipitatioo of 0.01 io. or more (days),
• Normal daily mean tcmperature CF), and
• Prccipitation rate (in./h).
Default values for the aforementioned statistics are available from the
National Climatic Data Center (NCDC) for 28410cations in the United States (11,
12). These default valucs are numerous, so they are not shown here. They are
listed in thc STREETVAL computational engioe available in the online fKM
Volume4.

Duration of Pavement Runoff

The duration of pavement runoff for a snow event is required. It is defined as
thc period of time after the snow stops falling that snowpack (or ice) covers the
pavcment. After this time period clapses, the pavement is exposed and dryiog
begios. This time is likdy a function of traffic volume, snow depth, and ageocy
snow rcmoval capabilities. An appropriate local value should be eslablished for
the subject facility if that is possible. If such a value is not available, the default
value provided in the last row oi Exhibit 17-8 can be used.

Chapter17jUrnanStreetReliability
andATDM CoreMethodology
Version 6.0 Page 17.17
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Incident Data
Crash LOCiJtioncategories
Chapter 16, Urban Street Facilities, defines segments as induding portions of
their bounding intersections (Le., segments extend from the upstream
intersection stop bar to the downstream intersection stop bar). For the porposes
of reliability analysis, this definition must be modified to eategorize eaeh crash in
aeeordanee with the location of its occurrenee (Le., on the segment or at the
intersection). The categorization of crashes by location is determined by using
the definitions given in Highway Safety Manual (HSM) Seetion A.2.3, found in
Appendix A of HSM Volume 2 (13). The HSM states: "lntcrsection crashes
indude erashes that oecor at an intersection (Le., within the curb limits) and
erashes that oecur on the interseetion legs and are intersection-related. AlI
crashes that are not classified as intersection or intersection-related crashes are
considered to be roadway segment crashes."

Base Segment and lntersection Crash Frequency

The methodology requires the base crash frequeney for each segment and for
each intersection along the subjeet fadlity. The base erash frequency is an
estimate of the expeeted crash frequeney for the segment or intersection when no
work zones are present or spedal events oecur. The estimate should indude all
severity levels, induding property.damage-only (POO) crashes. Crash frequeney
is provided in units of crashes per year, regardless of the duration of the
reliability reporting periodo

Crash Frequency Adjustment Faetors far Work Zones and Special Events
Crash frequeney adjustment factors must be supplied for each work zone or
spedal event for whieh an altemative dataset is assembled. One erash frequeney
adjustment factor is supplied for eaeh segment and one is supplied for eaeh
intcrseetion. They are used (at the appropriate step of the reliability
methodology) to estimate the expected crash frequency when a work zone or
spedal event is presento The estimate is obtained when the appropriate factor is
multiplied by the base crash frequeney for the segment or interseetion. The result
represents the expected crash frequency in a segment or at an intersection if the
work zone or spedal event were present for 1 year.
The factor value should indude consideration of the effect of the work zone
or spedal event on traffie volume and crash risk. Volume may be redueed
beeause of diversion, while changes in the roadway geometry and signal
operation for a work zone or spedal event may inerease the potential for a crash.
To illustrate this concept, eonsider a work zone that is envisioned to inerease
crash risk by 100% (Le., erash risk is doubled) and to deerease traffic volume by
50% (Le., volume is halved). In this situation, the erash frequency adjustment
factor is 1.0 (= 2.0 " 0.5). The analyst's experienee with similar types of work
zones or special events should be used to determine the appropriate adjustment
factor value for the subjeet facility.

Core Methodology Chapter 17/Urban $treet Reliability and ATDM

Page 17-18 Vef5ion 6.0
 Highway Capacity Manual: A Guide tor Multimodal Mobility Analysis

Crash Frequency Adjustment Faetors for Inclement Weather

Inclement weather conditions can increase the likelihood of crashes. Crash
frequeney adjustment factors are required for the following conditions:
• Rainfall,
• Snowfall,
• Wet pavement (not raining), and
• Snow or ice on pavemcnt (not snowing).
The crash frequeney adjustment factor is the ratio of hourly crash frequency
during the weather event to the hourly crash rate during c1ear,dry hours. lt is
computed by using one or more years of historical weather data and crash data
for the region in which the subject fadlity is located.
The adjustment factor for a specific weather condition is computed from (a)
the number of hours for which the weather condition exists for the year and (b)
the count of crashes during those hours. An hourly crash frequency for the weathcr
conditionJ..~_ is computed by dividing the crash count by the number of hours.
By a similar technique, the hourly crash frequeney is computed for dry pavement
hours J..dr¡¡' The crash frequency adjustment factor for the weather condition
CFAF""" is computed as the ratio of the two frequencies (i.e., CFAF= = J..~_/f,"dn)'

The crash frequeney adjustment factor ineludes consideration of the effect of

the weather event on traffic volume (Le., volume may be reduced because of bad
weather) and on crash risk (Le., wet pavement may increase the potential for a
crash). For example, if rainfall is anticipated to increase crash risk by 200% and to
decrease traffic volume by 10%, the crash frequency adjustment factor for rainfall
is 2.70 (= 3.0 " 0.9).
Exhibit 17-9 provides default values for the crash frequency adjustment
factor for inclement weather. The other input data e1ements listed in the exhibit
are discussed in the next subsection.

Input Data Element Default Values Exhibit 17-9

Crash frequeocy adjustment Rainfall: 2.0 Default Values for Incidents
factor fOf weather conditions Wet pavement (not raining): 3.0
Snowfall: 1.5
Snowor iCe on pavement (not snowing): 2.75
Incident detection time 2.0 min (all weather COnditions)
lncident response time Oear, dry: 15.0 min
Rainfall: 15.0 min
Wet pavement (not raining): 15.0 min
Snowfall: 20.4 min
Snow or ice on pavement (not snowing): 20.4 min

SOurce: Zegeeretal. (1).

Faetors Influencing Incident Duration

The duration of an incident depends on a number oí factors, including time
to detect an incident, time to respond, and time to clear the incident. The incident
detection time is the time period starting with the occurrence of the incident and
ending when the response officials are notified of the incidenL lncident response
time is the time period from the receipt oí incident notification by officials to the
time the first response vehicle arrives at the scene of the incidenL This time will

Chapter 17¡Urban Street ReliabiHty aOO ATDM Core Methodology

Versk¥7 6.0 Page 17-19
Highway Capacity Manual: A Guide tor Multimodal Mobility Analysis

Iikcly vary among jurisdictions and facilities, depending 00 the priority placed
on street systcm management aod the conncctivity of the street system. Incident
clearance time is the time from the arrival of the first response vehicle to the time
when the inddent and service vehicles no longer directly affect travel on the
roadway. This time varies by inddent location, type, and severity.
Response and clearance times are weather-dependent; clearance times are
also dependent on the incident severity and Iocation (e.g., shoulder versus travel
lanes). The following vaIues are required:
• Incident detection time, in minutes;
• Incident response times, in minutes, for five weather categories (dry,
rainfall, snowfall, wet pavement, snow or ice on pavement); and
• Incident clearance times, in minutes, by street location (segment or
intersection), incident type (crash or noncrash), lane location (shoulder,
one lane, two or more lanes), severity (fatal/injury or rOO), and weather
condition (dry, rainfall, wet pavement, snowfall or snow or ice on
pavement) (96 total values).
Default valucs for incident detection time and incident response time are
provided in Exhibit 17-9. Default values for incident clearance time are provided
in Exhibit 17.10. The default distributions for segments and intersections are thc
same in this exhibit. Segments and intersections are differentiated beeause the
method allows thc analyst to provide different c1earanee times for segments and
intersections when local values are available.

Exhibit 17-10
Default Incident Oearance
Times
Street
~:.t
La ••
C¡earanee

D~
Rain- w.,
Time by Weather Condition (mio)
Snow
or Iceb
location Location 5everitv6 faU Pavement
5egment <Ja'" One tane FI 56.4 42.1 43.5 76.7
'00 39.5 28.6 29.7 53.7
2+ lanes FI 56.4 42.1 43.5 76.7
'00 39.5 28.6 29.7 53.7
Shoulder FI 56.4 42.1 43.5 76.7
'00 39.5 28.6 29.7 53.7
Nan- One lane Breakdown 10.8 5.6 5.7 14.7
crash Olh" 6.7 2.4 2.8 9.1
2+ lanes Breakdown 10.8 5.' 5.7 14.7
Olh" '.7 2.4 2.8 9.1
Shoulder Breakdown 10.8 5.' 5.7 14.7
Olh" '.7 2.4 2.8 9.1
Signalized
intersection
C"", One lane FI
'00
56.4
39.5
42.1
28.6
43.5
29.7
76.7
53.7
2+ lanes FI 56.4 42.1 43.5 76.7
'00 39.5 28.6 29.7 53.7
Shoulder FI 56.4 42.1 43.5 76.7
'00 39.5 28.6 29.7 53.7
N"". One lane 8reakdown 10.8 5.' 5.7 14.7
,,,'" 2+ lanes
Olh"
8reakdown
6.7
10.8
2.4
5.'
2.8
5.7
9.1
14.7
Olh" 6.7 2.4 2.8 9.1
Shoulder 8reakdown 10.8 5.' 5.7 14.7
Olh" 6.7 2.4 2.8 9.1

Source: Zegeeretal. (1).

Notes: 6 A ~ fatal or inJury crash; POO '" property-damage-onlycrash.
~ Applies te snowfall and te sroow or Ice on pa~ement (but not snowing).

Cof"eMethodology Chapter 17¡Urban Street Reliability and ATOM

Page 17-20 Version 6.0
 Highway Capacity Manual: A Gujde for Multimodal Mobjlity Analysjs

In general, an analyst shouId supply local values for the incident duration
factors when the reliability analysis is testing the effects of traffie management
measures that influenee incident deteetion, response, or dearance.

Incident Location Distribution

The incident location distribution is used by the incident generation
procedure to assign incidents to specific ¡ocations on the facility. Research
indicates that this distribution varies by incident location, type, and severity. The
following incident proportions are required:
• Proportion of crash and noncrash inddents by street location (segment or
intersection) (four total values); proportions should total 1.000 for a given
street location;
• Proportion of shoulder, one-Iane, and two-or-more-Iane incidents by
street location and event type (crash or noncrash) (12 total values);
proportions should total 1.000for a given street !ocation and event type
combination; a 0.000 proportion should be assigned to values involving a
shoulder location if no shouIders exist on the facility;
• Proportion of fatal or injury and rDO erashes by street loeation and lane
location (12 total values); proportions shouId total 1.000for a given strcct
Iocation and lane !ocation combination; and
• Proportion of breakdown and other noncrash incidents by street location
and lane !ocation (12 total values); proportions should total 1.000 for a
given street location and lane loeation combination.
The four proportions identified in the previous list are multiplied together to
obtain the desired incident location distribution factors. One factor is obtained
for each combination of street location, incident type, incident location, and
incident severity. The computcd factors should total 1.000 for a given street
location.
Default values for these factors are provided in the last column of Exhibit 17-
11 and Exhibit 17-12.The default distribution of incident lane location is based
on facilities with outside shoulders. The distribution is modified accordingly
when shoulders are not present on the subject facility. The first exhibit provides
the distribution for urban streets with shoulders. The second exhibit provides the
distribution for urban streets without shoulders.

Qlapter 17jUrban 5treet Relíabílity and ATDM (ore Methodology

Ve~1On6.0 Page 17-21
Highway Capacfty Manual: A Guide for Multimodal MobUity Analysis

Exhibit 17.11 Incident Tyoe Inddent Location Incident 5everity

Default Incident Distribotion Street p",- Lanes p",- p",- Joint
with Shoulder Presence Location Tvoe DOrtion Affected DOrtion 5everitv'" DOrtion ProDOrtion
Segmeo' Oo,h 0.358
,."" 0.335 Fl
PDO
0.304
0.696
0.036
0.083
2+ lanes 0.163 Fl 0.478 0.028
PDO 0.522 0.030
5hoolder 0.502 Fl 0.111 0.020
PDO 0.889 0.160
Non- 0.642 11ane 0.849 Breakdown 0.836 0.456
aash 00" 0.164 0.089
2+ lanes 0.119 Breakdown 0.773 0.059
Othec 0.227 0.017
Shoulder 0.032 Breakdown 0.667 0.014
00" 0.333 0.007
Total: 1.000
Signalized Crash 0.310 11ane 0.314 Fl 0.378 0.037
intersection PDO 0.622 0.061
2+ lanes 0.144 Fl 0.412 0.018
PDO 0.588 0.026
Shoulder 0.542 Fl 0.109 0.D18
PDO 0.891 0.150
Non-
crash
0.690
,."" 0.829 Breakdown
00"
0.849
0.151
0.486
0.086
2+ lanes 0.141 Breakdown 0.865 0.084
00" 0.135 0.013
Shoulder 0.030 Breakdown 0.875 0.018
00" 0.125 0.003
Total: 1.000
SOurce: Z~etal.(l).
Note: • FI = fal<ll or inJury crash; PDO = property-damage-only crash; other '" not breakdown (e.g., debris).

Exhibit 17.12 Inciden' TyDe Inddent Location Inciden' seyerttv

Default Incident Distribution Street p",- Lanes p",- p",- Joint
Without Shoulder Presence Location TVDe DOrtion Affected DOrtion 5everitv" Dortion ProDOrtion
5egment Crash 0.358 1 tane 0.837 '1 0.304 0.091
PDO 0.696 0.209
2+ lanes 0.163 Fl 0.478 0.028
PDO 0.522 0.030
Non-
,,,'"
0.642
,."" 0.881 Breakdown
00"
0.836
0.164
0.473
0.093
2+ lanes 0.119 Breakdown 0.773 0.059
00" 0.227 0.017
Total: 1.000
Signalized
intersection
Oo,h 0.310
,."" 0.856 Fl
PDO
0.378
0.622
0.100
0.165
2+ lanes 0.144 Fl 0.412 0.018
PDO 0.588 0.026
Noo- 0.690 11ane 0.859 Breakdown 0.849 0.503
crash Othec 0.151 0.089
2+ lanes 0.141 Breakdown 0.865 0.084
Othec 0.135 0.013
Total: 1.000
SOurce: Zegeer et al. (J).
Note: • FI '" fatal Of inJury crash; PDO = property-damage-ooly crash; other = not breakdown (e.g., debris).

Work Zone and Specia/ Event Data

Work zones and spedal events require the use oí altemative datascts that
spedfy the demand, geometric, and traffic control conditions in effect during the
work zone or spedal evento A schedule (Le., start and end dates) is also required
that spedfies when the work zone is in effed or when the sredal event takes place.

Core MethodoIogy Chapter 17/Urnan Street Reliability and ATDM

Page 17-22 VefSiQn6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Traffie Counts
The date and nme of the traffic count represented in the base dataset are
required inputs. If the base dataset demands are computed by using planning
procedures, they are assumed to represent average day volumes. In this case, a
date does not need to be provided by the analyst. However, the time of day for
which the estimated volumes apply is still needed. The date and time of the
traffic count represented in an alternative dataset are also required inputs.

Data Sources
Reliability (as measured by Tri or PTI) can vary widely with the
charaderistics of a particular facility. Therefore, analysts are encouraged to use
local values representativc of local dcmand, weather, and incident pattcrns whcn
the data are available. In addition, analysts must supply local values for work
zones and special events if they wish to account for these effeds in a reliability
analysis. This subsection identifies potential sources of these data.

Demand Pattern Data

The best potential source of demand pattero data is a permanent traffic
recorder (PTR) located along the facility. Alternatively, an analyst ma)' be able to
use data fram a PTR located along a similar facility in the same geographic area.
Many state departments of transportanon produce compilations of data fram
thcir PTRs and pravide demand adjustment factors by time of day, day of week,
and month of year by facility and area type. The analyst is reminded that
measured volumes are not necessarily reflective of demands. Upstream
bottlenecks may limit the volume reaching a PTR or other observa non point.

Weather Data
NCDC provides rainfall, snow, and temperature statistics for thousands of
locations thraugh its website (11) and average precipitation rate data in the
Rain/ull FrequellCY Atlas (12).
A weather station that a transportation agency has installed along the study
facility may also be able to provide the required data, if the agency stores and
archives the data collected by the station. A lO-year weather dataset is desirable
for capturing weather events that are rare but have a high impact.
Final1y,analysts should consider the location of the facility relative to the
weather station. Elevation differences, praximity to large bodies of water, and
other factors that create micraclimates may resuIt in significant differences in the
probabilities of certain types of weather events (e.g., snow, fog) on the facility
and at the weathcr station.

Incident Data
The base crash frequency for a segment or intersection can be computed with
the predictive method in Chapter 12 of the 2010 HSM (13). U this method cannot
be used, the base crash frequency of a segment or an intersection can be
estimated on the basis of its 3-ycar crash history. However, crashes that OCCUf
when work zones and spedal events afe present should be removed fram the
crash data. In this situation, the expected crash frequency is computed as the

Chapter 17tUrban Street Reliability and ATDM (ore Methodology

Vet$iOn 6.0 Page 17-23
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

count of crashes during times when work zones and spedal eveots are oot
present divided by the time period when work zones and spedal events are not
presento Thus, if 15 crashes were reported during a reccnt 3-year period and five
of the crashes occurred during a 6-month period when a work zone was present,
the base crash frequency is estimated as 4.0 crashes per year [= (15 - 5)/(3 - 0.5)}.
A technique for distinguishing between segmento and intersection-related
crashes is described in Appendix A of Part e of the 2010 HSM (13).

Work Zone Data

A schedule of long-term work zones should be obtained from the roadway
operating agency. The schedule should indicate the days and times when the
work zone will be in force and the portions of the roadway that will be affected.
Work zones that vary in intensity (e.g., one lane c10sed 00 sorne days and
two lanes c10sedon others) or that affect different segments at different times
will need to be provided as separate altemative datasets.
When detailed traffie control plans for each work zone are available, they
should be consulted to determine the starting and ending locations of lane
c1osures, along with any reductions in the posted speed. When detailed plans are
not available, the agency's standard practices for work zone traffic control can be
consulted to determine the likely traffic control that would be implemented,
given the project's charaeteristics.

Special Event Data

Special events are short-term events, such as major sporting events, concerts,
and festivals, that produce ¡ntense traffie demands on a facility for limited
periods. Spedal traffic control procedures may need to be implemented to
aecommodate the traffic demands before, during.. or after these events.
The analyst should identify whether any events that occur in or near the
study area warrant spedal treatment. If so, a schedule for the evcnt (date,
starting time, duration) should be obtained. Sorne types of events also have
varying intensities that will require separate treatrnent (e.g., a sold-out baseball
game compared with a lower-attendance midweek game). Recurring events may
have developed spedal traffic control procedures; if so, these plans should be
consulted to identify any changes required from base conditions. Altemative
datasets will be needed for each combination of spedal event venue and event
intensity.

OVERVIEW OF THE METHOOOLOGY

This subsection describes the methodology for evaluating the reliability of an
urban street facility. The methodology is computationalIy intense and requires
software to implement. The intensity stems from the need to create and process
the input and output data associated with the hundreds or thousands of
scenarios considered for a typical reliability reporting periodo
The objective of this section is to introduce the analyst to the calculation
process and discuss the key analytic procedures. Important equations, concepts,
and interpretations are highlighted.

Core Methodology Chapter 17fUrban Str'eet Reliability and ATOM

Page 17.24 VersJon6.0
 Highway Capacity Manual: A Gujde far Multimadal MabiJjty Analysis

The computational details of the methodology are provided in Chapter 29,

Urban Street Facilities: Supplemental. The STREETVALcomputational engine
provided in the Teehnical Refercncc Library in the online HCM Volume 4
represents the most detailed description of the methodology.

Framework
The scquence of eakulations in the reliability methodology is shown in
Exhibit 17.13.There are five main steps: (a) establishing base and altemative
datasets, (b) generating seenarios, (e) evaluating each scenario with the Chapter
16 methodology, (d) compiling travel times for each analysis period in the
reliability reporting period, and (e) producing rcliability performance measures.

Base Dataset Alternatlve Dataset

Exhlbit 17.13
Reliability Methodology

-
Segment Geometry Changes to Base Dauset Framework
Trafflc Control Inpuls Due to Work Zones
Base Ol'mands and Special Events
Demand Factors SChedule forWork Zones
Weather iIInd Incldent Hlstorv and Special Events
Reliabillty Reportin¡: Period
Residual Queue by
Analysis Period
Input OatasetAdjustments
Scenario Generator Adjusted Demand Core HCM Facirtv Method
Demand Varialion f- Adjusled C<ipacity ~ Chapler 16
Weather Adjusled Runníne Speed
tUrban 5treet Facilitiesl
lncldents Adjusled Saluration Flow Rale
Residual Queue -+ Inltlal Queue j
Travet Time Distribution
Performance Measures
-
i'"
,-

\
Planning Time IndeK
80th Percenllle Travel TIme lndu
Reliability Ratine
~ 1::
"o
1,.

.. -
On.Time Performance
1•
," "' ., .,
Semi.Slandard Devlalion •
etc. " ...........
'-
Data Depository
Every reliability analysis requires a base dataset. This dataset describes the
traffie demand, geometry, and signal timing conditions for the intersections and
segments along the facility during the study period when no work zones are
present and no spedal events occur.
Additional datasets are used, as needed, to describe the conditions when a
specific work zone is present or when a special event occurs. These datasets are
caBed the altemative datasets. One alternative dataset is used for each time
period during the reliability reporting period when a specific work zone is
present, a spedfic spedal event occurs, or a unique combination of these
conditions oerurs during the study periodo
As a first step in the reliability evaluation, the analyst develops the
aforementioned datasets. Thcn the analyst assembles the input data needed for
the reliability methodology. These input data are deseribed in the previous
subseetion titled Required Data and Sourccs.

Chapter 17{Urban StJeet Reliability and ATDM Core Methodology

Version 6.0 Page 17-25
Hjghway Capacity Manual: A Guide for Multimodal Mobmty Analysis

SCenario Generation
The seenario generation stage consists of four sequential procedures: (a)
weather event generation, (b) traffie demand variation generation, (e) traffie
incident generation, and (d) scenario dataset generation. Each procedure
generates in ehronological order the set of analysis periods that make up the
reliability reporting periodo This subsection gives an overview of the scenario
generation process; a detailed description is provided in Chapter 29, Urban Street
Facilities: Supplemental.

Weather Event Generation

Future research may ¡ndiCate The weather event procedure genera tes rain and snow events during the
that additional weather types
affect arterial operation. At tilis reliability reporting periodo The dates, times, types (Le., rain or snow), and
time, avaifable research durations of severe weather events are generated. These data are used to adjust
supports assessment of rain
and snow events on arterial the saturation flow rate and speed of facility traffie for each analysis periodo The
operation. procedure alsa predicts the time after each weather event that the pavement
remains wet or cavered by snow or ice, sinee the presence of these conditions
influences running speed and intersection saturation flow rateo

Traffie Demand Variation Generation

The traffie demand variation proeedure identifies the appropriate traffie
demand adjustment factors for each analysis period in the reliability reporting
periodo A set of factors accounts far systematic demand variation by hour of day,
day of week, and month of year.

Traffie Ineident Generation

The traffie incident procedure generates incident dates, times, and durations.
It also determines incident types (Le., erash or nonerash), severity levels, and
loeations on the facility. Loeation is defined by the intersection or segment on
which the incident oecurs and whether the incident occurs on the shoulder, in
ane ¡ane, ar in muItiple lanes. The proeedure incorporates weather and traffie
demand variation information from the previous proeedures in generating
incidents.

Scenario Dataset Generation

The seenario dataset generation procedure uses the results from the
preceding procedures to develop one HCM dataset for each analysis period in
the reliability reporting periodo Each analysis period is considered to be one
scenario. The base dataset is modified to reflect eonditions present during a
given analysis periodo Traffic volumes are modified at each intersection and
driveway. Saturation f10w rates are adjusted at intersections influenced by an
incident or a weather event, and speeds are adjusted for segments influenced by
an incident or a weather event. Dates and times represent a common basis for
tracking events and conditions from one analysis period to the next.

Core Methodology Chapter 17/Urban 5treet Reliability and ATDM

Page 17-26 Version 6.0
 Highway Capacity Manual: A Guide for Mulfimodal MobUity Analysis

Facility Evaluation
The fadlity evaluation stage consists of two tasks that are repeated in
seguenee for eaeh analysis periodo The analysis periods are evaluated in
chronological arder.
First, the dataset assodated with a given analysis period is evaluated by
using the urban street facility methodology. The performance measurcs output
by the methodology are then archived.
Second, the dataset assodated with the next analysis period is modified, if
neeessary, on the basis of the resuits of the current analysis periodo SpedficaHy,
the initial queue input value for the next analysis period is set egual to the
residual qucue output for the currcnt analysis periodo

Performance Summary
The performance summary stage consists of two seguential tasks. First, the
analyst identifies a spedfie direction of travel and the performance measures of
interest. The desired perfonnance measures are extracted from the fadlity
evaluation archive for each analysis period in the reliability reporting periodo
Available rneasures, as defined in Chapter 18, Urban Street Segments, are as
follows:
• Travel time,
• Travel speed,
• Stop rate,
• Running time, and
• Through delay .
The analyst also indicates whether the performance measures of interest
should be representative of the entire facility or a spedfic segment. The first three
measures in the aboYe list are available for faeHity evaluation. AHfive measures
are available for segment evaluation. At the condusion of this task, the collected
data represent observations of the performance measurcs for each analysis period
occurring during the reliability reporting period (or a sampled subset thereof).
Next, the selected performance measure data are summarizcd by using the
following statistics:
• Average;
• Standard deviation;
• Skewness;
• Median;
• 10th, 80th, 85th, and 95th pereentiles; and
• Number of observations.
In addition, the "average" base free-flow speed is always reported. This
measure is computed as the arithmetic mean of thc base free-flow speed for each
scenario in the reliability reporting periodo It can be used with one or more of the
distribution statistics to compute various variability and reliability measures,
such as the TII and the reliability rating.

Chapter17/Urban5treetReliability
andATDM CoreMetI1odology
V£'f>iOn 6.0 Page17-27
Highway Capacify Manual: A Guide fer Multimedal Mebility Ana/ysis

Interpretation of Results
Identifying ReliabJ1ityProblems
In a perfect world, aH urban street facilities would be perfectly reliable. They
would have mean TrIs and PTIs of 1.00 or better. Since operating a perfedIy
reliable facility is not a realistic standard, an ageney must distinguish between
Iess than perfed-but still aceeptable-reliability and unacceptabIe reHability.
This is obviously a choice that each ageney must make. This subseetion provides
guidanee on the faetors and eriteria that a transportation agency may wish to
eonsider in making its selection, but the final decision is up to the ageney.

Criterion No, 1: How Does Reliabi/ity Compare with Agency Congestion

Management Policy?
An agency may have a policy of delivering a eertain minimum speed or
maximum travel time on its urban streets, or a maximum aceeptable deJay per
signal or per mile. If so, the eomputation of the reliability statistics can be
modified on the basis of this infonnation so that faHures to meet agency policy
can be identified more quickly. TItis approach is iIIustrated in the following
paragraphs.
Minimum speed paliey.lf the ageney has a minimum acceptabJe facility
speed policy, this information can be used to compute the reliability statisties
that use the aeceptable speed as a baseline (instead of the base free-f1ow speed).
Determining the extent to which the facility meets the ageney's target
performance level by eomparing the eomputed reliability statistie with the target
value of 1.00 is then reJatively easy, The result of using the poliey speed instead
of the base free-f1ow speed is to negleet travel time reliability when speeds
exceed the ageney's minimum acceptable threshoJd.
For exampIe, if the ageney's congestion management policy is to deliver
speeds in excess of 40 mi/h, the poliey travel times are computed by using the
facility Jength divided by 40 mi/h and converting the result to minutes. The
poliey travel time index is then eomputed with the foIlowing equation:
TTF
Equation 17-1 TTlpolicy = =-
TTp

with
_
TTF=---x--x
1
Na Nap
3.600
5.280 I I Is
a=l ap=1 ¡=1
L,
T,seg,l,ap,a

- 3,600
TT. =--x
p 5,280
I
m
Lj
-
40
jo'

where
ITlpolicy = palicy traveI time index, based on the agency's poliey (ar target)
traveI time for the facility (unitless);
TTF = average travel time far through trips on the facility during the
reliability reporting period (s);

CoreMethodology Olapter17/Urban5treetReliability
andATDM
Page 17-28 Ver90n 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

TTp "" ageney's maximum acceptable travel time for through trips on the
facility during the reliability reporting period (s);
Li "" length of segment i (ft);

m "" number of segments on the facility;

N"" "" number of analysis periods in 1 day (i.e., study period);

Nd "" numbcr of days in the reliability reporting period; and

Sr,><¡:.l"",d "" travel speed of through vehides for segment j during analysis period
ap and day d (mi/h).
Values of LOOor less for ITlpoli,..,. mean that the agency's congestion
management poliey is being met on average over the course of the reliability
reporting periodo Values greater than LOOmean that the facility is failing to meet
the agency's policy on average.
Maximum acceptable delay. If the ageney has a maximum acceptabJe delay
standard per mile or per signaJ, the mean ITI can be readily converted into
equivalent delay estima tes for the facility and compared with the agency
standard. The following equations illustrate this conversion.
du'ip = TT{O.F x (TTlmean - 1) Equation 17-2

TT{o.F
dmi1e = Em L, x (TT/me•n - 1)
1=1 1

TT{o,F
dsignal = -N- x (TT/mean - 1)
,
with

_
TTI
o.F -
1
----x--x
Nd Nap
3,600
5,280 I I Is
d=1 ap"'1 1=1
L,
{o.seg.l.ap.d

where
d1rip average deJay per trip (s/veh);

dm¡le average delay per mile (s/veh);

doignol average delay per signal (s/veh);

IT/"",an '" mean travel time index (unitless);

TT{o.F '" average travel time for through trips at the base free-flow speed on
the facility during the reliability reporting period (s);
Sfc,~i.I'l',d = base free-flow speed of through vehides for segment j during
analysis period ap and day d (mi/h); and
N. = number of signals within study section of facility (unitless).

These delay values can also be computed by using the PTI (or any other
percentile ITI value) by substituting the desired TII value for TTlmean in the
appropriate equation. A poliey ro value can also be substituted.

Qlapter 17/Urban 5treet Reliability and ATDM Core Methodology

VetWn 6.0 Page 17-29
Highway Capadty Manual: A Guide tor Multimodal MobiUty Analysis

Critenon No. 2: How Does ReliabJ'lity Compare with Other FacJ'lities?

To address this question, the agency ranks the reliability results far a given
facility against those of other facilities it operates and prioritizes improvements
to its facilities with the worst reliability accordingly. Of course, this approach
requires that the agency coUed reliability data for its facilities so that the
agency's facility investments can be properIy ranked according to need.

Critenon No. 3: How Does Reliability Compare with HCM LOS?

This criterion involves translating reliability results into more traditional
HCM LOS results with which decision makers may be more comfartable. The
reliability results are used to identify what percentage of time a facility is
operating at an unacceptabIe LOS and in detennining a percentage of time that is
unacceptable.
For example, the agency's standard may be LOS D. The reliability results
may show that the facility operates at LOS E or worse during 5% of the weekday
peak periods over the course of ayear. This may be an acceptable risk for the
agency if the costs of improvements to eliminate the 5% risk are high.
The inverse of the m represents the 95th percentile slowest through-trip
spced on the facility divided by the base free-flow speed. Thus, the inverse PII
can be used to determine whether the facility will operate at a LOS acceptable to
the ageney at least 95% of the time. In this regard, the inverse PII (multiplied by
100 to yield a percentage) is compared with the base-free-flow-spced percentage
associated with the LOS considered acceptable. The percentage associated with
each LOS is described in the subsection titled LOS Criteria, Motorized Vehicle
Mode, in Section 2 of Chapter 16.

Diagnosing the causes of Reliability Problems

Exhibit 17-14identifies seven sources of congestion and unrcliability and
shows how they interad with each other. The starting point in traditional
analysis is to take a Hxed capacity and a Hxed volume to develop an estimate of
delay, usually for "typical" conditions. However, in the fieId, both physical
capacity and demand vary because of roadway disruptions, travel patterns, and
traffic control devices. These conditions not onIy decrease available capacity ar
cause volatility in demand but also interad with each other. Far example, both
inclement weather and work zones can Iead to an increase in incidents.
Thus, diagnosing the relative contribution of different causes of unreliability
involves identifying the causes individualIy and in combination. Depending on
the purpose of the evaluation, various approaches may be taken far assigning the
proportional responsibility to individual causes when two or more are acting in
combination.

Core Methooology Olapter 17/Urban 5treet Reliability and ATDM

Page 17-30 Version 6.0
 Highway Capacity Manual: A Gujde for Multimadal MobiJity Analysis

Traffic Control Planned Exhibit 17-14

Daily/selSonal Speclal
<D Interrelationship Between
Devices VlnaliDn <Zl £Venls <Il Erner¡encies causes of Congestion and the
_dele,mir>e_
Facility

_deue ••• ,""p'dt'o'

_lnleracu with .• Demand .nd dllrce dem.nd .•
Volume
Roadway D1srup1lons

I w","" mi

Total
ConlesliDn and
Reliability

Selecting a Performance Measure

To identify the relative effects oC different causes on the travel time reliability
oC the Cacility,computation oC total vehide (or person) hours oC delay summed
over the entire reliability reporting period is recommended. This measure oC
eHectiveness takes into account both the severity of the event (Le.,demand surge,
incident, weather) and its frequency of occurrence within the reliability reporting
periodo Exceptionally severe bul rare events may add relatively Iiule to the total
annual delay experienced by the facility. Moderate but freguent events will often
have a greater eHect on total annual delay.

Generating a Simplified Matrix of Causes

Identifying paUems of results in several thousand scenarios is impractical, so
consolidation of the many scenarios into a matrix of congestion causes along the
lines of Exhibit 17-15is recommended. This is best done by combining similar
scenarios that individually contribute less than 1% to annual delay. In the
example shown in Exhibit 17-15,the numerous severe weather events (rain,
snow, etc.) have been consolidated into a single "bad weather" category because
severe weather is relatively infreguent at this site. The results fram the original
analysis of multiple demand leveIs have similarly been consolidated into three
levels (Iow, medium, high).

Chapter 17/Urban 5treet Reliability and ATDM Core Methodology

VelSiOn 6.0 Page 17-31
Highway Capadty Manual: A Guide for Multimodal MobUity Analys;s

Exhibit 17-15 low Demand Moderate pemand High Pemand

Example Matrix Allocating Fair BaO Fair BaO Fair BaO
Annual Vehicle Hours of Delay lncidents Weather Weather Weather Weather Weather Weather Total
by cause 596 407 818 362 6,240 '56 9,379
N"", 2% 1% 3% 1% 23% 4% 34%
1 lane 2,363 92 2,097 61 9,102 11' 13,834
dosed
21anes
,%
194
<1%
13
8%
189 ,
<1% 33%
907
<1%
17
51%
1,329
dosed 1% <1% 1% <1% 3% <1% 5%
3lanes 621 40 468 23 1,510 32 2,694
d_ 2% <1% 2% <1% 6% <1% 10%
3,774 552 3,572 455 17,759 1,124 27,236
Total 2% 13% 2% 65% 4% 100%
14%

Diagnosing Primary causes of Unreliability

The diagnosis proceeds by first examining the cells oí the matrix to identify
those with the largest annual delay values. For example, examination of the ceIls
in Exhibit 17.15 yields the following conclusioos:
• The single greatest cause of annual deJay 00 the example facility is
incidents closiog a single lane under high-demand conditioos on fair-
weather days. They account for 33% of the armual delay 00 the facility.
• The next largest occurrence of armual delay happens under high.demand,
fair-weather, no-incident cooditions. They account for 23% of the annual
delay 00 the facility.
• The third and fourth largest annual de!ays occur when incidents close a
single lane under fair-weather conditions with low- to moderate.demand
conditions. Together, these scenarios account for 17% of the armual delay
on the facility.
• The fifth largest annual delays are accumulated when incidents c10se
three lanes under high-demand and fair-weather conditions.
Exhibit 17.16 shows that the top five cells in Exhibit 17-15account for about
78% oí the aTInua!delay on the facility.
Exhibit 17.16
Example Pie Olart of
Congestion causes
High Demand, Fair
Weather¡ l-lane
Inddeflts
33%
High Demand, Fair
Weathel; 3'lane
loddeots
6%

Modefate
Demand, Fair
Weather, l-Lane
loddeots
8%

Low Demand, Fair High Demand, Fair

Weather; 1.lane --~ Weathel; No
Inooents Inooents
9% 23%

Core MethocIology Olapter 17/Urban $treet Reliability aOOATDM

Page 17.32 Versk1n 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

The next step is to examine the row and column total s to determine whether
a single cause stands out. For example, examination of the row and column totals
in Exhibit 17-15 yields the following conclusions:
• The highest row or column total annual delay occurs in high-demand,
fair.weather conditions. Recurring congestion is therefore a significant
source of delay on this example facility. High-demand conditions account
for 65% oE the annual delay on the facility.
• The next highest row or column total occurs when incidents close one
lane on the facility. Incidents blocking a single lane account for 51% oE the
delay on the facility.
• Bad weather is a minor cause oE annual delay on the facility.

Developing a Treatment Plan

The conclusions Erom the example shown in Exhibit 17-15 suggest the
following options that are likely to have the greatest effect on improving
reliability in the example facility:
• Measures to reduce high.demand conditions or to increase capacity to
address recurring congestion, and
• Measures to manage incidents that close a single lane.
The diagnostic process also indicates that in this example, bad weather and
extreme incidents (closures of two or more lanes), despite their severity when
they happen, are minor contributors to total annual delay on the Eacility.
The particular example used here was from a state with relatively mild
weather. The results would likely be different on facilities in other parts of the
country.

Chapter 17/Urban 5treet Reliability and ATDM Core Methodology

Version 6.0 Page 17-33
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

4. EXTENSIONS TO THE METHODOLOGY

ACTIVE TRAFFIC ANO OEMANO MANAGEMENT STRATEGIES

ATDM strategies are essentially real.time changes in geometry, lane
assignment, or traffic control that are implemented in the time frame of a few
seconds. However, macroscopic deterministic models (like the methods
described in the HCM) are generally limited to the evaluation of strategies that
have a fixed geometry, lane assignment, and traffic control plan during the
analysis time frame of 15 min to 1 h.
On the basis of these observations, the HCM appears unable to evaluate an
ATDM strategy reliably, and to a certain extent that is true. However, in making
investment dedsions, the focus is not just on the next few seconds. Instead, the
focus is on how the investment will perform over the coming years. For this kind
of ATDM strategy evaluation, macroscopic deterministic models (like the
methods in the HCM) are appropriate tools. The key is to idcntify where and
how thcsc tools might be adapted to recognize the superior efficiency of ATDM
strategies (by delivering improved productivity) when compared with more
static management approaches.
At this point relatively little research has becn conducted on the demand,
capacity, speed, and deJay effects of urban street ATDM strategies. Therefore, the
guidance in this section is general; it ¡caves sorne dedsions up to the discretion of
the analyst. The focus is on using the methodologies in HCM Chaptees 16 to 23 to
evaluate ATDM strategies for the purpose of making investment decisions.

EXTENSIONS FOR SPECIFIC TACTlCS

Arterial Monitoring Tacties
Arterial monitoring tactics consist of an improved ability to detect vehicles
and pedestrians at a finer level of detail than is possible with conventional
pavemcnt loop detectors. In the future, connccted vehicles and pedestrians
(using cell phones) may be able to place calls to the signal system in advance of
arrival and provide sufficient information to enable the system to sort out the
relative priorities of the vacious vchicles and pedestrians and assign signal times
accordingly. Spcdal guidance may be provided to visualIy impaired pedestrians
arriving at a signalized intersection.
AH of the aforementioned monitoring tactics provide additional (and more
accurate) information to the signal system controller. However, the impad of a
spedfic measure on urban street performance will depend on what the control
system does with thc additional information. The ability to evaluate more
accurate monitoring will generally require no modification or extension of the
HCM procedures bt.'Causethey are based on the assumption that the controller is
fully aware of when vehicles and pedestrians aerive at the intersection. One
exceptiún wúuld be if the monitoring tactiés fundion is tied to changes in demand
or the antidpatcd demando In this case, the HCM proccdures would need to be
revised tú replicatc the control system's response tú the changing demando

Extensions te the Methodology Olapter 17fUman 5treet Reliability and ATDM

Page 17-34 VeiSion 6.0
 Hjghway Capacity Manual: A Guide for Multimodal Mobility Analysis

Signal and Speed Control Tactics

Signal and speed control tactics indude adaptive control priority servicc for
emergency response vchicles, transit vehicles, freight vchicles, and pedestrians.
The means by which each measure might be evaluated by using the HCM is
described in the following paragraphs.

Adaptive Control
Adaptive control is accomplished through sccond-by.second optimization of
signal timings according to the current monitor information and the priorities
assigned to each vehicle and pedestrian type by the opcrating agency. To
cvaluate this measure by using the HCM, sorne extensions are needed for the
following computational steps of the motorized vehide methodology in Chapter
18, Urban Streets Segments:
• Step 2: Determine running time.
• Step 3: Determine proportion arriving during grecn.
In the case of adaptivc signal control, the phase splits used with the HCM
method should be large enough to accommodate aH of the prcdicted average
hourly fiows, subject to any limits on maximum phase length incorporated into
the adaptive control algorithm.
The intersection peak hour factor should be set to 1.00 and only average
hourly fiows used in the HCM analysis, since the adaptive control will adapt the
effective cyde length to address any fiuctuations in demand within the houe.
The free-fiow speed and running times produced by Step 2, Chapter 18, will
probably need to be adjusted to refied the lower likelihood of stopping with
adaptive control (for suitably favorable demand conditions).
Similarly, the proportion arriving during grccn, produced in Step 3,
Chapter 18, will nced to be adjusted to reflect the lower likelihood of arriving on
red with adaptivc control (for suitably favorable demand conditions).
Under high-demand conditions, an adaptive control algorithm may revcrt to
essentially pretimed control, in which case the HCM methodology for pretimed
control can be used without adaptation.
If a range oi demand conditions are expected, a set of high-, medium., and
low.demand scenarios should be evaluated with the reliability methodology
described in Section 3.

Priority of Emergency Vehic/es, Transit Vehicles, Freight, and Pedestrians

The HCM methodology can be used to evaluate the effects of priority
treatments by performing a reliability analysis that identifies different evaluanon
scenarios. One sccnario is developed for each priority treatment condition (e.g.,
bus present so transit priority invoked, no bus so transit priority not invoked).
The HCM methodology is then used to evaluatc performance associated with
each scenario. The results are weighted by the probability of each priority caIl to
obtain annual performance effects.

Chapter 17/Urban 5treet Reliability and ATDM Extensions lo the Methodology

Version 6.0 Page 17-35
Highwoy Copocity Monuol: A Guide tor Multimodol Mobility Anolysis

The HCM methods for estimating free.flow spL'ed, running time, and the
proportion arriving on green should be modified in each scenario to reflect the
conditions for each priority callo

Speed Harmon;zation
Speed harmonization involves dynamically slowing traffic in advance of
queues, incidents, and lane dosures and then directing traffic to the remaining
lanes. This ATDM strategy will affect lane utilization, free-flow speed, running
time, and the proportion of vehides arriving on green in the HCM analyses. The
appropriate speed harmonization scenarios should be created and then
evaluated with the reliability methodology described Section 3.

Geometric Configuration Tactics

Geometric configuration tactícs dynamically reassign traffic movements to
different lanes in response to demand surges. For example, the right.hand
through lane may become an exdusive right-tum lane during periods of high
right-tum demando
Dynamic lane assignments, reversible lanes, and dynamic tum lane
restrictíons are evaluated in the HCM by using scenarios. One scenario is created
for each desired lane configuration. The reliability methodology described in
Section 3 is then used to evaluate the performance of each scenario and combine
the results into an overall assessment of the whole year performance of the tactic.

Demand Modification Tactics

Demand modification tactics seek to change the demand for the facility.
Traveler information, route guidance, dynamic parking, and congestion pricing
are evaluated in the HCM by using scenarios. The analyst must estímate the
likely demand effects of the tactic and then input this demand into the HCM
analysis. One or more scenarios are created for each desired demand
modification tactic. The reliability methodology described in Section 3 is then
used to evaluate the performance of each scenario and combine the results into
an overall assessment of the whole-year performance of the tacHe.

Extensions to the Methodology Qlapter 17¡Urban 5treet Reliability and ATDM

Pa<]e 17-36 VefS'on 6.0
 Highway Capacify Manual: A Guide for Multjmodal Mobilify Analysis

5. APPLlCATlON5

EXAMPLE PROBLEMS
Chapter 29, Drban Street Facilities: Supplcmcntal, describes the application
of the reliability methodology through the use of example problems. There is one
example problem iIIustratíng the use of the methodology to diagnose causes of
reliability problems and one example problem iIIustrating the use of the
methodology for altematives analysis.

ANALYSIS TECHNIQUES
This subseetion describes techniques for eHectíve use of the reliability
methodology.

Work Zones and Special Events

Work zones and special evcnts influence traffic demand levels and travel
pattems. To minimize the impact of work zones and special events on traffic
operation, agencies responsible for managing traffic in the vicinity of a work
zone or special event often reallocate sorne traffic lanes or alter thc signal
operation to increase the capacity of specifie traffie movements. These
characteristics mean that the eHect of each work zone and spedal event on
fadlity performance is unique. Multiple work zones and special events can ocrur
during the reliability rcporting periodo
The reliability methodology incorporates work zone and spedal event
influences in the evaluation results. Howevcr, the analyst must describe each
work zone and spedal event by using an alternative dataset. Each dataset
describes the traHic demand, geometry, and signal timing conditions when the
work zone is present or the spedal event is under way. A start date and a
duration are assodated with each dataset.
The presence of a work zone can significantly aHect traffie demand l('veIs.
The extent of thc cHect will depcnd partly on the availability of altcrnative
routes, the number of days the work zone is in opcration, and the volume-to-
capadty ratio of the segment or intersection approach within the work zone.
When the reliability methodology is used, the analyst must provide
estima tes of traffic demand volumes during the work zone or spedal evento The
estimates should reflect the eHect of diversion and can be based on Held
measurements, judgment, or areawide traffic planning models. They are
recorded by the analyst in the corresponding altemative dataset.
The analyst must ha ve information about lane c1osures, alternative lane
assignmcnts, and spedal signal timing that is prescnt during the work zonc or
spedal evcnt. The information can be based on agency policy or expericnce with
previous work zones or events. The available lanes, lane assignments, and signal
timing are recorded by the analyst in the corresponding aIternative dataset.

Olapter 17/Urban 5treet Reliability and ATDM Applications

Ver.s1On 6.0 Page 17-37
Highway Copocity Manual: A Guide far Multimadal Mobilify Analysis

Multiple 5tudy Periods

The geometric design elements, traffic control features (induding signal
timing plans), and directional distribution of traffic are assumed to be constant
during the study periodo If any of these factors varies significantly during certain
periods of the day (e.g., moming peak or evening peak), each unique period
should be the focus of a separate reliability evaluation. In this regard, each
unique period represents one study periodo
When multiple study period evaluations are undertaken for a common
facility, the set of analysis period averages for each evaluation can be merged to
evaluate the overall reliability. In this manner, the combined data for a given
performance measure represent the distribution of interest. The various
reliability measures are then quantified by using this combined distribution.

Alternatives Analysis
Weather events; traffic demand; and traffic incident occurrence, type, and
locatíon have both systematic and random elements. To the extent practical, the
reliability methodology accounts for the systematic variation component in its
predictive models. Specifically, it recognizes temporal changes in weather and
traffic demand during the year, month, and day. It recognizes the influence of
geographic location on weather and the inf]uence of weather and traffic demand
on incident occurrence.
A Monte Gario approacn uses Models of the systematic influences are induded in the methodology. They
essentially random inputs
(wit!Jin realiStiC Jimits) lo model are used to predict average weather, demand, and incident conditions during
a system and produce probable each analysis periodo However. the use of averages to describe weather events
outromes.
and incident occurrence for such short time periods is counter to the objectives of
reliability evaluation. The random element of weather events, demand variation,
and traffic incident occurrence introduces a high degree of variability in the
collectíve set oE analysis periods that make up the reliability reporting periodo
Thus, replication of thcse random elements is important in any reliability
evaluation. Monte Carla methods are used for this purpose in the reliability
methodology.
A random number seed is used with the Monte Cario methods in the
reliability methodology. A seed is used so that the sequence of random events
can be reproduced. Unique secd numbers are separately established for weather
events, demand variation, and incidents. For a given set of three secd numbers, a
unique combination of weather events, demand levels, and incidents is estimated
for each analysis period in the reliability reportíng periodo
One, two, or three of the seed numbers can be changed to genera te a
different set of conditions, if desired. For example, if the secd number for
weather events is changed, a new series of weather events is created, and to the
extent that weather influences incident occurrence, a new series of incidents is
created. Similarly, the seed number for demand variation can be used to control
whether a new series of demand levels is created. The secd number for incidents
can be used to control whether a new series of incidents is created.
When altematives are evaluated, the analyst willlikely use one set of secd
numbers as a variance reduction technique. In this application, the same seed

AppliCations Olapter 17/Urban 5tTeet Reliability and ATDM

Page 17-38 Version 6.0
 Highway Capacity Manual: A Gujde for Multimodal Mobility Analysis

numbers are used for aH evaluations. With this approach, the results from an
evaluation of one alternative can bc comparcd with thosc from an evaluation of
the baseline condition. Any observed difference in the results can be aUributcd to
the changes assodated with the altcrnativc (i.c., thcy are not due to random
changes in weather or inddent events arnong the evaluations).

Confidence Intervals
A complete exploration of rcliability would likely entail the use of multiple,
separate evaluations of the sarne reliability reporting period, with each
evaluation using a separate set of random numher seeds. This approach may be
particularly useful when the fadlity has infrequent weather events or inddents.
With this approach, the evaluation is replicated multiple times, and thc
performance measures from each replication are averaged to produce a more
reliable estimate of their long-run value. The confidence interval (cxpressed as a
range) for the average produced in this manner can be computed with the
following equation:
S
C11_a = 2 X t(1-aj2).N-l x ..JN Equation 17-3

where
C/1-a confidence interval for the true average value, with a level of
confidence of 1 - a;
Student's t.statistic for the probability of a two-sided error of a, with
N - 1 degrecs of frecdom;
N number of rcplications; and
s = standard deviation of thc subjcct performance measure, computed
by using results from the N replieations.
The variable a equals the probability that the true average valuc lies outsidc
the confidenee interva1. Values selected for a typically range from 0.05 (desirable)
to 0.10. Selected values of Student's t-statistic are provided in Exhibit 17-17.

Exhibit 17-17
student's t=Statistie tor Two values of (J
Student's t.Statistic
Number of Replications a = O.OS a = 0.10
3 4.30 2.92
4 3.18 2.35
5 2.78 2.13
10 2.26 1.83
15 2.14 1.76
30 2.05 1.70

Scenario Sampling
Typical combinations of rcliability repocHng period, analysis period, weather
evcnt oecurrenee, and incident event oecurrence eould Icad to a large number of
uniquc seenarios. If the time rcquircd to evaluate all of these seenarios is
considered too great for sorne reliability applieations, a seenario sampling
approach can be used.
The seenario sampling teehnique is used to minimize the total evaluation
time while maintaining thc underlying distribution of event occurrenee. The

Chapter 17/Urban Street Reliability and ATDM Applications

VersíOn 6.0 Page 17-39
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

analyst will need to input the scenario evaluation interval/. This interval has
units of "days." The analyst can choose to evaluate every scenario for every day
(Le.,input "1"). Alternatively, the analyst can choose to evaluate every sccnario
for every other day (i.e., input "2"). More generally, the analyst can input any
integer number for the evaluation interval. Any number that is larger than ''1''
indicates that the evaluation will indude onIy a sample of the total number of
scenarios.
The evaluation interval is checkcd to ensure that aH days in the reliability
reporting period are equally sampled. In this manner, the subset of scenarios to
be cvaluatcd is not biased to indude more of one weekday than another. The
check examines the pattern produced by the input "days of week considered" D
and evaluation interval/. An interval factor F is computed (F = 1- int(lID) xDj. If
five or seven days of the week are considered, values of 1that yield F> Oprovide
the desired representative sample. If two, three, four, oc six days of the week are
considered, values of 1 that yield F = 1 or F = D -1 provide the desired sample.
Any value of 1 that does not ml.'etthese conditions should be avoided because it
will yield a scenario sample that is biased toward a spedfic weekday.

USE CASES
Travel time reliability measures can be used for a number of planning and
roadway operating agency applications, induding those given in Exhibit 17-18.
Each of the applications listed in Exhibit 17-18 has several potential uses for
travel time reliability. For example, reliability may be assessed for existing or
future facilities to identify current problem spots and future deficiencies in
system operation. Reliability may provide additional performance measures that
can be used in generating and evaluating aIternatives. Reliability may
supplement conventional measurements for prioritizing improvement projects.

Exhibit 17-18 Application Use Cases for Travel Time Reliability

Use cases for Travel Time
• Identifying existing facilities not meeting reliability
Reliability Long.range transportation plan standards
Transportabon improvement • Identifying fUturefacilities oot meeting reliability
program standards
Corridor or area plans • Generating altematives to address reliability problems
• Evaluating reliability benefits of improvement alternatives
Major investJT1entstudies
• Prioritizing operational improvements and traditional
Congesbon management capacity improvements
Operabons planning • Evaluating tl1e probability of achieving acceptable
reliability and lOS
long-range planning: • Improving modeling of destinabon, time of day, mode,
demand forecasting and route choice

Planning has traditionally focosed on capacity improvements and has been

relatively insensitive to the reliability improvements that come with operations
improvements. Thus, reliability can become an important new measure in
identifying improvement aIternatives, evaluating their benefits, and prioritizing
them more accorately in relation to conventional capadty improvemcnts.
Reliability adds another dimension of information on facility performance
that can aid travel demand models in predicting the conditions under which

Applications O1apter 17tUrban Street Reliability and ATOM

Page 17-40 Version 6.0
 Highway Capacity Manual: A Guide tor Mulfimodol Mobility Analysis

people will choose to paya toll for more reliable service. Reliability wilJ enable
better destination, time-of-day, mode, and route choice models.

Use Case No. 1: Detecting Existing Deficiencies

This use case for the reliability methods in the HCM involves monitoring
conditions on a facility, identifying unacceptable performance, and detecting the
primary causes of unreliable facility operation. It involves selecting the
appropriate study period, performance measures, and thresholds of acceptance;
calibrating the HCM operations models; and expanding limited data to a full
reliability dataset.

Use Case No. 2: Forecasting Problems

This use case evaluates future reliability conditions on a facility, induding
the following:
• Expanding average annual (daily, peak period, or peak hour) volumes
(forecast demand) to the full variety of study period demands,
• Estimating facility travel times by time slice within the full study period,
and
• Comparing future with existing performance and identifying
"significant" changes in performance.
The following are among the forecasting questions that Use Case No. 2
addresses:
1. How to forecast weather:
a. Use of Monte Cario or expected value techniques to forecast the
frequency of future weather events.
b. Number of years that the forecast must be carried into the future
tu obtain a reasonably Iikely set of scenarios.
2. How to forecast incident frequency:
a. Use of Monte Cario or expected value techniques.
b. Number of future years that must be forecast to obtain a
reasonably likely sel of scenarios.
c. Prediction of the cHect of capacity improvements, demand
changes, and ATDM improvements on crash frequendes.
3. Dealing with congestion overflows (e.g., over thc entry link, over the
last analysis period) when performance measures are computed and
compared with existing conditions.
4. Calibrating this chapter's forecast reliability for future conditions to
field-measured reliability under existing conditions (for data-rich
agencies).

Use Case No. 3: Generating Alternatives

This use case identifies alternative operational and capadty improvements
for addressing reliability problems. Selection of operational and capadty
improvements that are Iikely to be best in addressing the primary causes uf
rcliability problems on the facility is induded.

Chapter 17tUman $treel Reliabitity and ATDM Applications

V~6.0 Page 17-41
Highway Capadty Manual: A Guide tar MultimOdal Mability Analysis

This case requires that the analyst

1. Determine that a reliability problem exists (see Use Case No. 6),
2. Diagnose the causes of the reliability problem, and
3. Identify promising treatment options for addressing the problem.
As part of the diagnostic process, the analyst needs to be able to identify the
fadlity's primary causes of unreliability and then identify two or three courses of
action to address those causes. This approach requires guidancc linking causes of
unreliability to cost-effective solutions that can be considered.

Use Case No. 4: Reliability Benefits of Alternatives

This use case computes the reliability effects of alternative operational and
capadty improvements for addressing reliability problems, inc1uding traditional
capacity improvements as well as more innovative ATDM measurcs.
While Use Case No. 3 was primarily about diagnosis, Use Case No. 4 focuses
on evaluating candidate trcatment options. The analyst fleshes out possible
treatments, estimatcs their effectiveness, and estimates their costs. This analysis
requires proccdures and parameters for computing the effects of capacity,
operational, and ATDM improvements on existing or predicted reliability.
Once an agency has performed enough of these analyses, it can probably
develop its own Case No. 3 diagnosis chart with locally specific treatment
options.

Use Case No. S: Prioritizing lmprovements

This use case applies reliability performance measures in combination with
other performance measures to prioritize investments in operational and
capacity improvements. Estimation of the relative values of mean travel time
improvements and travel time reliability improvements is inc1uded in this case.
While the reliability methodology provides results for only one facility at a
time, agencies putting together a regional program will want to combine the
results of individual facility analyses (freeways and urban streets) into a
prioritized tableoIn essence, the issue is how to weight the relative benefits of
reliability improvements versus more traditional capacity improvements. How
much is average travel time worth to the agency and the public, compared with
95th percentile travel time or sorne other measure of reliability?

Use Case No. 6: Achieving Acceptable Performance

This use case estima tes the probability of failure or the probability of
achieving acceptable performance. Performance may be reported as achieving a
minimum acceptable LOS.
This use defines and determines acceptable and unacceptable reliability
performance. Thus, it is a critical input to the diagnostic process of Use Case
No. 3. No diagnosis is nceded when it is determined that no reliability problem
exists. However, if Use Case No. 6 determines that a problem exists, Use Case
No. 3 is used to diagnose the causes and identify promising treatment options.

ApplicatiOns Chapter 17/Urban Street Reliability and ATDM

Page 17-42 Verron 6.0
 Highway Capacity Manual: A Guide for Multimodol Mobility Anolysis

Use Case No. 6 shares much with Case No. 5, but it introduces a new
concept, acceptability or failure. The numerical results produccd in Use Case
No. 5 are comparcd with sorne standard -a nationa!, state, or agency-specific
standard of acccptable performance.
This use case introduces the concept of defining a standard both as a
minimum acceptable performance level (such as LOS ar PTI) and as thc
probability af failing to achieve that level (Le.,probability of failure). The
standard is thus defined in two dimensions: a value and a probability of
exceeding that value.
Use Case No. 5 deals with nurnerical outputs that are compared with each
other (relativistic evaluation). In contrast, Use Case No. 6 compares the
numerical outputs with an absolute standard (faUure analysis).

Use Case No. 7: Improved Demand Modeling

This use case applies HCM methods to develop volume-reliability curves by
facility type for use in a demand modeling environment to estimate reliability
and to irnprove destinatian, time-of-day, rnode choice, and route choice models.

USE OF ALTERNATIVE TOOLS

In sorne cases, a finer temporal sensitivity to dynamic changes in thc system
will be required far a reliability analysis than can be provided by the typical 15-
min analysis periad used by HCM methads. This situation may occur in
evaluating traffic-responsivc signa! timing. traffic adaptive contro!, dynamic
ramp metering. dynamic congcstion pricing. or mcasures affecting the
prevalence or duration of incidents with less than \O-min durations. There may
also be scenarios and configurations that the HCM cannot address, such as cyclic
spillback or adaptive signal control.
Far such situations, this chapter's conceptual framev.,'orkfor evaluating
travel time reliability can be applied to altcrnativc analysis tools. The sarne
conceptual approach-generating scenarios, assigning scenario probabilities,
evaluating scenario performance, and summarizing the results-applies when
aIternative analysis tools, such as microsimulation, are used to estimate the
reliability cffects of opcrations improvements.
Befare embarking on the use of alternative to01sfor rcliability analysis, the
analyst should consider the much greater analytical demands imposed by a
reliability analysis following this chapter's conceptual anaIysis framework.
Thousands of scenarios may need to be analyzed with the aIternative to01in
addition to the number of replications per scenario required by the tool itself to
establish average conditions. Extracting and summarizing the results from
numerous applications oE the altcrnative tool may be a significant task.
If a microscopic slmulabon tool is used, sorne portions oE this chapter's
analysis framework that were fit to the HCM's lS-min analysis periods and
tailored to the HCM's speed-flow curves will no longcr be needed:
• Sccnarios may be defined differently from and may be oE longcr or
shorter duration than thosc used in HCM analysis.

O1apter 17jUrban 5treet Reliability and ATDM Applications

VersíOfl6.0 Page 17-43
Highway Capacity Manual: A Guide tor Multimodal Mobility Analysis

• Incident start times and durations will no longer need to be rounded to

the nearest 1S-min analysis periodo
• Weather start times and durations will no longer need to be rounded to
the nearest 1S-min analysis periodo
• Demand wiII no longer need to be held constant for the duration of the
1S-min analysis periodo
• The peak hour factors used to identify the peak lS-min flow rate within
the hour will no longer be applied. They will be replaced with the
analysis tool's built-in randomization process .
• The reliability methodology's recommended saturation flow rate and
free-flow speed adjustments for weather events and incidents will have to
be converted by the analyst to the microsimulation model equivalents:
desired speed distribution and desired headway distribution.
• Acceleration and deceleration rates will also be affected for sorne weather
events.
lf a less disaggregate tool is used (e.g., mesoscopic simulation analysis tool.
dynamic traffie assignment tool, demand forecasting tool), many of this chapter's
adaptations of the conceptual analysis framework to the HCM may still be
appropriate or may need to be aggregated further. Thc analyst should consult
the appropriatc tool documentation and determine what further adaptations of
the conceptual analysis framework might be required to apply the altemative
tool to reliability analysis.

Applications Olapter 17fUrban 5treet Reliability and ATDM

Page 17-44 Version 6.0
 Highway Capacity Manual: A Guide fer Multimedal Mebility Analysis

6. REFERENCES

1. Zegeer, L J. Bonneson, R Dowling, r. Ryus, M. Vandchcy, W. Kittclson, N. Sorne of these references can
be found in ttle Technicaf
Rouphail, B. Schroeder, A. Hajbabaie, B. Aghdashi, T. Chase, S. Sajjadi, R Reference Library in VoIume 4.
Margiotta, and L. Elefteriadou. lllcorporatillg Travel Time Reliability illto the
Highway Capacity Manual. SHRP 2 Rcport S2-L08-RW-1. Transportation
Research Board of the National Academies, Washington, D.C., 2014.
2. Dowling, R. G., and A. Elias. NCHRP SYlltilesis 447: Acti"e Traffic Mmwgemerzt
for Arterials. Transportation Research Board of thc National Acadcmies,
Washington, D.C., 2013.
3. Federal Highway Administration. Active Transportation and Demand
Managcmcnt websi te. http://www.ops.fhwa.dot.gov/atdm/about/overview.htm.
Accessed Nov. 7, 2014.
4. Stevanovic, A. NCHRP Syrzthesis 403: Adaptive Traffic COlltrol Systems:
Domestk alld Forcign State oJ Practice. Transportation Research Board of the
National Aeademies, Washington, D.C., 2010.
5. Urbanik, T., A. Tanaka, B. Lozner, E. Lindstrom, K. Lee, S. Quayle, S. Beaird,
S. Tsoi, P. Ryus, D. Gettman, S. Sunkari, K. Balke, and D. BuJlock. NCHRP
Report 812: Signal Timing Manllal, 2nd ed. Transportation Research Board of
the National Academies, Washington, D.C., 2015.
6. Margiotta, R, T. Lomax, M. Hallenbeck, R Dowling, A. Skabardonis, and S.
Tumer. Analytical Procedllres Jor Determillillg the lmpacts of ReIiability
Mitigation 5trategies. SHRP 2 Report S2-L03-RR-1. Transportation Research
Board of the National Academies, Washington, D.C., 2013.
7. Habib, P. A. Transportation System Management Options for Downtown
Curbside Pickup and Delivery ol Freight. In Trmtsportatioll Research Record
758, Transportation Research Board, National Research CounciL
Washington, D.C., 1980, pp. 63-69.
8. A Po/iey 011 Geometric Design of Highways and Streets. American Association of
State Highway and Transportation OHicials, Washington, D.C., 2011.
9. Hallenbcek, M., M. Rice, B. Smith, C. Cornel1.Martinez, and J. Wilkinson.
Vehicle Vollllne Distrihtltiolls by Classificatioll. Report FHWA.PL.97.025.
Chaparral Systcms Corporation, Santa Fe, N.M., 1997.
10. Maki, P. Adverse Weather Traffic Signal Timing. Presented at 69th Annual
Meeting of the Institute of Transportation Enginccrs, Las Vegas, Nev., 1999.
11. Comparative Climatk Data for tile Ullited States Through 2010. National Climatic
Data Center, National Oceanic and Atmospheric Administration, Ashevillc,
N.C., 2011. http://www.ncdc.noaa.gov. Accessed Sept. 21, 2011.
12. RaillfaIl FrequeJlCYAtlas of the U.S.: RainfaIl Evellt Statistks. National Climatic
Data Center, National Oceanic and Atmospheric Administration, Asheville,
N .c., 2011. http://www .ncdc.noaa.gov/oa/documentlibrary/rainfall.html.
Aeccssed Sept. 21, 2011.
13. HigIJway Safety Manllal, 1st ed. American Association of Statc Highway and
Transportation Officials, Washington, D.C., 2010.

Chapter 17¡Urban 5treet Reliability and ATDM References

Version 6.0 Page 17-45
 Highway Capacity Manual: A Guide for Multimodal MobWty Analysis

CHAPTER 18
URBAN STREET SEGMENTS

CONTENTS

1. INTRO DUCTI ON .•...•...•...•......••..•...•...•...•...•...........••..••...•...•...•...•...•...•...•..••...•...

18.1
Overview 18-1
Chapter Organization 18-1
Related HCM Content 18.2

2. CON C EPTS 18.4

Analysis Type 18-4
Urban Street Segment Defined 18.5
LOS Critcria 18-6
Scope oí the Methodologies 18-8

3. MOTORIZED VEHIClE METHODOlOGY •...................•...•...•............•.......18.10

Scope oí the Methodology 18-10
Required Data and Sources 18.14
Overview oí the Methodology 18-22
Computational Steps 18-24

4. PEDESTRIAN METHODOlOGY ................................••...•...•...•....•......•....•...•

18-41
Scope oí the Mcthodology 18-41
Required Data and Sources 18.43
Overvicw oí thc Methodology 18-48
Computational Steps 18-48

5. BICYClE METHODOlOGY ..•........•......•...••...•...•...•....•......••...•...•...•...............

18-57
Scope oí the Mcthodology 18-57
Rcquired Data and Sources 18-58
Overview oí the Methodology 18-61
Computational Stcps 18.62

6. TRANSIT METHODOlOGY 18.66

Scope oí the Methodology 18-66
Required Data and Sources 18-67
Overvicw oí the Methodology 18-71
Computational 5teps 18-71

Qlapter IB/Urban Street 5egments Contents

Version 6.0 Pa<]e 18-i
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

7. APPLI CA TIO NS •.•.•.•••.•..•.•...••.••..•.•.•••.••••.•••.•••..••..•••.•••..••..•••.•••..••.••..••..••.••..•..•..

18-79
Examp1e Prob1ems 18~79
Generalized Dail y Service Volumes 18-79
Anal ysis Type 18.79
Use of Altemative Tools 18.80

8. REFERENCES •••..••..••..•••..••..••..•••..•••.••..•••..••..••..•••..•••.••..•••..••..••..••••••..••.•••.••..••.•••
18-83

eoolent> Chapter 18¡Urban 5treet Segments

Page 18-1i Version 6.0
 Highway Capacity Manual: A Guide for Mu/timodal Mobility Analysis

LIST OF EXHIBITS

Exhibit 18-1 LOS Cntena: Motorized Yehicle Mode 18-7

Exhibit 18-2 LOS Criteria: Pedestrian Mode 18-8
Exhibit 18-3 LOS Criteria: Bicycle and Transit Modes 18-8
Exhibit 18-4 Three Alternative Study Approaches 18.11
Exhibit 18-5 Required Input Data, PotentiaI Data Sources, and Default
Yalues for Motorized Yehicle Analysis 18-15
Exhibit 18-6 Default Turn Propartions far Access Point Intersections 18.16
Exhibit 18.7 Default Access Point Density YaIues 18-19
Exhibit 18-8 Motorized Yehicle MethodoIogy for Drban Street Segments 18.24
Exhibit 18-9 Entry and Exit Yolume on Examplc Segment 18-26
Exhibit 18.10 Example Origin-Destination Distribution Matrix 18-26
Exhibit 18-11 Base Frec-Flow Spced Adjustmcnt Factors 18-28
Exhibit 18-12 Speed-Flow Relationship foe Drban Street Segments 18.30
Exhibit 18-13 Delay due to Turning Yehicles 18-31
Exhibit ]8-14 Use of an Arrival Flow Profile to Estimate the Yolume
Arriving During Green 18-33
Exhibit ]8-15 Qualitative Desceiption of Pedestrian Space 18.43
Exhibit ]8-16 Required Input Data, PotentiaI Data Sources, and DefauIt
Yalues for Pcdcstrian Analysis 18-44
Exhibit 18.17 Pedestrian Methodology for Drban Street Segments 18.48
Exhibit 18-18 Width Adjustments for Fixed Objccts 18-SO
Exhibit 18-19 Yariables far Pedestrian LOS Score for Link 18-53
Exhibit 18-20 Diversion Distance Components 18-55
Exhibit 18-21 Rcquired Input Data, Potential Data Sourees, and Dcfault
Yalues tor Bicyc1e Analysis 18-59
Exhibit 18.22 Pavement Condition Rating 18-60
Exhibit 18-23 Bicycle Methodology toe Drban Street Segmcnts 18-61
Exhibit 18.24 Variables foe Bicycle LOS Seore far Link 18-64
Exhibit 18-25 Required Input Data, Potential Data Sourees, and Default
Yalucs for Transit Analysis 18-67
Exhibit 18-26 Transit Methodology tor Drban Street Segments 18-71
Exhibit 18-27 Transit Yehicle Running Time Loss 18-75

Olapter 18/Urban Street 5egments Contents

Version 6.0 Page IB-iii
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

1. INTROOUCTION

OVERVIEW VOlUME 3: INTERRUPTED FLOW

16. Urban Street Fadlities
This chapter describes methodologies for evaluating the operation of each of 17. Urban Street Reliability and
ATOM
the following urban street traveI modes: motorized vehicle, pedestrian, bicyclc, 18. Urban Street Segments
and transit. Each methodology is used to evaluate the quality of service provided 19. Signalized Intefsections
20. TWSC Intersections
to road users traveling along an urban street segment. A detailed description of 21. AWSC Intersections
each travel mode is provided in Chapter 2, Applications. 22. Roundaoouts
23. Interchaoge Ramp Terminals
The methodologies are much more than just a means of cvaluating quality of and Altemative Intersections
24. Off.Street ?edestrian and
service. They inelude an array of performance measures that fully describe
Bicycle Facilities
segment operation. These measures serve as clues in identifying operational
issues and provide insight into the development of effective improvement
strategies. Thc analyst is encouraged to consider the full range of mcasures
associated with each methodology.
This chapter describes methodologies for evaluating urban street segment
performance from the perspective of motorists, pedestrians, bicyelists, and transit
riders. The methodologies are referred to as the motorized vehicle methodology,
the pedestrian methodology, the bicyele methodology, and the transit
rnethodology. Collectivcly, thc methodologies can be used to evaluate an urban
street segmcnt operation from a multirnodal perspective.
Each rnethodology in this chapter is focused on the evaluation of a street
segrnent (with consideration given to the intersections that bound it). The
aggregation of scgment performance rneasures to obtain an estimate oí facility
performance is described in Chapter 16, Drban Street Facilities. Methodologies
for evaluating the intersections along the urban street are describcd in Chapters
19 to 23.
A street segrnent's performance is described by the use of one or more
quantitative measures that characterize sorne aspect of the service provided to a
specific road-user group. Performance measurcs citcd in this chapter inelude
motorized vehiele travel speed, motorized vchiele stop rate, automobilc travelcr
perception score, pedestrian travel spccd, pedestrian space, pcdestrian lcvcl-of-
service (LOS)score, bicyele travel speed, bicycle LOS score, transit vchiele travel
sJX>ed,transit wait-ride senre, and transit passenger LOS score.
The four methodologies described in this chapter are based largely on the
products of two National Cooperative Highway Research Program (NCHRP)
projects (1, 2). Contributions from other research are rcferenced in the relevant
sections.

CHAPTER ORGANIZATlON
Section 2 of this chapter presents concepts used to describe urban street
operation. lt ineludes guidance for establishing the segment analysis boundaries
and the analysis period duration and describes how an urban street segment is
defined for the purpose of this chapter. lt concludes with a discussion of the
service measures and LOS thresholds used in the methodology.

Chapter lB/Urnan Street 5egments Introduction

Version 6.0 Page 18-1
Highway Capacity Manual: A Guide for Mulfimodal Mobility Analysis

$ection 3 presents the methodology for evaluating motorized vehide service

along an urban street segment. It indudes a description of the scope of the
methodology and its required input data. It condudes with a description of the
computational steps that are followed for each application of the methodology.
Section 4 presents the methodology for evaluating pedestrian service along
an urban stn.."Ct
segment. It indudes a discussion of methodology scope, input
data, and computational steps.
SecHon5 presents the methodology for evaluating bicyele service along an
urban strcet segment. lt ineludes a discussion oC methodology scope, input data,
and computational steps.
$ection 6 presents the methodology for evaluating transit rider service along
an urban street segmento It ineludes a discussion of methodology scope, input
data, and computational steps.
Section 7 presents guidance on using the results of the segment evaluation. lt
indudes example rcsuIts from each methodology and a discussion of situations
in which alternative evaluation tools may be appropriate.

RELATED HCM CONTENT

Other Highway Capacity Manual (HCM) content related to this chapter
indudes the following:
• Chapter 16, Urban Street Facilities, which describes concepts and
methodologies for the evaluation of an urban street facility;
• Chapter 17, Urban Street Reliability and ATDM, which provides a
mcthodology for evaluating travel time reliability and guidance for using
this methodology to evaluate aIternative active traffic and demand
management (ATDM) strategies;
• Chapter 19, Signalized lntersections, which provides methods for
evaluating pedestrian and bieyc1eLOS at intersections, the results of
which are used in this chapter's facility-Ievel pedcstrian and bicyde
methods;
• Chapter 29, Urhan Strcet Facilities: Supplemental, which provides details
of the reliability methodology, a procedure for sustained spillback
analysis, information about the use of alternative evaluation tools, and
example problems demonstrating both thc urban street facility
methodologies and the reJiability methodology;
• Chapter 30, Urban Street Segments: Supplemental, which describes
procedures for predicting platoon flow, spillback, and delay due to turns
from the majar street; a planning.level analysis application; and example
problems demonstrating the urban street segment methodologies;
• Chapter 31, Signalized Intersections: Supplemental, which describes
procedures for predicting actuated phase duration; lane volume
distribution; saturation f10wadjustment factors for pedestrian, bicyde,
and work zone presence; and queue length; and presents a planning-level

Introdoction Chapter 18fUrban 5treet 5egments

Page 18-2 VeISion 6.0
 Highway Capacity Manual: A Guide tor Multimodal Mobility Analysis

analysis application, as well as example problems demonstrating thc

signalizcd intersection methodologies;
• Case Study 3, Krome Avcnue, in the HCM Applications Guide in Volume
4, which demonstrates the application of HCM methods to the evaluation
of a real-world urban strect; and
• Section K, Drban Streets, in Part 2 of the Planning and Preliminary
El1gil1eeringApplícafiolls Guide fa the HCM, which describes how to
incorporate this chapter's methods and performance measures into a
planning or preliminary engineering eHort.
A procedure for determining free-flow speed when a work zone is prescnt
along the scgment is provided in the final report for NCHRP Project 03-107,
Work Zone Capacity Methods for the HCM. This report is in the Technical
Reference Library in online Volume 4.
Methodologies for quantifying the performance of a downstream boundary
intersection are described in Chapters 19 to 23.

Chapter 18/Urban Street 5egments Introduction

Version 6.0 Page 18-3
 Highway Capadty Manual: A Guide tor Multimodal Mobi/ity Analysis

2. CONCEPTS

This section presents concepts uscd to describe urban street operation. The
first subsection assists the analyst in determining thc type oí analysis to be
conducted and indudes guidance íor estabJishing the segment analysis
boundaries and the analysis period duration. The second describes how an urban
strcet segment is defined in terms oí points and Iinks. The third discusses the
service measures and lOS thresholds used in the mcthodology. The last
identifies the scope oí the collcctive set oí methodologics.

ANALYSIS TYPE
The phrase analysis type is used to describe the purpose íor which a
methodology is used. Each purpose is associated with a different level oí detail,
sincc it relates to the precision of the input data, the number of dcfault values
used, and the desircd accuracy of the results. Three analysis types are recognizcd
in this chapter:
• Operational,
• Dcsign, and
Spatiaf aOO Temporal Umirs • Planning and preliminary engineering.
These analysis types are discussed in more dctail in Chapter 2, Applications.

Analysis Boundaries
The segment analysis boundary is defined by the roadway right-of-way and
the opcrational influence area of each boundary intersection. The influence area
of a boundary intcrscction extends upstream from the intersection on each
~
./ o
t intersection leg. lt indudes al! geomctric fcatures and traffic conditions that
~ o

"
influence segment or intersection operation during the study periodo For these
"'~
~ Ó
N
o
reasans, the analysis boundaries should be established for each segmcnt and
~
~~ intersection on the basis of the conditions present during the study periodo

Travel Directions to Be Evaluated

Previous editions of the HCM have allowed the evaluation of one direction
of travel along a scgment (even when it sefVed two-way traffic). That approach is
retained in this edition for thc analysis of bicyele and transit performance. For
the analysis oí pedestrian performance, this approach translates into the
evaluation of sidewalk and street conditions on one side of thc scgmcnt.
For the motorized vehide For the analysis of motorized vehiele performance, an analysis of only one
methodology, a segment
evaluation considers both travel direction (when the street serves two-way traffic) docs not adequately
directions of travef (when the recognize the interactions between vehieles at the boundary intcrsections and
street serves two-way traffic).
thcir influence on segment operation. For examplc, thc motorizcd vchiclc
methodology in this edition oí the HCM explicitly rnodels the platoon formed by
the signal at one end oí the segment and its influencc on thc opcration of the
signal at the other end oí the segmento For this rcason, evaluation oí both travel
directions on a two-way segment is important.

Concepts Olapter IB/Urban Street 5egments

Page lB-4 Version 6.0
 Highway Capacity Manual: A Guide tor Multimodal Mobility Analysis

5tudy Period and Analysis Period

The study period is the time interval represented by the performance
evaluation. It eonsists of one or more consecutive analysis periods. An analysis
period is the time interval evaluated by a single application of the methodology.
The methodology is based on the assumption that traffie conditions are
steady during the analysis period (i.e.}systematic change over time is negligible).
For this reason, the duration of the analysis period is in the range of 0.25 to 1 h.
The longer durations in this range are sometimes uscd for planning analyses. In
general. the analyst should use caution with analysis periods that exceed 1 h
because traffic conditions are not typicalIy steady for long time periods and
because the adverse impact of short peaks in traffic demand may not be detected
in the evaluation.

URDAN STREET SEGMENT DEFINE O

Terminology
For the purpose of analysis, the roadway is separated into individual
elements that are physicalIy adjacent and operate as a single entity in serving
travelers. Two elements are commonly found on an urban street system: points
and links. A poin/ is the boundary betwecn links and is represented by an
intersection or ramp terminal. A link is a length of roadway between two points.
A link and its boundary points are referred to as a sexment.

Points and Segments

The link and its boundary points must be evaluated together to provide an A segment performance
aecurate indication of overall segmcnt performance. For a given direction of measure combines link
performance and point
travel along the segment, link and downstream point performance measures are performance.
eombined to determine overall segment performance.
If the subjeet segment is within a coordinated signal system, the following
rules apply when the segment boundaries are identified:
• A signalized interscction (or ramp terminal) is always used to define a
segment boundary.
• Dnly intersections (or ramp terminals) at which the segment through
p p
movement is uncontrolled (e.g., a two-way STOP-controlledinterscction)
can exist along the segment between the boundaries.
If the subjcct segment is not within a coordinated signal system, the
following rules apply when the segment boundaries are identified: -•
o
E
• An intersection (or ramp terminal) having a type of control that can .!'
impose on the segment through movement a legal requirement to stop or
yield must always be used to define a scgment boundary.
• An intersection (or ramp terminal) at which the segment through
movement is uncontrolled (e.g., a two.way STOP-controlledintersection)
may be used to define a segment boundary, but it is typicalIy not done.
A midsegment lraffie control signal provided for the exclusive use of - segmeot perl. measu.e
pedestrians should not be used to define a scgment boundary. This reslriction I P • point perf. measure

Chapter 18¡Urban 5treet 5egments eoo<epts

VetSbn 6.0 Page 18-5
Highway Capacity Manual: A Guide for Mulfimodal Mobility Analysis

refleets the faet that the methodologies described here were derived for and
calibrated with data from street segments bounded by an intersection.
An access point intersection is an unsignalized intersection with one or two
access point approaches to the segmento The approach can be a driveway or a
publie street. The through movements 00 the segment are uncontrolled at an
access point intersection.

LOS CRITERIA
This subsection describes the LOS crHeria for the motorized vehicle,
pedestrian, bicycle, and transH modes. The criteria for the motorized vehicle
mode are different from the crHeria used for the other modes. Specifieal1y, the
crHeria for the motorized vchicle mode are based on performance measures that
are field.measurabJe and pereeivable by travelers. With one exception, the
criteria for the pedestrian and bicycle modes are based on scores reported by
travelers indicating their perception of service quality. The exception is the
pedestrian space measure (used with the pedestrian mode), which is field-
measurable and perceivable by pedestrians. The criteria for the transH mode are
based on measufl_>dchanges in transit patronage due to changes in service quality.

AIf uses 01 the word "voIume' Motorized Vehicle Mode

or the phrase "voIume-to-
Two performance measures are used to characterize vehicular LOS for a
GJpacity ratio' in this chapter
refer lo demand voIume or given direction of travel along an urban street segmento One measure is travel
demand-voIume-to-capacity speed for through vehicles. This spced refleets the factors that influence running
ratio.
time along the link and the delay incurred by through vehicles at the boundary
interseetion. The second measure is the volume.to-capadty ratio for the through
movement at the downstream boundary interseetion. These performance
measures indicate the degree of mobility provided by the segment. The following
paragraphs characterize each service leve!'
"Free-flow speetr is the • LOS A describes primarily free-flow operation. Vehicles are completely
average running speed of
through vehides traveling unimpeded in their ability to maneuver wHhin the traffic stream. Control delay
along a segment under Iow. at the boundary intersection is minimal. The travel speed exceeds 80% of the base
voIume conditions and not
delayed by traffic control free-flow speed, and the volume-to-capacity ratio is no greater than 1.0.
devices or other vehides.
LOS B describes reasonably unimpeded operation. The ability to maneuver
The "base ~ffow spee,r is
defined lo be the free-ffow within the traffie stream is only slightly restricted, and control deJay at the
speed on longer segments. boundary intersection is not signifieant. The travel speed is between 67% and
80% of the base free-flow speed, and the volume-to-eapacity ratio is no greater
than 1.0.
LOS e describes stable operation. The ability to maneuver and change lanes
at midsegment locations may be more restricted than at LOS B. Longer queues at
the boundary intersection may contribute to lower travel speeds. The travel
speed is betwcen 50% and 67% of the base free-flow speed, and the volume-to-
capacity ratio is no grcater than 1.0.
LOS O indicatcs a less stable eondition in which small increases in £Iow may
cause substantial increases in delay and decreases in travel speed. This operation
may be due to adverse signal progression, high volume, or inappropriate signal

Concepts Chapter 18/Urban 5treel 5egments

Page 18.6 Vet5ion 6.0
 Highway Capacity Manual: A Guide tor Multimodal Mobility Analysis

timing at the boundary intersection. The travel speed is between 40% and 50% of
the base free.flow speed, and the volurne.to.capadty ratio is no greater than 1.0.
LOS E is characterized by unstable operation and significant delay. Such
operations may be due to sorne combination of adverse progression, high
volume, and inappropriate signa! timing at the boundary intersection. The travel
spt'ed is between 30% and 40% of the base free-flow speed, and the volume.to.
capacity ratio is no greater than 1.0.
LOS F is characterized by flow at extremel)' low spl.'Cd.Congestion is likely
occurring at the boundary intersection, as indicated by high dela)' and extensive
queuing. The travel spl.'Cdis 30% or less of the base free.flow speed, or the
volume-to-capacity ratio is greater than 1.0.
Exhibit 18.11ists the LOS thresholds established for the motorized vehicle
mode on urban streets. The threshold value is interpolated when the base free.
flow speed is between the va!ucs shown in the column headings of this exhibit.
For example, the LOS A threshold for a segment with a base free-flow speed of
G~~~M~~~~-~~-~.~-~+~
Ttavel Speed IhreshQld by Base Free.Flow Speed (milh) Volume.to- Exhibit 18-1
LOS 55 50 45 40 35 30 25 CaDacitv Ratio. LOSeriteria:Motorized
A >44 >40 >36 >32 >28 >24 >20 :'S 1.0 Vehícle Mode
B >37 >34 >30 >27 >23 >20 >17
e >28 >25 >23 >20 >18 >15 >13
o >22 >20 >18 >16 >14 >12 >10
E >17 >15 >14 >12 >11 >9 >8
F :'S17 <15 :'S14 :S12 :S11 <9 <8
F AA > 1.0
Note: • Volume.to""Vlpodty ratio of through movement at oownstream ooundary intersection.

Pedestrian, Bicycle, and Transit Modes

Historically, this manual has used a single performance measure as the basis
for defining LOS. However, research ducumented in Chapter 5, Quality and
Level.of.$crvice Concepts, indica tes that travelers consider a wide vnriety of
factors in assessing the quality of service provided to them. Some of these factors
can be described as performance measures (e.g., spt."'Cd), and others can be
describcd as basic descriptors of the urban street character (e.g., sidewalk width).
The methodologies for evaluating the pedestrian, bicycle, and transit modes
combine these factors lo determine Ihe corresponding rnode's LOS.
Pedestrian quality of service can be evaluated for the segment, the link, or The Spatfal Limn-s subsections
of Sections 4 and 5 provide
both. A segment-based pedestrian evaluation uses the worse of the LOS letters guidance on wIlen fa use
resulting from pedestrian space and the segment pedestrian LOS score to segment- and link.base(!
analyses for the per:festrian and
determine the overall segment pedestrian LOS. The !eft side oí Exhibit 18.2 Iists bicycJe modes, respectively.
the threshold values associated with each LOS fur the scgment.based evaluation
of the pedestrian travcl mode. The LOS is determined by consideration of both
the LOS score and the average pedestrian space on the sidewalk. The applicable
LOS for an evaluation is determined from the table by finding the intersection of
the row corresponding to the computed score value and the column
corresponding to the computed space value.

Chapter 18jUrban 5treet 5egments Concepts

V~6.0 Page 18-7
Highway Capacity Manual: A Guide for Multimodal Mobi/ity Analysis

Exhibit 18-2 segment- 5egment-Based lOS bV Unk-Based

LOSCriteria: Pedestrian Mode
oa",. Average Pedestrian 5Dace etr/o) Pedestrian LOS
Pedestrian >40- >24- >15- >8.0- link-Based
lOS 5core >6. 6. 4. 24 15' <8.0. lOS 5core LOS
$2.00 A B e D E F <1.50 A
>2.00-2.75 B B I e D E F >1.50-2.50 B
>2.75-3.50 e e e D E F >2.50-3.50 e
>3.50-4.25 D D D D E F >3.50-4.50 D
>4.25-5.00 E E E E E F >4.50-5.50 E
>5.00 F F F F F F >5.50 F
Note: "In O'OSS-fIowsituabons, the LOS E1Fthreshold is 13 ft'!p. Chapter 4 describes the concept of"ooss fIow"
¡lIld sitwtions where it should be coosidered.

A link.based pedestrian evaluation uses the link pedestrian scare to determine

the overalllink pedestrian LOS. The right side of Exhibit 18-2lists the threshold
values associated with each LOS for the Iink-based evaluation of the pedestrian
travel mode. The LOS is determined by consideration of onIy the LOS seore.
Exhibit 18-3lists the range of scores that are assodated with each LOS far the
bicyde and transit modes. Similar to the pedestrian mode, bicyde LOS can be
evaluated far the link, the segment, or both. Transit LOS is only evaiuatcd for the
segment.

Exhibit 18-3 5egment-Based link-Based

LOS Oitelia: 6icyde and LOS Bicycle lOS 5core Bieycle lOS 5core Transit LOS 5core
Transit Modes A :;;2.00 $1.50 $2.00
B >2.00-2.75 >1.50-2.50 >2.00-2.75
e >2.75-3.50 >2.50-3.50 >2.75-3.50
D >3.50-4.25 >3.50-4.50 >3.50-4.25
E >4.25-5.00 >4.50-5.50 >4.25-5.00
F >5.00 >5.50 >5.00

The association between LOS score and LOS is based on traveier perception
research. Travelers were asked to rate the quality of service assodated with a
specific trip along an urban street. The letter A was used to represent the best
quality of service, and the letter F was used to represent the worst quality of
servire. "Best" and "worst" were ieft undefined, allowing the respondents to
identify the best and worst conditions on the basis of their traveling experience
and perception of service quality.

SCOPE OF THE METHODOLOGIES

This subsection identifies the conditions for which each methodoiogy is
applicable.
• Boulldary intersectiol1s. AH methodologies can be used to evaluate segment
performance with signalized ar two-way Slúr-controlled houndary
intersections. In the latter case, the cross street is STOP controlled. The
motarized vehide methodoIogy can al50 be used to evaIuate performance
with all-way SlúP- or YIELD-control1ed(e.g., roundabout) boundary
intersections.
• Street types. The four methodologies were deveIoped with a focus on
arterial and collector street condition5. If a methodoIogy is used to

Concepts Chapter 18/Urban Street 5e9ments

Page 18-8 Version 6.0
 Highway Capacify Manual: A Guide tor Mulfimodal Mobilify Analysis

evaluate a local street, the performance estima tes should be carefully

reviewcd for accuracy.
• Flow conditiolls.The four methodologies are based on the analysis of
steady traffie conditions and are not well suited to the evaluation of
unsteady eonditions (e.g., congestion, cyclic spillback, signal preemption).
• Target road usas. Colleetively, the four methodologies were developed to
estímate thc LOS perceived by motorized vchicle drivers, pedestrians,
bicyclists, and transit passengers. They were not devclopcd to provide an
estimate of the LOS perceived by other road users (e.g., commerdal
vehíde drivers, automobile passengers, delivery truek drivers, or
reereational vehide drivers). However, it is likcly that the perecptions oí
these other raad users are reasonably well represented by the raad users
for whom the methodologies were developed .
• J"jluellces righl+of-way. A raad uscr's perception oí quality of service
in fhe
is influeneed by many factors inside and outside of thc urban street right-
of-way. However, the rnethodologies in this chapter were spedfically
eonstructcd to exclude factors that are outside oí the right-of-way (e.g.,
buildings, parking lots, scenery, landscaped yards) that rnight influenee a
traveler's perspective. Th.is appraaeh was followed beeause factors
outside oí the right-of-way are not under the direct contral of thc agency
operating the street.

Chapter 18¡Urban Street 5egments Concepts

Version 6.0 Page 18-9
Highway Capacity Manual: A Guide for Mu/timoda/ Mobility Analysis

3. MOTORIZEO VEHICLE METHODOLOGY

This section describes the methodology for evaluating the capacity and
quality of service provided to motorizcd vehicles on an urban street segment.
Extcnsions to this methodology for evaluating more complex urban street
operational elements are described in Chapter 30, Drban Street Segments:
Supplemental.

SCOPE OF THE METHODOLOGY

The overall scope of the four methodologies was provided in Section 2. This
section identifies the additional conditions for which the motorized vehicle
methodology is applicable.
• Target travel »lode. The motorized vehide methodology addresses mixed
automobile, motoreycle, truck, and transit traffic streams in which the
automobile represents the largest percentage of all vehicles. The
methodology is not designed to evaluate the pcrformance of other types
of vehiclcs (e.g., golf carts, motorized bicycles).
• Mobility The motorized vehicle methodology is intended to facilita te
/OCIIS.
the evaluation of mobility. Accessibility to adjacent properties by way of
motorized vehicle is not directly cvaluated with this rnethodology.
Regardless, a segment's accessibility should also be considercd in
evaluating its performance, especially if the segment is intended to
provide such access. Oftentimes, factors that favor mobility reflect
minimalleveIs of access and vice versa.

Spatial and Temporal Limits

Analysis Boundaries
An analysis of only one travel direction (when the street serves two-way
traffic) does not adequately recognizc the interactions between vehicles at the
boundary intersections and their influence on segment operation. For this reason,
evaluation of both travel dircctions on a two-way segment is important.
The analysis boundary for each boundary intersection is defined by the
operational influence area of the intcrsection. It should inelude the most distant
extent of any intersection-related queue expected to occur during the study
periodo For thcse rcasons, the influence area for a signalized intersection is likely
to extend at least 250 ft back from the stop line on each intersection leg.

Study Period and Analysis Period

The concepts of study period and analysis period are defined in Section 2 in
general terms. They are defined more predse1y in this subsection as they relate to
the motorizcd vehicle mcthodology.
Exhibit 18-4 demonstratcs three altcrnative approaches an analyst might use
for a given evaluation. Other aItematives exist, and the study penod can exceed
1 h. Approach A is the approach that has traditionally beeo used and, unlcss
otherwisc justified, is the approach that is recommendcd for use.

Motorized Vehicle Methodology Olapter 18jUrban 5treet 5egments

Page 18-10 Version 6.0
 Hjghway Capadty Manual: A Guide for Multimodal Mobility Analysis

APProach A Aporoacb B Aoproach e Exhibit 18-4

Three Alternative 5tudy
Study Period " 1.0 h Study Period " 1.0 h Study Period " 1.0 h Approaches

Single analysis period Single analysis period Multiple analysis peroos

T,,0.25h T=LOh T=O.2Sh

Time

~ - analysis periOd

Approach A is based on evaluation of the peak 15-min period during the rhe use of peak 15-min traffic
multiplfed by tour is preferred
study periodo The analysis period T is 0.25 h. The equivalent hourly fIow rate in for existing conditiorls when
vehic1es per hour (veh/h) used for the analysis is based on either (a) a peak 15- traffic cvunts are available. !he
use of a l-h demand voIume
min traffic count multiplied by faur or (b) a l-h demand valume dividcd by the divided by a peak hour factor
peak hour factor. The former option is preferred for existing conditions when is preferred when volumes are
projected or when hourfy
traffic caunts are available; the laUer option is preferred when hourly volumes projeded voIumes have been
are projected or when hourly projected volumes are added to existing volumes. added lo existing vo/umes.

Additional discussion on use of the peak hour factor is provided in the

subsection titled Required Data and Sources.
Approach B is bascd on the evaluation of one l-h analysis period that is
coincident with the study periodo The analysis period T is 1.0 h. The tlow rate
used is equivalent to thc l-h demand volume (Le., the peak hour factor is not
used). This approach implicitly assumes that the arrival rate of vehides is
constant throughout the period of study. Therefore, the effects of peaking within
the hour may not be identified, and the analyst risks underestimating the delay
actually incurred.
Approach e uses a l-h study period and divides it into four 0.25-h analysis
periods. llis approach accounts for systematic tlow rate variation among
analysis periods and for queues that carry over to the next analysis periodo lt
produces a more accorate representation of dcla)'. It is caBed "multiple time
period analysis" and is described in the next subsection.
Regardless of analysis periad duration, a single-period analysis (Le.,
Approach A or B) is typical for planning applications.

Mu/tip/e Time Period Ana/ysis

If the analysis period's demand volume cxceeds capadty, a multiple time
period analysis should be undertaken in which the study period ineludes an
initial analysis periad with no initial queue and a final analysis periad with no
residual qucue. On a movement-by-movement and intersection-by-intersection
basis, the inHial queue for thc sccond and subsequent periods is equal to the

Chapter 18/Urban street 5egments Motorized Vehide Methodology

VersiOn 6.0 Page 1B-11
Highway Capacity Manual: A Guide for Multimodal MobiJjty Analysis

residual queue from the previous periodo This approach provides a more
accurate estimate of the delay associated with thc congestiono
If evaluaban of multiple analysis periods is determined to be important, the
performance estima tes for each period should be separately reported. In this
situation, reporting an average performance for the study period is not
encouraged because it may obscure extreme values and suggest acceptable
operation when in reality sorne analysis periods have unacceptable operation.

5egment Length Considerations

The motorized vehicle methodology descrihed in this section is not
appropriate for the analysis of "short" segments that are bounded by signalized
interse<:tions. In cantrast, the methodology described in Chapter 23, Ramp
Terminals and Alternative lntersections, is appropriate for the analysis of short
segments at signalized interchanges. The analyst may also consider using an
altemative analysis tool that is able to model the operation of c10sely spaced
intersections.
When a segment has a short length, the interaction between traffic movements
and traffic control devices at the two boundary intersections is sufficientIy
complex that the motorized vehicle methodology may not provide an accorate
indication of urban street performance. This complication can occur regardless of
the type of control present at the two boundary intersections; however, the
situation is particularly complicated when the two intersections are signalized.
Demand staryation ocevfS A short segment can experience "eyclic spillback." This spillback occurs
when a portion of the green at
the downstream intersection is when a queue extends back from one intersection into the other intersection (Le.,
not used because the spills back) during a portion of each signal eycle and then subsides. A short
upstream intersection
signaliZation prevents vehides segrnent can also experience "demand starvation." Demand starvation occurs
from readJing the stop line. when a portion of the grecn at the downstream intersection is not used because
the upstrcam intersection signalization prevents vehicles from reaching the stop
line. Demand starvation leads to the ineffieient use of the downstream through
phase and the retention of unserved vehicles on the approaches to the upstream
intersection.
Specific conditions under which a segment bounded by signalized
interscctions should be considered "short" are difficult to define. As a general
rule of thumb, eyclic spiUback and demand starvation are unlikely to occur if the
subject segment exceed.s about 700 fl. They are also unlikely to occur on segments
less than 700 ft provided that the following two conditions hold. First, the major
traffic movement through the segment has coordinated signal timing that
provides very favorable progression. Second, the coordinated traffic movernent
has about the same green-to-cycle-Iength ratio at each signal and each ratio is
about 0.50 or largee. lf the application of these rules to a specific segment indica tes
that cyclic spillback and starvation are unlikely to occur, the methodology
described in this section can be used to evaluate the subject segment.
The mcthodology described in this section is applicable to segments having a
length of 2 mi or less. This restriction is based on the fact that STOp., YIELD.,or
signal-controlled intersections are Iikely to have negligible effect on urban street
operation when segment length exceeds 2 mi. Therefore, if a segment exceeds

Motorized Vehicle Methodology Chapter 18¡Urban 5treet 5egments

Page 18-12 Version 6.0
 Highway Capacify Manual: A Guide for Multimodal Mobnity Analysis

2 mi in length, the analyst should evaluate it as an uninterrupted-flow highway

segment with isolated intersections.

Performance Measures
Performance measures applicable to the motorized vehicle travel mode
include travel specd, stop rate, and automobile traveler perception score. The
latter measure provides an indication of the traveler's perception of service
quality.
LOS is .lIso considered a performance measure. It is useful for describing
segment performance to clected officials, policy makers, administrators, or the
public. LOS is based on travel specd and volume-to-capacity ratio.

Limitations of the Methodology

This subscction identifies the known limitations of the motorized vehicle
methodology. If one or more of these limitations are bclieved to have an
important influence on the performance of a specific street segment, the analyst
should consider using altemative methods or tools for the evaluation.
The rnotorized vehicle methodology does not account for the effect of the
following conditions on urban street operation:
• Delay due to on-street parking maneuvers occurring along the link (see me toIlowlIJg parking-relaled
effects are addressed in !he
margin note for exceptions), methodology: (a) the effect on
saturation ffow rate of parking
• Significant grade along the link, on tf1e approach ro a signalized
mt~amtm)theeffect
• Queuing at the downstream boundary intersection backing up to and en free-fIow speed of parkJi"lg
interfering with the operation of the upstream intersection or an access stail presence along the street.

point intersection on a cyclic basis (e.g., as may occur at sorne interchange

ramp terminals and c10selyspaced intersections),
• Stops incurred by segment through vehicles as a result of a vehicle ahead
tuming from the segment into an acccss point,
• Bicycles sharing a traffic lane with vehicolar traffic, and
• Cross-street congestion or a railroad crossing that blocks through traffic.
In addition, any limitations associated with the methodologies used to
evaluatc the intersections that bound the urban street segment are sharcd with
this methodology. These limitations are listed in Chapters 19 to 23.

Lane Groups and Movement Groups

and mOl'emellt grollp are phrascs used to define combinations of
Lmre grollp
intersection movements for the purpose of cvaluating signalized intersection
operation. These two terms are used extensively in the motorized vehicle
methodology in Chapter 19, Signalized Intersections. They are .lIso used in the
motorized vehicle methodology when the boundary intersection is signalized.
The motorized vehicle methodology in Chapter 19 is designed to evaluate
the performance of designated lanes, groups of lanes, an intersection approach,
and the entire intersection. A lane or group of lanes designatcd for separate
analysis is referred to as a falle grol/p. In general, a separate lane group is

Chapter 18fUrbiln 5treet 5egments Motorized Vehicle Methodology

VetSiOn 6.0 Page 18-13
Highway Capacify Manual: A Guide for Mulfimodol Mobility Anolysis

established for (a) each lane (or combination of adjacent Janes) that exclusively
serves one movement and (b) each lane shared by two or more movements.
,
The concept of nlovemellf groups "isestablished to facilitate data entr)' to the
methodology.ln this regard, input data dcscribing intcrsection traffic are
traditionally spedfic to the movernent (e.g., lefHum movernent volume) and not
specific to the lane (e.g., anal)'sts rarel)' have the volume for a Jane shared by left.
tuming and through vehicles). A sepadte movement group is established for (a)
each tum movement with one or more exclusive tum lanes and (b) the through
movemcnt (inclusive of any tum rnovements that share a lane).

REQUIRED DATA ANO SOURCES

Additionaf required input data, This subsection describes the input data needed for the motorized vehicle
potentiiJf data SOUTreS, and
default vaiues tor the methodology. The required data are listed in Exhibit 18.5. They must be
roundabout segment separately specified for each direction of travel on the segment and for each
methodofogy can be found in
Section 9 of ClIapter JO. boundary intersection. The exhibit al50 lists default values that can be used if
local data are not available (3).
The entries in the Hrst column in Exhibit 18.5 indicatc whether the input data
are needed for a movement group at a boundary intersection, the overall
intersection, or the segmento The input data needed to evaluate the boundary
intersections are identified in the appropriate chapter (Le., Chapters 19 to 23).
The data elements Iisted in Exhibit 18-5 do not include variables that are
considered to represent calibration factors (e.g., acceleration rate). A calibration
factor typically has a relativeJy narrow range of reasonable values or has a small
impact on the accuracy of the performance estimates. The recommended value
for each calibration factor is identiHed at relevant points in the presentation of
the methodology.

Traffic Characteristics Data

This subsection describes the traffic characteristics data listed in Exhibit 18.5.
These data describe the motorized vehicle traffic stream traveling along the street
during the analysis periodo

Demand Flow Rate

The demand flow rate for an intersection traffic movement is defined as the
count of vehicles arriving at the intersection during the analysis period, divided
by the analysis period duration. It is expressed as an hourly flow rate, but it may
represent an analysis period shorter than 1 h. Guidance for estimating this rate is
provided in the chapter that corresponds to the boundary intersection
configuraban (i.c., Charters 19 to 23). The "eount of vehicles" can be obtained
from a variety of sources (e.g., from the Held or as a forccast frarn a planning
model).

Motorized Vehide Methodology Chapter 18/Urban Street Segments

Page 18-14 VetSiOn 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Reguired Data and Units Potential Data Source(s) Suggested Default Value Exhibit 18-5
TrafflC Characteristics Data Required Input Data, Potential
Data Sources, and Default
Demand fIow rate by movement group Field data, past counts Must be provided
Values for Motorized Vehicle
at bOlJndary intersection (veh/h)
Analysis
Access point f10w rate by movement FJeld data, past counts 5ee discussion in text
group (vehjh)
Midsegment flow rate (veh/h) Field data, past counts Estimate by using demand
flow rate at the downstream
boundary int. approach
Geometric Data
Number of lanes by movement group Field data, aerial photo Must be provided
at boundary intersection
Upstream interseetion width (ft) Field data, aerial photo Must be provided
5egment approach tum bay length at Field data, aerial photo Must be provided
boondary lntersection (ft)
Number of midsegment through lanes FJeld data, aerial photo Must be provided
Number of lanes at access points- Field data, aerial photo (a) Number of through lanes
segment approach on approach •• number of
midsegment through lanes.
(b) No right-tum lanes.
(e) It median present, one
left-tum lane per approach;
otherwise, no left-tum lanes.
Number of lanes at access points- Field data, aeria! photo One left-tum and one right-
access point approach tum lane
5egment approach tum bay length at Field data, aerial photo 40% of the access point
access points (ft) spadng, where spacing = 2 x
(5,280) I D~in feet, but not
more than 300 ft nor less
than SO ft
5egment length (ft) FJeld data, aerial photo Must be provided
Restrictive median length (ft) Field data, aerial photo Must be provided
Proportion of segment with curb Field data, aerial photo 1.0 (curb present on both
(decimal) sides of segment)
Number of access point appr03Ches F¡eld data, aerial photo 5ee discussion in text
Proportion of segment with on-street Field data Must be provided
parking (decimal)
0tJuy Data
Analysis period duration (h) 5et by analyst 0.25 h
Speed limit (mi/h) Field data, road inventory Must be provided
Performance Measure Data
Through control delay at boundary HCM methbd output Must be provided
intersection (sjveh)
Through stopped vehicles at boondary HCM method output Must be provided
intersection (veh)
2nd- and 3rd-term back-of-queue size HCM method output Must be provided
for through movement at boundary
intersection (veh/lane)
C3pacity by movement group at HCM method output Must be provided
boundary intersection (veh/h)
Midsegment delay (s/veh) Field data 0,0 sjveh
Midsegment stop:; (stopslveh) Field data 0,0 stopslveh
Notes: lot. '" IntersectiOn,
D. '" ao:.es:spoint (iensJtyen segment (points,lmi).

Access Point Flow Rate

The access point fIow rate is defined as the count of vchides arriving at an
access point intersection during the analysis period, divided by thc analysis
period duration. It is expressed as an hourly flow rate, but it may represent an
analysis period shorter than 1 h. It should represent a demand flow rate,

Chapter 18/Urban 5treet 5egments Motorized Vehicle Methodology

VersiOn 6.0 Page 18-15
Highway Capacity Manual: A Guide for Mulfimodal Mobilify Ana/ysis

This fIow rate is needed for aH movements on each "active" access point
approach and for all major-strect movements at the intersection with one or more
"active" access point approaches. An access point approach is considered to be
active if its volume is suffident to have sorne irnpact on segment operations
during the analysis period. As a rule of thumb, an access point approach is
considered active if it has an entering fIow rate of 10 veh/h or more during the
analysis periodo
If the segment has many access point intersections that are considered
¡nactive but collectively have some impact on traffic fIow, those intersections can
be combined into one equivalent active access point intersection. Each
nonpriority movement at the equivalent access point intersection has a fIow rate
that is equal to the sum of the corresponding nonpriority movement fIow rates of
each of the individual inactive access points.
If a planning analysis is being conductcd in which (a) the projected demand
flow rate coincides with a 1.h period and (b) an analysis of the peak lS-min period
is desired, each movement's hourly demand should be divided by the intersecnon
peak hour factor to predict the fIow rate during the peak lS.rnin periodo The
peak hour factor should be based on local traffic peaking trends. If a local factor
is not available, the default value in Exhibit 19-11 ofChapter 19 can be used.
Default value. The default access point flow rate can be estimated from the
midsegment flow rate by using default toro proportions. These proportions are
shown in Exhibit 18-6 for a typical access point intersection on an arterial street.
The proportion of 0.05 for the left-tum movements can be reduced to 0.01 for a
typical access point on a collector street. These proportions are appropriate for
segments with a typical access point density. They are applicable to access points
serving any public.oriented land use (this excludes single-family residentialland
use and undeveloped property).

Exhibit 18-6 Aa:ess Point

O]V. OC
Default Tum Proportions for
Access Point intersections

Major Street '---0.01 v~

0.05 v~--l
Aow rate '" V.
r-0.05 v~
_ O.OlV~~

"1 (
0.05 Vb 0.01 V~

If one of the movcments shown in Exhibit 18-6 docs not exist at a particular
access point intersection, its volume is not computed (its omission has no eHect
on the proportion used for the other movement fIow rates). The flow rate for the
through movement on an access point approach is nol needed for the motorized
vehicle methodology because this movement is considered to have negligible
effeel on major-street operation.

Motorized Vehicle Methodology Olapter l8jUrban 5treet 5egments

Page 18-16 Vet5ioo 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Midsegment F/ow Rate

The midsegment flow rate is defined as the eount of vehieles traveling along
the segment during the analysis period, divided by the analysis period duration.
It is expressed as an hourly flow rate, but it may represent an analysis period
shorter than 1 h. This volume is spccified separately for each direetion of travel
along the segmento
lf one or more access point intersections exist along the scgment, the Two--Way Vehicufar Trave!

midsegment flow rate should be measured at a loeation betw('('n thesc

intersections (or between an access point and boundary intersection). The
t:::
11

[ocation chosen should be representative in terms of its having a flow rate similar
to other loeations along the segmento If the flO\••.rate is believed to vary
:
.1 If

signifieantIy aIong the scgment, it shouId be measured at severaI Iocations and

:9 - ~
~
an average used in the methodology.
If a pIanning analysis is being conducted in which (a) the projected demand "¡ - t
e
e
o
e
o
¡;
o
flow rate coincides with a l-h period and (b) ao analysis of the peak 15-min N

••.s"
~ e
.s•• B
period is desired, each movement's hourly demand should be divided by the
peak hour factor to predict the flow rate during the peak 15-min periodo The
peak hour factor uscd should be based on local traffic peaking trends. o
E
• ~ i5
i E
•~
b
•
~ ~
Geometric Design Data => j'
This subsection describes the geometric design data listed in Exhibit 18-5. It,
Thesc data describe the geometric elements of the segment or interscctions that l.
are addressed in the motorized vehiele methodology.
~
Number of Lanes at Boundary lntersection
One-Way Vehicufar Trave!
The number of lanes at thc boundary intersection is the couot of lanes that
I

'. !=
are provided for each intersection traffie movement. For a tum movement, this I
count represcnts the lanes reserved for the exclusive use of tuming vehieles. IT
Tum mowment lanes inelude tum lanes that extcnd backward for the length of
the segment and lanes in a tum bayoLanes that are shared by two or more I
,

movements are included in the eouot of through lanes and are described as ~ I

" -
,

shared Imles. If no exclusive tum lanes are provided, the tum movement is e
o e
indieatcd to have O lanes. ¡; o
¡;
~ ~
Upstream lntersection Width ••.s i5

The intersection width applies to the upstream boundary intersection for a E

•
o
~i~
I
given direction of travel and is the effective width of the eross street. On a two- b
way street, it is the distance betw('('n the stop (or yield) line for thc two opposing
•
~ I
=>
segment through movements at the boundary intersection, as measured along L
the centerline of the segment. On a one-way street, it is the distance fmm the stop
line to the far side of the most distant traffic lane on the eross street.

Turn Bay Length at Boundary lntersection

~
I
.
I I
Tum bay length is the length of the bay at the boundary intersection for
which the lanes have full width and in which queued vehieles can be ston'd. Bay
length is measured parallel to the roadway centerline. If there are multiple lanes

Chapter18/Urban5treet5egments Motorized
VehícleMethodology
VetSlCn 6.0 Page18-17
Highway Capacify Manual: A Guide for Mulfimodal Mobility Analysis

in the bay and they have differing lengths, the length entered shouJd be an
average value.
If a two-way left-turn Jane is provided for left-tum vehiele storage and
adjacent access points exist, the bay length entered should represent the effective
storage length available to the left-turn movement. The determination of
effective length is based on consideration of the adjacent access points and their
associated left-turning vehieles that store in the two-way Jeft-turn lane.

Number of Through Lanes on the Segment

The number of through lanes on the segment is the count of lancs that extend
for the length of the segment and serve through vehieles (even if a Jane is
dropped or added at a boundary intersection). This count is specified separately
for each direction of travel along the segmento A lane provided for the exclusive
use of turning vchieles is not included in this count.
If there is a rnidsegrnent Jane restriction, the number of through lanes equals
the number of lanes through the restriction. For example, if a work zone is
present and it requires one through Jane to be e1osed, the nurnber of through
lanes equals the munt of through lanes that remain open through the work zone
(and does not inelude the count of lanes that are e1osed).

Number of Lanes at Access Points

The number of lanes at an access point intersection is the count of Janes that
are provided for each traffic movement at the intersection. The method for
determining this nurnber follows the same guidance provided in a previous
paragraph for the nurnber of lanes at boundary intersections.
This input data elernent is needed for aHmovements on each active acccss
point approach and for aH major-street movements at the intersection with one
or more active access point approaches. Guidance for determining whether an
acccss point is "active" is provided in the section titled Access Point Flow Rate.

Turn Bay Length at Access Points

Turn bay length is the length of the bay at the access point intersectian for
which the lanes have fuIl width and in which queued vehieles can be stored. This
length is nccded for both segment approaches to the access point intersection.
The method for determining this length follows the same guidance provided in a
previous paragraph for turn bay Jength at boundary intersections.
This input data clement is needed for aH major-street turo movements at the
intersection with ane or more active access point approaches. Guidance for
determining whether an access point is "active" is provided in the section titled
Access Point Flow Rate.

5egment Length
Segment Jength is the distance bctween the boundary intersections that
define the segmento The point oí measurement at each intersection is the stop
line, the yield line, or the functional equivalent in the subject direction oí travel.

Motorized Vehicle Methodology O1apter 18/Urban 5treet 5egments

Page 18-18 Vmial6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

This length is rneasured along the centerHne of the street. If it differs in the two
travel directions, an average length is used.
The link length is used in sorne calculations. It is cornputed as the segment
length minus the width of the upstream boundary intersection.

Restrictive Median Length

The restrictive median length is the length of street with a restrictive median
(c.g., raised curb). This length is measured from median nose to median nose
along the centerlioe of the street. It does not indude the length of any median
openings on thc street.

Proportion of Segment with Curb

The proportion of the segment with curb is the proportion of the link length
with curb along the right side of the segment that is within 4 ft of the travelcd
way (i.e., within 4 ft of the nearest edge of traffic lane). This proportion is
computed as the ¡ength of street with a curb present (and within 4 ft) divided by
the link length. The l('ngth of street with a curb present is measured from the
start of the curbed cross section to the end of the curbed cross section on the link.
The width of driveway openings is /Jot deducted from this length. This
proportion is computed separately for each direction of travcl along the segmento

Number of Access Point Approaches

The number of access point approaches along a segment is the count of all
unsignalized driveway and public-street approaches to the segment, regardless
of whether the access point is considered to be active. This number is counted
separately for each side of the segmento It must equal or exceed the number of
active access point approaches for which delay to segment through vehides is
computed. Guidance for determining whcther an access point is "active" is
provided in the section titled Access Point Flow Rate. 1£the downstream
boundary intersection is unsignalized, its cross-street approach on the right-hand
side (in the direction of travel) is induded in the count.
Defau/t valuc. When the number of access points is not known, it can be
estimated from a specified access point density by using the following equation:
Do L Equation 18.1
Nap•s = 0.5 5,280
where Nap .• is the number of access point approaches on the right side in the
subject direction of travel (points), D. is the access point density on the segment
(points/mi), and L is the segment length (fcet). A default number af access points
can be dctermined from the default access point density obtained from
Exhibit 18-7.

Area Median Default Acq:ss Point Density Cooints'mil by Speed Limit (mjlbl Exhibit 18-7
Type Type 25 30 35 40 45 50 55 Default Access Point Density
Values
U""o Restrictive 62 50 41 35 30 26 22
Othec 73 61 52 46 41 37 33
Suburban Restrictive 40 II 19 12 7 3 O
or rural Othec 51 38 30 23 18 14 11

Olapter 18jUrban Street 5egments Motorized Vehide Methodology

VefSion 6.0 Page 18-19
Highway Capacity Manual; A Guide tor Multimodal Mobility Analys;s

Proportion of Segment with On-Street Parking

The proportion of the segment with on-street parking is the proportion of the
link length with parking stal1s (either marked or unmarked) available along the
right side of the segmento This proportion is computed as the length of street
with parking stalls divided by the link length. Parking stalls considered indude
those described as having either a parallel or an angle designo This proportion is
separately computed for each direction of travel along the segmento

Other Data and Performance Measures

This subsection describes the data listed in Exhibit 18.5 that are categorized
as "other data" or "performance measure data."

Analysis Penoo Duration

The analysis period is the time interval considered for the performance
evaluation. Its duration is in the range of 15 mio to 1 h, with longer durations in
this range sometimes used for planning analyses. In general, the analyst should
use caution in interpreting the results from an analysis period of 1 h or more
because the adverse impact of short peaks in traffic demand may not be detected.
Any 15.min period of interest can be evaluated with the methodology;
however, a complete evaluation should always indude an analysis of conditions
during the 15.min period that experiences the highest traffic demand during a
24-h periodo
Operatiollal analysis. A 15-min analysis period should be used for operational
analyses. This duration will accurately capture the adverse effects of demand
peaks.
Planning flllalysis. A 15-min analysis period is used for most planning
analyscs. However, a 1-h analysis period can be used, if appropriate.

Speed Limit
The methodology is based on the assumption that the posted speed limil is
(a) consistent with that found on other streets in the vicinity of the subject
segment and (b) consistent with agency policy concerning specification of speed
limits. If the posted speed limit is known not to satisfy these assumptions, the
speed limit value that is input to the methodology shouJd be adjusted so that it is
consistent with the assumptions.

Through Control Delay

The through control deJay is the control delay to the through movement at
the downstream boundary intersection. lt is computed by using the appropriate
procedure provided in one of Chapters 19 to 23, depending on the type of control
used at the intersection.
If the intersection procedure provides delay by lane groups and the through
movement is served in two or more lane groups, the through-movement delay is
computed as the weighted sum of the individuallane-group delays, where the
weight for a lane group is its proportion of through vehides.

Motorized Vehicle Methodology Qlapter 1a/Uman Street 5egments

Page 18-20 VersiotJ 6. O
 Highway Capacity Manual: A Guide for MultimodaJ Mobility Analysis

Through 5topped Vehic/es and Second- and Third- Term Back-of-Queue Size
Three variables are needed for the calculation of stop rate when the
downstream boundary intersection is signalized. They apply to the through lane
group at this intersection. A procedure Eorcomputing the number oE fully
stopped vchicles Nf second-tcrm back-of-queue size Q2' and third-term back-of-
queue size Q3 is provided in Chapter 31, Signalized Intersections: Supplemental.
If the procedure provides the stop rate by lane groups and the through
movement is served in two or more lane groups, the through-movement stop
rate is computed as the weighted sum of the individuallane-group stop rates,
whcre the weight for a lane group is its proportion of through vehieles.

capaCity
The capacity of a movement group is the maximum number of vehicles that
can discharge from a gueue during the analysis period, divided by the analysis
period duration. This value is needed for the movements entering the segment at
the upstream boundary intersection and for the movements exiting the segment
at the downstream boundary intersection. With one exception, it is computed by
using the appropriate procedure provided in one of Chapters 19 to 23,
depending on the type of control used at the interscction. Chapter 20, Two-Way
STOP-Controlled Intersections, does not provide a procedure Eorestimating the
capacity of the uncontrolled through movement, but this capacity can be
estimated by using Eguation 18-2.
Cth = 1,800 (Nrh -1 + Po.¡)
where
C'h through-movement capacily (veh/h),
N'h number of through ¡anes (shared or exclusive) (In), and
Po.¡ probability that there will be no qucue in the inside through lane.
The probability Po.¡ is computed by using Eguation 20-43 in Chapter 20. lt is
egual to 1.0 if a left-turn bay is provided for left turns from the major street.
If the procedure in Chapters 19 to 23 provides capacity by lane group and the
through movemcnt is served in two or more lane groups, the through-movemcnt
capacity is computcd as the weighted sum of the individuallane-group
capacities, where the weight for a lane group is its proportion of through
vehieles. A similar approach is used to compute the capacity for a tum
movement.

M/dsegment De/ay and Stops

Through vehic1es traveling along a segment can encounter a variety of
situations that cause them to slow slightly or e\'en come to a stop. These
encounters delay the through vehicles and cause their segment running time to
increase. Situations that can cause this dclay inelude
• Vehiclcs tuming from thc segment into an access point approach,
• Pedestrians crossing at a midsegment crosswalk,

Chapter 18¡Urban 5treet 5egments Motorized Vehicle Methodology

Vetsioo 6.0 Page 18-21
Highway Capacity Manual: A Guide for Mulfjmodal Mobility Analysis

• Vehicles mancuvering into or out of an on-street parking space,

• Double-parked vehicles blocking a lane, and
• Vehicles in a dropped lane that are merging into the adjacent lane.
A procedure is provided in the methodology for estimating the delay due to
vehicles tuming left or right into an access point approach. This edition of the
HCM does not inelude procedures for estimating the deJay or stops due to the
other sourccs listed. If they exist on the subject segment, they must be estimated
by the analyst and input to the methodology.

OVERVIEW OF THE METHODOlOGY

This subsection provides an overview of the methodology for evaluating the
performance of the urban strect segment in terms of its service to motorized
vehicles. The methodology is computationally intense and requires software to
implement. The intensity stems from the need to roodel the traffic movements
that enter or exit the segment in terms of their interaction with each other and
with the traffic control elements of the boundary intersection.
A planning-level analysis application for evaluating segment performance is
provided in Section 5 of Chapter 30, Urban Street Scgments: Supplemental. This
method is not computationalIy intense and can be applied by using hand
calculations.
Each travel direction along the segment is separately evaluated. Ullless
otherwise stated, all variables are specific lo tlle subject directioll o/ trovel.
The methodology has been developed to evaluate motorized vehicle
performance for a street segment bounded by intersections that can have a
variety of control types. The focus of the discussion in this subsection is on the
use of the methodology to evaluate a coordinated signal system because this type
of control is the most complexoHowever, as appropriate, the discussion is
extended to describe how key elements of this methodology can be used to
evaluate motorized vehiele performance in noncoordinated systems.
The objective of this overview is to introduce the analyst to the calculation
process and to discuss the key analytic procedures. This objective is achieved by
outlining the procedures that make up the methodology while highlighting
important equations, concepts, and interpretations. A more detailed discussion
of these procedures is provided in Sections 2, 3, and 4 of Chapter 30, Urban Street
Segments: Supplemental.
The computational engine developed by the Transportation Research Board
Committee on Highway Capacity and Quality of Scrvice represents the most
detailed description of this methodology. Additional information about this
engine is provided in Section 7 of Chapter 30.
A methodology for evaluating the performance of the motorized vehicle
mode on an urban street segment bounded by one or more roundabouts is
provided in Section 9 of Chapter 30.
Exhibit 18-8 illustrates the calculation framework of the motorized vehicle
methodology. It identifies the sequence of calculations nceded to estimate

Motorized Vehicle MethodoIogy Olapter 18/Urban 5treet 5egments

Page 18-22 Vef5ioo 6.0
 Hjghway Capadty Manual: A Guide for Multimodal Mobility Analysis

selected performance measures. The calculation process flows from top to boUom
in the exhibit. These calculations are described more fully in the next section.
lOe framework ilIustrates the calculation process as applicd to two system
types: coordinated and noncoordinated. The analysis of coordinated systems
recognizes the influence of an upstream signalized intersection on the
performance of the street segmento The analysis of noncoordinated systems is
based on the assumption that arrivals to a boundary intersection are random.
lOe framework is further subdivided into the type of traffic control used at DetiJils on (he methcdology for
segments wlth roundabouts as
the interscctions that bound the segmento This approach recognizes that a boundary fntersections can be
boundary intersection can be signalized, two-way SToP-controlled, all-way STOP- found in 5ection 9 of
Chapter 30.
controlled, or a roundabout. Although not indicated in the exhibit, the boundary
intersection could also be an interchange ramp terminal.
lOe methodology is shown to be iterative within Steps 1 to 4, with
convergence achieved when the predicted discharge volume, phase duration,
and capacity from successive iterations are effectively in agreemcnt. Several
iterations are typically nceded for coordinated systems. In contrast, only one
iteration is needed for noncoordinated systems unless there is a downstream lane
c10sure (e.g., a midsegment work zone), in which case multiple iterations are
nceded to ensure that the vehic1esdischarged upstream of the lane closure do
not exceed the lane closure capacity. The procedure far analyzing midsegment
lane restrictions is described in Section 3 of Chapter 30.
Procedures in othcr chapters are needed to evaluate an urban street segmento
For example, the procedure in $cction 3 of Chapter 19 for computing actuatcd
phase duration is nceded for the analysis of actuated intersections on both
coordinated and noncoordinated segments. AIso, the procedure in Section 3 of
Chapter 19 for computing control deJay is needed for the estimatian uf segment
through-movement delay. lOe capacity and control deJay estimation procedures
for roundabouts and all-way STOP-controlledintersections are nceded from their
respective chapters for the analysis of noncoordinated segments.

Chapter 18/Urnan Street 5egments Motorized Vehide MeI:t1odology

Ver5ion 6.0 Page 18-23
Highway Capacity Manual; A Guide tor Multimodal Mobility Analysis

Exhibit 18-8 Noocoor dinate d System Coordina ~System

Motorized Vehicle
Methodology for Urnan Street
5egments
Pretimed Actu,,,. Pretimed f-. Actuated
j j I
I Step 1: Determine Traffic Demand Adjustments

1 j I
I Step 2: Determine Running Time I
StoD. Vield I
UncornrOllell

Step 3: Determine the Proportion Arriving Step 3: Determine the Proportion Arriving
During Green DuringGreen
(e5li"",!e as "'lU"I ti> glCretiol (cnmpule by U5"'9 procedure)

j I 1
Step 4: Determine Step 4: Determine Step 4: Determine Step 4: Determine
Signal Phase Duration Signal Phase [)uration Signal Phase Duration Signal PI1aseDuration
(e<tinwe by u5ing known (corn¡:¡o.«eby u<ing (H:iI'Mte byusin¡;¡1mown (compute by usin¡;¡
procedu~)
Iimingl """"'"~) tirning)

1
I Converge?
No I Converge? ~
1 y" y"
I Step s: Determine Through Delay

1 1 1
I Step 6: Determine Through Stop Rate

1 1 1
I 5tep 7: Determine Travel Speed

1 1 1
I Step 8: Detenmne Spatial Stop Rate I
1 1
I $tep 9: Determine leve! of Service I
1 1
I $tep 10: Determine Automobile Traveler Perception 5core I
COMPUTATIONAL STEPS
5tep 1: Determine Traffie Demand Adjustments
During this step, various adjustments are undcrtaken to ensure that the
volumes evaluated accuratcly reflect segment traffic condUions. The adjustments
inelude (a) limiting entry to the segment because of capacity constraint, (b)
balancing the volurnes entering and exiling the segment, and (e) mapping entry-
to-exit flow paths by using an origin-destination matrix. Also during this step, a
check is made for the occurrence oí spillback from a lum bay or írom one
segment into anothcr segmento
The procedures for making the aíorementioned adjustmcnts and checks are
described in Section 2 oí Chapter 30. These adjustrnents and checks are not
typically used íor planning and preliminary engineering analyses. 1£spillback

Motorized Vehide Methodology Cllapter 18¡Urban 5treet 5egments

Page 18-24 V~6.0
 Highway Capacity Manual: A Guide tor Multimodal Mobility Analysis

occurs, the sustaincd spillback procedure should be used. It is described in

Scction 3 of Chapter 29, Urban Street Facilities: Supplementa!.

CCpaCity Constraint
When the demand volume for an intersection traffic movement exceeds its
colpolcity,the discharge volume from the intersection is restricted (or metered).
Whcn this metering occurs for a movement that enters the subject segment, the
volume arriving at the dovo,'nstreamsignal is reduced below the unrestricted
value.
To determine whether metering occurs, the capacity of each upstream
movement that discharges into the subject segrncnt must be computed and then
checked against the associated demand volume. If this volume exceeds
mavement colpacity,thc volume entering the segment must be reduced to equal
the movement colpacity.

Vo/ume Balance
Volume balance describes a conditian in which the combined volume from
aH movements entering a segment equals the combined volume exiting the
segmcnt, in a givcn direction of trave!. The segment is balanced when entering
valume equals exit volume for both directions af trave!. Unbalanced vnlumes
often exist in turo movement counts when the count at ane intersection and that
at the adjacent intersedion are taken at different times. They are .liso likcly when
access paint intersections exist but their volume is not counted.
The accuracy of the performance evaluation may be adversely affeded if the
volumes are nat balanced. The extent of thc impact is bolsedon the degree ta
which the valumes are unequal. To balance the valumes, the methadology
assumcs tholtthe valume for each mavement entering the segment is correct and
adjusts the valume for each movement exiting the segment in a proportional
molnncr so that a balance is achieved. The exiting volumes computed in this
manner represent a best estimate af the actual dcma/ld volumes, such that the
adjustment process does not preclude the possibility af queue buildup by one ar
more exit movements at the dawnstream baundary intersection during the
analysis periodo

Origin-Destination Distnbution
The volume nf traffic that arrives at a downstrcam intcrscctian for a givcn
downstream mavcment reprcscnts the combined valume from each upstreolm
paint of entry wcighted by its percentagc contribution to thc downstreolm
movement. Thc distribution of thesc contribution percentagcs between each
upstream and downstream pair is represented as an origin-destination
distribution matrix.
The concept af an origin-destination distribution matrix is illustrated by
example. Cansider the segment shown in Exhibit 18-9, which has four entry
points and four exit points. Thcre are thrce entry volumes at upstream
Intersection A that cantribute to threc exit volumes at downstream lntersedion
B. There are .lIso an entrance and an exit valume at the access point intersection
located between the two intersedians. The volume entering the segment,

Chapter 18/Urban 5treet 5egments Motorized Vehicle Methodology

VersiOn 6.0 Page 18-25
Highway Capacity Manual; A Guide for Multimoda/ Mobi/ity Ana/ysis

1,350 veh/h, is the same as that exiting the segment; thus, there is volume balance
for this example segmento The origin-destination distribution matrix for this
sample street segment is shown in Exhibit 18-10.

Exhibit 18-9
Entry and Exit Volume on
Example 5egment

II
Exhibit 18-10
Example Otigln-Destil'laoon
Distribution Matrix
-----:::::=:=::-=::::=::::::::::-----r--¡D~e;;"~;~n ••~ti~º~n~V~º~I¡;"~m~e
OriginVolumeby Moyement(yeh/hl To~1Vo_lu~e
left Throu h Ri ht AccessPoiot Movement (veh/hl
2 46 2 O left 50
188 877 95 50 Through 1,210
3 36 1 O Right 40
7 41 2 O Accessooint 50
200 1 000 100 50 1 350

The column totals in the last row oí Exhibit 18-10correspond to the entry
volumes shown in Exhibit 18-9.The row tota1s in the last column oí Exhibit 18-10
indicate the exit volumes. The individual cell values indicate the volume
contribution oí each upstream movement to each downstream movement. For
example, oí the 1,000through vehic1esthat enter the segment, 877 depart the
segment as a through movement, 46 depart as a left-tum movement, and so on.
The volumes in the individual cells are sometimes expressed as a proportion of
the column total.
The motorized vehide mcthodology computes one origin--destination matrix
íor movements between the upstream boundary intersection and a downstream
junction (Le., either an access point or the downstream boundary intersection).
When the boundary intersections are signalized, the matrix foc movements
between the upstream and downstream boundary intersections is used to
compute the proportion of vehides arriving during the green indication for each
exit movement. The matrix íor movcments between the upstream boundary
intersection and a downstream access point is used to compute the proportion of
time that a platoon is passing through the access point and effectively blocking
nonpriority movements from entering or crossing the street.

Spi//back Occurrence
Segment spillback can be characterized as one oí two types: cydic and
sustained. Cyc1icspillback occurs when the downstream boundary intersection is
signalized and its queue backs into the upstream intersection as a result oí queue
growth during the red indication. When thc grecn indication is prescnted, the

Motortzed Vehide Methodology Chapter 18/Urban 5treet 5egments

Page 18-26 Version 6.0
 Highway Capacify Manual: A Guide for Mu/timodal Mobility Analysis

queue dissipates and spillback is no longer present for the remainder of the cycle.
This type of spillback can ocrur on short street segments with relatively long
signal cyde lengths. The methodology may not provide a rcliable estimate of
segment performance if cyclic spillback ocrurs.
Sustained spillback ocrurs at sorne point during the analysis period and is a
result of oversaturation (i.e., more vehides discharging from the upstream
intersection than can be served at the subject downstream intersection). The
queue does not dissipate at the end of cach cycle. Rather, it remains present until
the downstream capacity is increased or the upstream demand is reduced.
The preceding discussion has forused on segment spillback; however, the
concepts are equally applicable to turn bay spillback. In this case, the queue of
tuming vehides exceeds the bay storage and spills back into the adjacent lane
that is used by other vchirular movements.
The occurrence of both sustained segmcnt and bay spillback must be
checked during this step. A procedure is described in Section 3 of Chapter 30 for
this purpose. 1£thc spillback does not ocrur during the analysis period (i.e., it
never ocrurs, or it ocrurs alter the analysis period), the methodology will
provide a reliable estimate of segment performance.
A procedure is described in Section 3 of Chapter 29 for evaluating the
ocrurrencc of sustained segment spilJback during the analysis periodo
If turn bay spillback occurs during the analysis period, the methodology may
not yield reHable performance estima tes. In this situation, the analyst should
considcr either (a) reducing the analysis period so that it ends before spillback
occurs or (b) using an alternative analysis tool that can modeI the eHect of
spillback conditions.

Step 2: Determine Running Time

A procedure for determining segment running time is describcd in this step.
This procedure indudes the calrulation of free-f1owspeed, a vehicle proximity
adjustment factor, and the additional running time due to midsegment deJay
sources. Each calrulation is disrussed in the following subparts, which rulminate
with the calrulation of segment running time.

A. Determine Free-Flow Speed

Free-f1owspeed is the average running speed of through vehicles traveling
along a segment under low-volume conditions and not delayed by traffie control
devices or other vehides. It reflects the eHect of the street environment on driver
speed choice. Elements of the street cnvironment that influenee this choice under
frce-flow conditions include speed Iimit, access point density, median type, curb
presence, and segment length.
The determination of free-flow speed is based on the ealculation of base free-
flow speed and an adjustment factor for signal spacing. These ealrulations are
described in the next fe\\' paragraphs, which culminate in the calrulation of frec-
flow spced.

Chapter 18/Urnan 5treet 5egments Motorized Vehicle Methodology

V~6.0 Page 18.27
Highway Copacity Manual: A Guide tor Multimodal Mobility Analys;s

Base Free--F/ow Speed

The base free-flow spccd is defined to be the free.flow speed on longer
scgments. It ¡neludes the influence of speed limit, access point density, median
type, curb presence, and on-street parking presence. It is computed by using
Eguation 18-3. AHernatively, it can be measured in the field by using the
technique described in $cction 6 of Chaptcr 30.
Equation 18-3 Sto = Scalib + So + fes + fA + fpk
where
Sji> base frec-flow speed (mi/h),
Sadil> base free-flow speed calibration factor (mi/h),
So speed constant (mi/h),
fes adjustment for cross section (mi/h),
fA adjustment for access points (mi/h), and
fpt adjustment for on-strcct parking (mi/h).
The spt.'Cdconstant and adjustment factors used in Equation 18-3 are Iisted in
Exhibit 18-11 (1). Eguations providcd in thc table footnote can also be used to
compute these adjustment factors for condUions not shown in the exhibit.

Exhibit 18-11 Speed Percent with Adjustment for Cross

BaseFree-FlowSpeed Restrictive Section fq(milht
AcIjustmentFactors S~/~;nit co(~~~~~~
mi h ~ MedianTvoe Median f%l No Curb Curb
25 37.4 Restrictive 20 0.3 O.,
30 39.7 40 0.6 -1.4
35 42.1 60 O., -1.8
40 44.4 80 1.2 -2.2
45 46.8 1001.5 -2.7
50
55
49.1
51.5
Noorestrictive
Nomechan ","
licable
licable
N""
0.0
0.0
-0.5
0.5
Access Adjustment for Access Points !A by Lanes Percent with Adjustment
Density D~ Nfh(mi/hY On-Street
(points/mi) 3 Lanes Parkin~(%l fO~r
~~r:~~g
mi hd
1 Lane 2 Lanes
O 0.0 0.0 0.0 O 0.0
2 -0.2 -0.1 -0.1 20 -0.6
4 -0.3 -0.2 -0.1 40 -1.2
10 -0.8 -0.4 -0.3 60 -1.8
20 -1.6 -0.8 -0.5 80 -2.4
40 -3.1 -1.6 -1.0 100 -3.0
60 -4.7 2.3 1.6
Notes: ~ ~ = 25.6 + 0.47~ where s,.. = posted speed Ilmlt (ml/h) .
• fes = 1.5 p",,- 0.47 Av¡,- 3.7 p"",p_ where p"" = propertion of link.length with restr~ median
(dedmal) and Av¡, = proportioo of segment with curb on the right'hand sKle (decimal).
e t.. = -0.078 D~IN", with D~'" 5,280 (N_ + N"",,)/(L- 111,),where D~= access polnt density on segment
N",
(pointslmi); = number of through lanes en !he segment In the subject directioo Of lravel (In);
N.",. = number of access polnt appr0ache5 on the right side in the subject direction of lravel (points);
N.,." = number of access peint approaches en the right side In the opposing direction of travel (points);
L = segment length (ft); and ¡.v, = wKlth of signalized intersection (ft).
"f", = -3.0 x proportion of link.length wittl on-street park.lng available on the right'hand side (dedmal).

Eguation 18-3 has been calibrated by using data for many urban street
segments collectively located throughout the United Sta tes, so the default value
of 0.0 mi/h for Sc.olib is believed to yield results that are reasonably representative
of driver behavior in most urban areas. However, if desired, a locally

Motorized
VehicleMethodology Qlapter18/Urban5treetSegments
Page 18-28 Vt'iWn 6.0
 Highway Capacity Manual: A Guide for Multimodal MobiJjty Analysis

representative value can be determined from Held-measured estimates of the

base free-flow spced for several street segments. The local default value can be
established for typical street segments or for specific street types. This calibration
factor is determined as the one value that provides a statisHcally based best-fit
between the predicHon from Equation 18-3and the field-measured estimates. A
procedure for cstimating the base free-flow speed from ficld data is dcscribed in
Section 6 of Chapter 30.

Adjustment for Signal Spacing

Empirical evidence suggests that a shorter segment length (when defined by
signalized boundary intersections) tends to influence the driver's choice of free-
flow speed (1). Shorter segments have been found to have a slower free-flow
speed, all other factors being the same. Equation 18.4 is used to compute the
value of an adjustment factor that accounts for this influence.
Sro - 19.5
h=1.02-4.7 ( )$1.0 Equation 18-4
max Ls,400
where
fL signal spacing adjustment factor,
S", base free-flow speed (mi/h), and
L, distance between adjacent signalized intersections (ft).
Equation 18-4 was derived by using signalized boundary intersections. For
more general applications, the definition of distance L, is broadened so that it
equals the distance between the two intersections that (a) bracket the subject
segment and (b) have a type of control that can impose a legal requirement to
stop or yield on the subject through movcmcnt.

Free-Flow Speed
The predicted free-flow speed is computed by using Equation 18-5on the
basis of estima tes of base free-flow speed and the signal spacing adjustment
factor. Alternatively, it can be entercd directly by the analyst. It can also be
measured in the Held by using the teclmique described in Chapter 30.
Sr = Sro fL ;:::Spl Equation 18-5

where SI is the free-flow speed (mi/h), SpI is the posted speed limil, and al1othcr
variables are as previously defined. The speed obtained from Equation 18-5 is
always greater than or equal to the speed Iimit.

8. Compute Adjustment for Vehiele Proximity

The proximity adjustment factor adjusts the free-flow running time to
account for the eHect of traHic density. The adjustment results in an increase in
running time (and corresponding reduction in speed) with an increase in
volume. The reduction in speed is a result of shorter headways associated with
thc higher volumc and drivers' propensity to be more cautious when headways
are short. Equation 18-6 is uscd to compute the proximity adjustment factor.

Chapter 18/Urban 5treet 5egments Motorized Vehicle Methodology

Version 6.0 Page 18-29
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

2
Iv = 0.21
1 + (1_ 52.8 v",Nth SI
)
where
fp proximity adjustment factor,

v'" midsegment demand flow rate (veh/h),

N1h number of through lanes on the segment in the subject direction of
travel (In), and
SI = free-flow speed (rni/h).
The relationship between running speed [= (3,600 L)I(5,280 tR)' where L is the
segment length in fcet and iR is the segment running time in scconds] and
volume for an urban street segment is shown in Exhibit 18-12. Trend lines are
shown for three specific free-flow speeds. At a flow rate of 1,000 vehicles per
hour per lane (veh/h/ln), each trend line shows a reduction of about 2.5 mi/h
relative to the free-flow speed. The trend lines extend beyond 1,000 veh/h/ln.
However, a volume in excess of this amount is unlikely to be experienced on a
segment bounded by intersections at which the through rnovement is regulated
by a traffic control device.

Exhibit 18-12
Speed-Flow Relationship for
Urban 5treet 5egments
:<
50 r=~===~=======~1
r-----'- I
I40~
~ r----+--::- __--:: _
~ 30 ~
~
~
.~ 20 -~~ ,--
•
~ T
10

OL---~-----------'-----'--'
O 500 1,000 1,500 2,000
Midsegment une Flow Rate (veh/h/ln)

e Compute De/ay due to Turning Vehic/es

Vehides tuming from the subject street segment into an access point
approach can cause a deJay to following through vehicIes. For right-tum
vehicles, the delay results when the following vehides' speed is reduced to
accommodate the turning vehide. For Idt-tum vehicles, the delay results when
the following vehicles must wait in queue while a vehide ahead executes a left-
tum maneuvcr at the access point intersection. Oclay due to left-tuming vehicles
occurs prirnarily on undivided streets; however, it can occur on divided streets
when the Icft-tum queue exceeds the available storage and spills back into the
inside through lane. A procedure for computing this delay at each access point
intersection is described in $echon 4 of Chapter 30.

Motorized Vehicle MetI10d0logy O1apter 18¡Urban Street 5egments

Page 18-30 VetWn6.0
 Highway Capacity Manual: A Gujde for Mulfimodal Mobi/ify Ana/ysjs

For planning and preliminary engineering analyses, Exhibit 18-13can be

used to estímate thc dciay due to tuming vehicles at one representatíve access
point intersection by using a midsegment volume that is typical for aHsuch
access points. The values in the exhibit represent the delay to through vehicles
due to left and right tums at one access point intersection. The selected value is
multiplied by the number of access point intersections on the segment to
estímate delay due to left and right toms (= I d.p in Eguation 18-7).

Midsegment Through yehicle Delav {slveh/ptl by Number Qf1hrough Lanes Exhibit 18.13
Volume veh h In 1 Lane 2 Lanes 3 Lanes Delay due to Tuming Vehicles
200 0.04 0.04 0.05
300 0.08 0.08 0.09
400 0.12 0.15 0.15
500 0.18 0.25 0.15
600 0.27 0.41 0.15
700 0.39 0.72 0.15

The values listed in Exhibit 18-13represent 10% left tums and 10% right
tums from the segment al the aecess point intersection. Jf the actual tum
percentages are less than 10%, the delays can be redueed proportionally. For
exampJe, if the subject aeeess point has 5% left turns and 5% right turns, the
vaJues Usted in the exhibit should be multiplied by 0.5 (= 5/10). AJso, if a tum bay
of adequate length is provided for one tom movement but not the other, the
values listed in the exhibit should be multiplied by 0.5. If both tum movements
are provided a bay of adeguate length, the dela}' due to tums can be assumed to
equal 0.0 second per vchide per access point (s/veh/pt).

D. Estimate De/ay due to Other Sources

Numerous other factors eould cause a driver traveling along a segment to
reduce speed or to incUTdela}'. For example, a vehicle that js campleting a
paraUel parking maneuver ma}' cause following vehiclcs to incur sorne deja}'.
Also, vehides that yield to pedestrians at a midsegment crosswalk may ineur
delay. Finally, bicyclists riding in a traffie Jane ar an adjacent bicyde lane may
directly or indirectly cause vehicular traffie to adopt a lower speed.
Among the many sources of midsegment delay, the matorized vehide
methodoJogy only ineludes procedures foc estimating the deja}' due to tuming
vehicles. However, if the delay due to other sourees is known or estimated by
other means, it can be included in the eguation to compute running time.

E. Compute Segment Running Time

Eguation 18.7 is used to compute segment running time on the basis of
consideration of through movement control at the boundary intersection, fr('('-
flow speed, vehicle proximity, and various midsegment delay sOUTces.
N"
6.0 - tI 3,600 L "\'
tR = 0.0025 L fx + 5,280 S fv + L. dap.i + dother Equation 18-7
f 1=1

Chapter 18JUman 5treet 5egments Motorized Vehicle Methodology

Version 6.0 Page 18.31
Highway Capacity Manual: A Guide for Mulfimodal Mobilify Analysis

with
LOO (signalized or STOP-controlledthrough movement)
f,x-- 0.00 (uncontrolled through movement)
Equation 18-8 Vth
{ min [Cth ,1.00] (VIELD-controlledthrough movement)

where
tR segment running time (s);
11 start-up lost time = 2.0 if signalized, 2.5 if STOPor YlELDcontrolled (s);

L segment length (ft);

Ix control.type adjustment factor;
V,h through-demand flow rate (veh/h);
C'h through-movement capacity (veh/h);
d"1'.i delay due to left and right turos from the street into access point
intersection i (s/veh);
N"1' number of influential access point approaches along the segment = N"1'.'
+ p."."N.".o (points);
N"1'o' number of access point approaches on the right side in the subject
direction of travel (points);
N.",o number of access point approaches on the right side in the opposing
direction of travel (points);
P"1'.u proportion of N.p ••• that can be accessed by a left turo from the subject
direction of travel; and
dOl"", = delay due to other sources along the segment (e.g., curb parking or
pedestrians) (s/veh).
The variables 11./x' v,•. and C¡h used with the first term in Eguation 18-7 apply
to the through movement exiting the segment at the boundary intersection. This
term accounts for the time required to accelerate to the running speed, less the
start-up 10st time. The divisor in this term is an empirical adjustment that
minimizes the contribution of this term for longer segments. lt partially reflects a
tendency for drivers to offset this added time by adopting slightly higher
midsegment spceds than reflectcd in the start-up lost time estimate.

Step 3: Determine the Proportion Arriving During Green

This step applies to the downstream boundary intersection when the
operation of a signalizcd urban street segment is evaluated. If the downstream
boundary intersection is not signalized, this step is skipped.
The methodology includes a procedurc for computing the proportion of
vehicles that arrive during the effective green time for a phase serving a segment
lane group (Le.,the lane groups "interna)" to the segment). That procedure is
described in this step. The platoon ratio (as described in Section 3 of Chapter 19,

Motorized Vehicle Methodology Olapter 18/Urban 5treet 5egments

Page 18-32 Version 6.0
 Highway Capacity Manual: A Guide lor Multimodal Mobility Analysis

Signalized lntersections) should be used to compute this proportion for phascs

serving externallane groups.
lf the upstrcam intersection is not signalized (or it is signalized but not
coordinated with the downstream boundary intcrsection), the proportion
arriving during thc grecn is equal to the effective green-to-cyele-length ratio and
this step is completcd. This relationship implies that arrivals are effectively
unifonn during the cyelc when averaged over the analysis periodo
If the boundary intersections are coordinated, the remaining discussion in
this step applies. Thc calculation of the proportion arriving during green is based
on the signal timing of the upstream and downstream boundary intersections.
However, if the signals are actuated, thc resulting estimate of the proportion
arriving during green typically has an eHect on signal timing and capacity. In
facl, the process is circular and requires an iterativc sequencc of calculations to
arrive at a convergence soIution in which all computed variables are in
agrccment with their initially assumed values. This process is illustrated in
Exhibit 18-8.This exhibit indica tes that the calculation ol average phasc duration
is added to this process \••..
hen the intersection is actuated.
Typically, there are threc signalized traHic movements that depart the
upstream boundary intersection at different times during the signa! cyele. They
are the cross-street right tu m, the major-street through, and the cross-street 112ft
turno Traffie may also enter the segment at vacious access point intersections. The
signalized movements often enter the segment as a platoon, but this platoon
disperses as the vehides move down the segmento
A platoon dispersion model is used to predict the dispcrsed flow rate as a
function of running time at any specified downstream location. The dispersed
flow rates for the upstream intersection movement are combined with aecess
point £low rates to predict an arrival flow prome at the downstream loeation.
Exhibit 18-14illustratcs the predicted arrival flow profile at the stop linc of the
downstrcam interseetion. This profile reflects the eombination of the 112ft-tu m,
through, and right-turn movements from the upstream intersection plus the tum
movements at the aeeess point intcrsection. The platoon dispersion model and
the manncr in which it is used to predict the dispersed £low rates for each of the
individual movemcnts are described in Section 3 of Chapter 30.

Exhibit 18-14
Comtlined Arri\lill E19w Profik:'oc IhrQugtH,¡¡oe Group
Use of an AITival Flow Profile
to Estímate the Volume
AITivíng During Green

Shaded area = arrival couot dUl"ing green ng

o green

o Time (steps)

Chapter 18/Urban 5treet Segments Motorized Vehicle Methodology

Version 6.0 Page 18-33
Highway Capocity Manual: A Guide for Multimodal Mobility Analysis

The gray shaded arca in Exhibit 18-14represents the arrival count during
green nlf This count is computcd by summing the flow rate for each time "step"
(or interval) that occurs during the effective green periodo The proportion of
vehicles arriving during the effective green period for a specified lane group is
computed with Equation 18-9.
Equation 18-9 n
p=-g-
q, e
where
P proportion of vehicles arriving during the green indication,
llx arriva! count during green (veh),
qd arrival flow rate for downstream lane group (veh/s), and

e cycle length (s).

Step 4: Determine Signal Phase Duration

This step applies to the downstream boundary intersection when the
operation of a signa!ized urban street segrnent is evaluated. If the downstream
boundary intersection is not signalized, this step is skipped.
If the downstream boundary intersection has pretirned signa! control, the
signal phase duration is an input value. If this intersection has sorne form of
actuated control, the procedure described in $cction 3 of Chapter 19 is used to
cstimate the average phase duration.
Steps 1 to 4 are repeated unti! the duration of each phase at each signalized
intersection converges to its steady-state value. Convergence is indicated when
the estimate of phase duration on two successive repetitions is thc same.

Step S: Determine Through Delay

Thc delay incurred by through vehicles as they exit the segment is the basis
for travel time estimation. In this context, a through vehicle is a vehicle that
enters and exits the segment as a through vehicle. The nature of the delay models
used in this manual makes it difficult to separate the delay to through vchiclcs
from the de!ay to nonthrough vehicles. However, these models can provide a
reasonab!e estimate of through de!ay whenevcr the through movement is the
dorninant movement on the segrnent.
Through delay is the sum of two delay sources. One source, caBed control
delay, is the delay due to the traffic control at the boundary intersection. The
other, caBed geometric de!ay, is that duc to the negotiation of intersection
gcometry, such as curvature.
Procedures for computing control delay are dcscribed in thc following
chapters of this manual:
• Signal control (Chapter 19 or 23),
• All-way STOPcontrol (Chapter 21), and
• YIELDcontrol at a roundabout intersection (Chapter 22).

Motofized Vehicle Methodology O1apter 18(Urban Street: 5egments

Page 18-34 VetWn6.0
 Highway Capacity Manual: A Guide for Mulfimoda/ Mobility Analysis

The analyst should refcr to the appropriate chapter for guidance in

estimating the through control delay for the boundary intersection. U the
through movement is uncontrolled at thc boundary intersection, the through
control deJay is 0.0 s/veh.
The geometric delay for conventional thrcc-Icg or four.leg intersections (i.c.,
noncircular intcrscctions) is considercd to be negligible. In contrast, the
geometric delay foc a roundabout is not negligible. This dclay can be estimated
by using the procedure provided in Section 9 of Chapter 30.
U thc segment is not in a coordinated system, the through delay estimate
should be based on isolated operation. The methodologies in Chapters 19 to 22
can be used to provide this estimate.
If the segment is within a coordinated signal system, the methodoJogy in
Chapter 19 (for most signalized intersections) or Chapter 23 (for signalized camp
terminals and aJtemative intersections) is used to determine the through delay.
The upstream filtering adjustment factor is used to account for the effect of the
upstream signal on the variability in arrival volume at thc downstream
intersection. The equation foc calculating this factor is desccibed in $ection 3 of
Chapter 19.
U the through movement shares one or more lanes at a signalized boundary
interscction, the through dclay is computed by using Equation 18-10.

dth vt Nt + ds1 vsl(l - Pd + dsr vsr(t - PR)

dt = ------------------ Equation 18.10
v<h
where
dr through deJay (s/veh),
Vlh through-demand flow rate (veh/h),
dlh delay in exclusive-through lane group (s/veh),
Vr demand flow rate in exclusive-through lane group (veh/hlln),
NI number of lanes in exc1usive-through lane group (In),
d,¡ delay in shared left-tum and through lane group (s/veh),
Vol demand flow rate in shared left.turn and through lanc group (veh/h),
d" delay in sharcd right-turn and through lane group (s/veh),
v" demand flow rate in shaced right-turn and through lane group (veh/h),
PL proportion of left-turning vehicles in the shaced lane (decimal), and
PR peoportion of right.turning vehicles in the shared lane (decimal).
The procedure described in Section 2 of Chapter 31, Signalized Intersections:
Supplemental, is used to cstimate the variables shown in Equation 18-10.

Chapter 18/Urban 5treet 5egments Motorized Vehicle Methcxlology

Version 6.0 Page 18-35
Highway Capacity Manual: A Guide lor Multimodal Mobility Analysis

Step 6: Determine Through Stop Rate

Through stop rate describes the stop rate of vehicles that enter and exit the
segment as through vehicles. The nature of the stop rate models described in this
step makes it difficult to separate the stops incurred by through vehicles from
those incurred by nonthrough vehides. However, the models can provide a
reasonable estimate of through stop rate whenever the through movement is the
dominant movement on the segment.
Stop rate is defined as the average number of full stops per vehicle. A full
stop is defined to occur at a signalized intersection when a vehide slows to zero
(or a crawl speed, if in queue) as a consequence of the change in signal indication
from green to red, but not necessarily in direct response to an observed red
indication. Afull stop is defined to occur at an unsignalized intersection when a
vehicle slows to zero (or a crawl speed, if in queue) as a consequence of the
control device used to regulate the approach. For example, if a vehide is in an
overflow queue and requires three signal cycles to dear the intersection, it is
estimated to have three full stops (one stop for each cyde).
The stop rate for a STOP-controlledapproach can be assumed to equal
1.0 stops/veh. The stop rate for an uncontrolled approach can be assumed to
equal 0.0 stops/veh. The stop rate at a YIELD-control1ed approach will vary with
conflicting demand.1t can be estimated (in stops per vehide) as equal to the
volume-to-capacity ratio of the through movement at the boundary intersection.
This approach recognizes that YIELD control does not require drivers to come to a
complete stop when there is no conflicting traffic.
The through stop rate at a signalized boundary intersection is computed by
using Equation lB-ll.

Equation 18-11
h = 3,600 [ N,
. (1'N sge ) gs
Vrll
+ Nth Qm]
Vtlle
mm
tll

with
Equation 18-12 NO Nt + Nf..sl(l- P,.) + N¡,rr(1- PR)
N¡
= ~---~--N-'-h--~----

St Nt + ssl(1- PL) + s.sA1- PR)

Equation 18-13
s = -------------
Nth
_ (Q", + Q,.,)N, + (Q2$'+ Q",)(l- P,) + (Q,." + Q,,,)(l- PR)
Equation 18-14 Q2+3 -
Nth
where
h full stop rate (stops/veh),
NI: number of fully stopped vehides (veh/ln),

g effective greeo time (s),

s = adjusted saturation flow rate (veh/h/ln),
back-of-queue size (veh/ln),

Motorized Vehicle Methodology Chaprer 18/Urban 5treet Segments

Page 18-36 Version 6.0
 Highway Capacify Manual: A Guide tor Multimodal MobiJjty Analysis

NJ,I number of fully stopped vehicles in exclusive-through lane group

(veh/ln),
Nfo! number of fully stopped vehicles in shared left-turn and through lane
group (veh/ln),
NJ.•• number of fully stopped vehic1esin shared right-turn and through
lane group (veMn),
Ntn number of through lanes (shared or exclusive) (In),
St saturation flow rate in exclusive-through lane group (veMl/ln),
S,¡ saturation flow rate in shared left-turn and through lane group with
permittcd operation (veh/h/ln),
s,. satura non flow rate in shared right-turn and through ¡ane group with
permitted operation (veh/h/ln),
Qtt second-term back-of-queue size for exc1usive-through lane group
(veh/ln),
Qt.<I second.term back-of-queue size for shared left-turn and through lane
group (veh/ln),
Qv, second-term back-of-queuc size for shared right-turn and through lane
group (veh/ln),
Q3.t third-term back-of.queue sizc for exclusive-through lane group
(veMn),
Q3,'¡ third.term back-of-queue size for shared lefHurn and through lane
group (veh/ln). and
Q3,>r third-terrn back-of-queue size for shared right-turn and through lane
group (veh/ln).
The procedure for computing Ni' Q:u and Q3 is provided in Scction 4 of
Chapter 31, Signalized lntersecnons; Supplemental.
The first terrn in Equation 18.11 represents the proportion of vehides
stopped once by the signa!. For sorne of the more complex arrival-departure
polygons that ¡nelude lefHurn movements operating with the permitted mode,
the queue may dissipate at two or more points during the cycle. If this occurs, N¡.;
is computed for each of the i periods between queue dissipation points. The
value of NI then equals the suro of the N¡.¡ values computed in this marmer.
The second terrn in Eguanon 18-11 rcprcsents the addinonal stops that may
occur during overflow (Le., cycle failure) conditions. The contribunon of this
term becornes significant when the volume-to-capacity ratio excecds about O.s.
The full stop rate typically varies from 0.4 stops/veh at low volume-to-capacity
ratios to 2.0 stops/veh when the volume-to-capacity ratio is about 1.0.

O1apter lB¡Uroan 5treet 5egments Motorized Vehicle Methodology

VernOn 6.0 Page lB-37
Highway Capacity Manual: A Guide tor Multimodal Mobility Analysis

Step 7: Determine Travel Speed

Equation 18-15 is used to compute the travel specd for the subject direction
of trave! along the segment
3,600 L
S ----
Equation 18-15 T.seg - 5,280 (t + d )
R t
where
Sr,gg travel speed of through vehides for the segment (mi/h),

L '" segment length (ft),

tR segment running time (s), and
dI through delay (s/veh).
The deJay used in Equation 18-15 is that incurred by the through lane group
at the downstream boundary intersection.

Step 8: Determine Spatial Stop Rate

Spatial stop rate is the stop rate expressed in units of stops per mile. It
provides an equitable means of comparing the performance of alternative street
segments with differing lengths. Equation 18.16 is used to compute the spatial
stop rate for the subject direction of travel along the segment.

Equation 18-16 ~ 5280 h + hother

Hseg' L
where
H~ '" spatial stop rate for the segment (stops/mi),

h ful! stop rate (stops/veh),

hot~, foil stop rate due to other sourc€s (stops/veh), and
L segment length (ft).
The full stop rate h used in Equation 18-16 is that incorred by the through
lane group at the downstream boundary intersection. In sorne situations, stops
may be incorred at midsegment locations due to pedestrian crosswalks, bus
stops, or turns into access point approaches.lf the foil stop rate assodated with
these other stops c~n be estimated by the analyst, it can be induded in the
calculation by using the variable h",,,,,,.

Step 9: Determine LOS

LOS is detennined scparately for both directions of trave) along the segment.
Exhibit 18.1lists the LOS thresholds established for this purpose. As indicated in
this exhibit, LOS is defined by two performance measures. One measure is the
trave) speed for through vehides. The second is the volume-to-capacity ratio for
the through movement at the downstream boundary intcrsection.
The travel speed LOS threshold value is shown in Exhibit 18-1 to be
dependent on the base free-flow speed. The base free-flow speed was computed
in Step 2 and the trave) speed was computed in Step 7.

Motorized Vehide MethodoIogy Chapter 18/Urban 5treet 5egments

Page 18-38 V~6.0
 Highway Capacity Manual: A Guide lor Multimodal MobWty Analysis

The volume-to-capacity ratio ior the through movement at the boundary

intersection is oomputed as the through volume divided by the through-
movement capacity. This capacity is an input variable to thc mcthodology.
The LOS determined in this step applics to the overall segment ior the
subject direction oi trave1.This LOS describes conditions ior the combined link
and downstream boundary intersection. If desired, the methodologies in
Chapters 19 to 23 can be used to determine the LOS ior travel through just the
downstream boundary intersection. The HCM does not indude a methodology
iar describing the LOS for just the link portion oi the segmento
LOS is probably more meaningiul as an indicator oi traffic performance
along a iacility rather than a single street segment. A procedure ior estimating
facility LOS is described in Chapter 16.

Step 10: Determine Automobile Traveler Perception Score

The automobile traveler perception score for urban street scgments is
provided as a useful performance measure. It indicates the traveler's perception
of service quality. The soore is computed with Equation 18-17 to Equation 18-22.
la.seg = 1+ PBCDEF + PCDEF + PDEF + PEF + PI' Equation 18-17
with
PBCDEF = (1 + e -1,1614 - 0.253 Hseg + 0,3434 PLTL.seg )-1 Equation 18-18

PCDEF = (1 + eO.6234 - 0.253 Hseg + 0.3434 PLTL.seg )-1 Equation 18-19

P
DEF
= (1 + e 1.7389 - 0.253 Hseg + 0.3434 PLTL.seg )-1 Equation 18-20

PEF = (1 + e2.7047 - 0.253 Hseg + 0.3434 PLTL.seg )-1 Equation 18-21

PI' = (1 + e 3.8044 - 0.253 Hseg+ 0.3434 PLTL.seg )-1 Equation 18-22

whcre
automobile traveler pcrception scare ior segment;
probability that an individual will respond with a rating oi B, C, D, E,
or F;

probability that an individual will respond with a rating of C, D, E, or

F;
PD[f probability that ao individual will respond with a rating oi D, E, or F;
probability that an individual will respond with a rating of E or F;
P, probability that an individual will rcspond with a rating oi F; and
proportion oi interscctions with a left-turn lane (or bay) on the
segment (decimal).
The derivation of Equation 18-17 is based on the assignment oi soores to cach
letter rating, in which a soore of "1" is assigned to the rating oí A (denoting
"best"), "2" is assigned to B, and so on. The survey results were used to calibrate
a set oi modcls that collectively predicts the probability that a traveler will assign
various rating combinations ior a specified spatial stop rate and proportion oi

Chapter 18¡Urban 5treet 5egments Motorized Vehide Methodology

Vf'rSk)n6.0 Page 18-39
Highway Capacify Manual: A Guide for Multimoda! Mobility Ana/ysis

intersections with lefHum lanes. The seore obtained fram Equation 18-17
represents the expected (or Iong.nm average) score for the population of travelers.
The praportion of intersections with left-tum lanes equals the number of Ieft-
tum lanes (oc bays) eneountered while driving along the segment divided by the
number of intersections encountered. The signalized boundary intersection is
eounted (if it exists). AH unsignalized interseetions of public roads are counted.
Private driveway interseetions are not eounted unIess they are signa! eontroIled.
The seore obtained fram Equation 18-17 pravides a useful indication of
performance from the perspeetive of the traveler. Seoees of 2.0 or less indicate the
best pereeived serviee, and values in excess of 5.0 indicate the worst perceived
service. Although this seore is dosely tied to the concept of service quality, it is
tlOt used to determine LOS for the urban street segment.

Motorized Vehide Methodology Qlapter 18/Urban Street Segments

Page 18-40 Version 6.0
 Highway Capacity Manual: A Guide for Mu/timodal Mobility Analysis

4. PEDESTRIAN METHODOLOGY

This section describes the methodology for evaluating the qualit)' of service
provided to pedestrians traveling along an urban street segment.

SCOPE OF THE METHOOOlOGY

The overall scope of the four mcthodologies was provided in Section 2. This
section idcntifies the additional conditions for which the pedestrian
methodology is applicable.
• Target fraveI modes. The pedestrian methodology addresses travcl by
walking in the urban street right-of-way. It is not designed to evaluate the
performance of other travel means (e.g., Scgway, roUer skates).
• "YypicaI pedestrian" focus. The pedestrian methodology is not designed to
reflect the perceptions of any particular pedestrian subgroup, such as
pedestrians with disabilities. The performance measures obtained from
the methodology are not intended to be indicators of a sidewalk's
compliance with U.5. Access Board guidelines related to the Americans
with Disabilities Act requirements. For this reason, they should not be
considered as a substitute for a formal compliancc assessment of a
pedestrian fadlity.

Spatiallimits
Travel Directions to Be Evaluated
Urban street performance from a pedestrian perspective is separately
evaluated for each side of the street. Unless otherwise stated, aII variables identijieli
ill /I,is sectioll are spccific fo fhe subject side of the strect. lf a sidewalk is not available
for the subject side of the str('('t, pedestrians are assumed to walk in the strcct on
the subject side (even if there is a sidewalk on the other side).
The typical evaluation will focus on the performance of the scgment (Le., the
link and boundary intersection combined). However, in sorne situations, an
evaluation of just the link is appropriate. Each approach is discussed in this
subsection.

Segment-Based Evaluation
For a segmcnt-based evaluation, the pedestrian methodology considers the
performance of thc link and the boundary intersection. It is applied through a
series of 10 steps culminating in the determination of the segment LOS.
A segment-based evaluation considers both pcdestrian space and a
pedestrian LOS score to determine segment LOS. It uses the worse of the LOS
letters resulting from pcdestrian spaee and thc scgment pedestrian LOS seore to
determine the overall segment pedestrian LOS. A segment-based evaluation is
recommended for analyses that compare the LOS of multiple travel modes
because each mode's segment LOS seore and letter can be directly compared.
Pedestrian spaee refleets the level of crowding on the sidewalk. Pedestrian
space typically onl)' influenees overall pedestrian LOS when pedcstrian facilities

Chapter 18/Urban 5treet 5egments Pedestrian Methodology

Version 6.0 Page 18-41
Highway Capacity Manual: A Guide lar Multimadal Mability Analysis

are very narrow, pedestrian volumes are very high, or both. For example, with
an effective sidewalk width oE 4 ft, pedestrian volumes need to be in excess of
1,000pedestrians per hour for the space-based pedestrian LOS to drop below
LOS A. Pedestrian space is not applicable when the pedestrian facility does not
existo
The methodology supports the analysis of a segment with either signal-
controlled or two-way STOP-controlledboundary intersections. $ection 5 of
Chapter 19 describes a methodology for evaluating signalized intersection
performance from a pedestrian perspective. No methodology exists for
evaluating two-way STOP-controlledintersection performance (with the cross
street STOPcontrolled). However, it is reasoned that this type of control has
negligible influence on pedestrian service along the segmento This edition of the
HCM does not inelude a procedure for evaluating a segmcnt's performance
when the boundary intersection is an all-way STOP-controlledintersection, a
roundabout, or a signalized interchange ramp terminal.

Unk-Based Evaluation
Only two of the 10 steps of the pedestrian methodology are used for link-
based evaluation of pedestrian service. This approach is regulariy used by local,
regional, and state transportation agencies. It offers the advantage oEbeing less
data.intensive than the fulllO-step methodology and produces results that are
generalIy reflective of pedestrian perceptions oE service along the roadway.lt can
be especially attractive when agencies are performing a networkwide evaluation
for a large number of roadway links.
The analyst should recognize that the resulting link LOS docs not consider
sorne aspects of pedestrian travel along a segmcnt (e.g., pedestrian space,
crossing difficuity, or intersection service). For this reason, the LOS score for the
link should not be aggregated to characterize facility performance. The analyst
should also be aware that this approach predudes an integrated muitimodal
evaluation because it does not reflect all aspects of segment performance.

Performance Measures
Performance measures applicable to the pedestrian travel mode inelude
pcdestrian travel speed, pedestrian space, and pedestrian LOS score. The LOS
score is an indication of the typical pedestrian's perception of the overall
segment travel experience.
LOS is also considered a performance measure. lt is useful for describing
segment performance to elected officials, policy makers, administrators, or the
publico LOS is based on pedestrian space and pedestrian LOS score.
"Pedestrian space" is the average amount of sidewalk area available to each
pedestrian walking along the segmento A larger area is more desirable from the
pedestrian perspective. Exhibit 18-15provides a qualitative description of
pedestrian space that can be used to evaluate sidewalk performance from a
circulation-area perspective.

Pedesbian Methodology Olapter 18/Urban 5treet 5egments

Page 18-42 Version 6.0
 Hjghway Capadfy Manual: A Guide tor Mulfimodal Mobility Analysis

pedestrlan SDaee rtr/p) Exhibit 18-15

Qualitative Description ot
Random Platoon
Pedestrian Space
Flow Flow Descri tion
>60 >530 Ability to move in desired path, no need to alter movements
>40-60 >90-530 occasional need to adjust path te avoid conflicts
>24-40 >4Q-90 Frequent need 10 adjust path to avoid conflicts
>15-24 >23-40 5peed and ability to pass slower pedestrians restricted
>8-15 >11-23 Speed restricted, very limited ability to pass slower pedestrians
~8 ~11 S severel restricted fr uent contact with other users

The first two columos io Exhibit 18.15 indicate a sensitivity to flow

condition. Random pedestrian flow is typical of most segments. Platoon flow is
appropriate for shorter segments (e.g., in downtown areas) with signalized
boundary intersedions.

Limitations of the Methodology

This subsection identifies the known limitations of the pedestrian
methodology. If one or more of these Iimitations are believed to have an
important influence on the performance of a specific street segment, the analyst
should consider using altemative methods or tools for the evaluation.
The pedestrian methodology does not account for the eHect of the following
conditions on the quality of sen'ice provided to pedestrians:
• Segments bounded by an all.way STOP-controlledinterscction,
roundabout, or signalized interchange ramp terminal;
• Midsegment unsignalized crosswalks;
• Grades in excess of 2%;
• Pedestrian overcrossiogs for service across or along the segment;
• Points of high-volume pedestrian access to a sidewalk, such as a transit
stop or a doorway from a large office building;
• Points where a high volume of vehides cross the sidewalk, such as a
parking garage entrance; and
• Presence of railroad crossings.

REQUIRED DATA ANO SDURCES

This subsection describes the input data needed for the pedestrian
methodology. The required data are listed in Exhibit 18-16.They must be
separately specified for each direction of travel on the segment and for each
boundary intersection. The exhibit also lists default values that can be used if
local data are not available (2, 3).
The data elements listed in Exhibit 18-16do not indude variables that are
considered to represent calibration factors. A calibration factor typically has a
relatively narrow range of reasonable values or has a small ¡mpad on the
accuracy of the performance estimates. The recommended value for each
calibration factor is identified at relevant points in the presentation of the
methodology.

Chapter 18/Urban 5treet 5egments Pedestrian Methodology

VersiOn 6.0 Page 18--43
Highway Capacity Manual: A Guide for MultimOdal Mobility Anolysis

Exhibit 18-16 Reguired Data and Units Potential Data Source(s) Suggested Default Value
Required Input Data, Potential
Traffic Characteristics
Data Sources,and Default
Midsegment motorized vehicle fIow Field data, past counts, Must be provided
Values for Pedestrian Analysis
rate. (veh/h) foreca"
Midsegment pedestrian flow rate (p/h) Fleld data, past counts Must be províded
Proportion of (lfl-street parking Field data 0.50 (if parking lane present)
oc:cupied(decimal)
Geometric Design
Downstream intersect:ionwidth. (ft) Field data, aerial piloto Must be provided
5egment length. (ft) Field data, aerial photo Must be provided
Number of midsegment through Ianes. Field data, aerial photo Must be provided
Outside l:hrough lane width (ft) Field data, aerial photo 12ft
Bicyde lane width (ft.) Field data, aerial piloto 5.0 ft. (if provided)
Paved outside shoulder width (ft) Field data, aerial photo Must be provided
Striped parking lane width (ft) Field data, aerial photo 8.0 ft. (if provided)
Curb presence (yes or no) Field data, aerial piloto Must be provided
5idewalk presence (yes or no) Field data, aerial photo Must be provided
Total walkway width (ft) Field data, aerial piloto 9.0 ft. (business(office uses)
11.0 ft (residential(industrial

Effective width of fixed objects (ft.) Fleld data

"""1
2.0 ft. ¡nside, 2.0 ft. outside
(business(office uses)
0.0 ft ¡nside, 0.0 ft outside
(residentialjindustrial uses)
Buffer width (ft) Fleld data, aerial photo 0.0 ft (business(office uses)
6.0 ft (residentialjindustrial

5paeing of objects in buffer (ft.) F¡elddata, aerial photo

"""1
Must be provided
Othe< Data
Oistanceto nearest signal-eontrolled Field data, aerial photo One-third the distance
crossing (ft) between signal-eontrolled
crossingsthat bracket the
segment
legality of midsegment pedestrian Field data, local traffie laws Must be provided
crossing (legal or iUegal)
Proportion of sidewalk adjacent to Field data 0.0 (non-CBOarea)
window, building, er feoce (decimal) 0.5 building. 0.5 window (eBO)
Performance Measures
Motorized vehicle midsegment running HCM method output Must be provided
speed. (mi{h)
Pedestriandelay at boundary HCM method output Must be provided
intersect:ion(S/p)
Pedestriandelay at midsegment HCM method output 20 s/p (if present)
signalized crosswalk (s(p)
PedestJiandelay at Ul)(()ntrolled HCM method output Must be provided
crossing (S/p)
PedestrianLOSscore fer intersection HCM method output Must be proYided
(decimal)
Notes: CBO= <;entralbusinessdistrict;p = persono
o AIsousedor Cil1culilted
bythe motoriredvet1iclemethodology.

Traffie Characteristics Data

This subsection describes the traffic eharacteristics data listed in Exhibit 18-
16. These data describe the motorized vehicIe and pedestrian traffie streams
traveling along the segment during the analysis periodo Midsegment f10w rate is
defined in a similarly titled section for the motorized vchicIe mcthodology.

Pedestrian Aow Rate

The pedestrian f10w rate is based on the count of pcdestrians traveling along
the outside of the subject segment during the analysis periodo A separate eount is

Pedestrian Mettlodology O1apter lB/Urban Street 5egments

Page 18-44 Version 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobmty Analysis

taken for each direction of travel along the side of the segmento Each count is
divided by the analysis period duration to yield a directional hourly flow rate.
These rates are then added to obtain the pedestrian flow rate for that side.

Proportion of On-Street Parking Occupied

This variable represents the proportion of the segment's right-hand curb line
on which parked vehieles are present during the analysis period.lt is computed
as the sum of the curb line lengths occupied by parked vehic1es divided by the
link length. The use of pavement markings to delineate the parking lane should
also be noted.
If parking is not allowed on the segment, the proportion equals 0.0. If
parking is allowed along the segment but the spaces are not used doring the
analysis period, the proportion equals 0.0. lf parking is allowed along the full
length of the segment but only one-half of the spaces are occupied during the
analysis period, the proportion equals 0.50.

Geometric Design Data

This subsection describes the geometric design data listed in Exhibit 18-16.
These data describe the geometric elements that influente pedestrian
performance. Al! input data should be representative of the segment for its entire
length. An average value should be used for each element that varies along the
segmento Segment length and number of through lanes are dcfined in a similarly
titled section for the motorized vehic1e methodology.

Downstream Intersection Width

The intersection width applies to the downstream boundary intersection for
a given direction of travel and represents the effedive width of the cross street.
On a two-way street, it is the distance between the stop (or yield) line for the two
opposing segment through movcments at the boundary intersection, as
measured along the centerline of the segmento On a one-way street, it is the
distance from the stop line to the far side of the most distant traffic lane on the
cross strect.

Width of Outside Through Lane, Bicyc/e Lane, Outside Shou/der, Parking Lane
The widths of several individual elements of the cross section are considered
input data. These elements inelude the outside lane that serves motorized
vehic1es traveling along the segment, the bicyc1elane adjacent to the outside lane
(if used), paved outside shoulder, and striped parking lane.
The outside lane width does not inelude the width of the gutter. If curb and
gutter are present, the width of the gutter is inc1uded in the shoulder width (Le.,
shoulder width is measured to the curb face when a curb is present).

Curb Presence
The presence of a curb on the right side edge of the roadway is determined
for each segment travel direction.

Chapter 18/Urban 5treet 5egments Pedestrian Methodology

Version 6.0 Page 18-45
Highway Capacity Manual: A Guide tar Multimadal Mability Analysis

Presence of a Sidewalk
A sidewalk is a paved walkway that is provided at the side of the roadway.
Pedestrians are assumed to walk in the street if a sidewalk is not presento

Tota/ Wa/kway Width

Total walkway width is measured from the outside edge of the road
pavement (or face of curb, if present) to the far edge of the sidewalk (as sometimes
delineated by a building face or landseaping). It includes the width of any buffer
(see below), if presento If this width varies along the segment, an average value is
used. A paved shoulder is not includcd in this width measurement.

Effective Width of A)(ed Objects

Two input variables are used to describe fixed objects along the walkway.
One represents the effective width of objects along the inside of the sidewalk.
These objccts include light poles, traffie signs, planter boxes, and so forth.
Typieal widths for these objects are provided in Chapter 24, Off-Street Pedestrian
and Bieycle Facilities. AH objects along the sidewalk should be considered and an
average value for the length of the sidewalk input to the methodology.
The second variable represents the effcctive width of objects along the outside
of the sidewalk. lt is determined in the same manner as was the first variable.

Buffer Width and Spacing of Objects in Buffer

The buffer width is the distance bctween the outside edge of the paved
roadway (or face of curb, if present) and the near edge of the sidewalk. This
element of the cross scction is not designed for use by pedestrians or motorized
vehicles. lt may be unpaved or inc1ude various vertical objeets that are
eontinuous (e.g., barrier) or diseontinuous (e.g., trees, bollards) to prevent
pedestrian use. If vertical objects are in the buffer, the average spadng of objects
that are 3 ft or more in height should also be recorded.

Other Data
This subsection describes the data Iisted in Exhibit 18-16that are categorized
as "other data."

Distance to Nearest Signa/-Control/ed Crossing

This input variable is needed if there is an identifiable pedestrian path (a)
that intersects the segment and continues beyond the segment and (b) on which
most crossing pedestrians trave1. This variable defines the distance pedestrians
must travel along the segment should they divert from the path to cross the
segment at the nearest signalized crossing. The crossing will typically be at a
signalized intersection. However, it may also be at a signalized crosswalk
provided at a midsegment location. If the crossing is at a signalized intersection,
it willlikely occur in the crosswalk on the side of the intersecnon that is nearest
to the segmento Occasionally, it will be on the far side of the intersection because
the near-side crosswalk is closed (or a crossing at this location is otherwise
prohibited). This distance is measured along one side of the subject segment; the

Pedestrian MethodoIogy Chapter 18/Urban Street 5egments

Page 18-46 Version 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

methodology accounts far the returo distance once the pedestrian arrives at the
other side of thc scgment.

Legality of Midsegment Pedestrian Crossing

This input indicates whethcr a pedestrian can cross the segmcnt at any point
along its length, rcgardless of location. If making this crossing at any point is
iIIega!. thc pcdcstrian is assumed to be required to divert to the nearest
signalized intersection to cross the segmento

Proportion of Sidewalk Adjacent to Window, Building, or Fence

Thrce proportions are input far a sidewalk. One proportion represents thc
length of sidewalk adjacent to a fence or low wall dividcd by thc Icngth of thc
link. The second represcnts thc length of the sidewalk adjacent to a building face
divided by the length of the link. The final propartion represents the Jength of
the sidewaJk adjaccnt to a window display divided by the length of the link.

Performance Measures
This subsection describes the data listed in Exhibit 18.16 that are categorized
as "performance measures."

Motorized Vehicle Running Speed

The motorized vehicle running speed is based on the segment running time
obtained from the motorizcd vchicle methodology. The running spt..'edis equal to
the segment Icngth divided by the segment running time.

Pedestrian De/ay
Three pedestrian delay variables are needed. The first is the deJay to
pedestrians who travel through the boundary intersection along a path that is
parallel to the segment centerline. The pedestrian movement of interest is
traveling on the subject side of the street and heading in a dircction that is "with"
or "against" the motarized traffic stream. For a two-way STOI'-controllcd
boundary intersection, this deJay is reasoncd to be ncgligible. For a signaJ-
controlled boundary intersection, the procedure described in Section 3 oí
Chapter 19 is used to compute this delay.
Thc sccond deJay variable describes the delay incorred by pedestrians who
cross thc subject segment at the /leares! signaJ-controlled crossing. If the nearest
erossing is at a signalized intersection, the procedure deseribed in Seetion 3 of
Chapter 19 is used to compute this delay. If the nearest erossing is at a midscgment
signalized crosswalk, this deJay shouJd equal the pedestrian's average wait for
service after the pedestrian push button is pressed. This wait will depcnd on the
signal settings and eould range from 5 to 25 seconds per pedestrian (s/p).
Thc third delay variable needed is the pedestrian waiting delay. This delay is
incorred when pedestrians wait at an uncontrolled crossing loeation. If this type
of crossing is legal, the pedestrian waiting delay is determined by using thc
procerlure in Chapter 20, Two-Way STOP-Controlledlntersedions. If it is iIIegal,
the pedestrian waiting delay does nol need to be calcolated.

Chapter 18/Urban Street 5egments Pedestrian Methodology

VersiOn 6.0 Page 18-47
Highway Capacity Manual: A Guide for Multimodal MobUify Analysis

Pedestrian LOS SCOre for lntersection

The pedestrian LOS score for the signalized intersection is obtained from the
pedestrian methodology in Chapter 19.

OVERVIEW OF THE METHODOLOGY

This subsection provides an overview of the methodology for evaluating the
performance of an urban street segment in terms of its service to pedestrians. The
methodology consists of 10 calculation steps. These steps are illustrated in
Exhibit 18.17. AlllO steps are completed for a typical segment-based evaluation.
Only Steps 6 and 7 are nceded for a link-based evaluation.

Exhibit 18-17 1
Pedestrian Methodology fof
Urban 5treet 5egments Step 1: Determine Free-Aow Step 6: Determine Pedestrian LOS
Walking Speed 500re for link

I I
Step 2: Determine Average
Step 7: Determine link lOS
Pedestrian Space

I I
Step 3: Determine Pedestrian Delay Step 8: Determine Roadway
atIntersection Crossing DiffiQJlty Factor

I I
Step 4: Determine Pedestrian Travel Step 9: Determine Pedestrian LOS
Speed Score for Segment

I I
Step 5: Determine Pedestrian LOS
Step 10: Determine Segment lOS
$eore for Intersection

A methodology for evaluating off-street pedestrian facilities is provided in

Chapter 24, Off-Street Pedestrian and Bicycle Facilities.

COMPUTATIONAL STEPS
Step 1: Determine Free.Flow Walking Speed
The average free-flow pedestrian walking speed Sr! is needed for the
evaluation of urban street segment performance from a pedestrian perspective.
This speed should reflect conditions in which there are negligible pedestrian-to-
pedestrian conflicts and negligible adjustments in a pedestrian's desired walking
path to avoid other pedestrians.
Research indicates that walking speed is influenced by pedestrian age and
sidewalk grade (4). If 0% to 20% of pedestrians traveling along the subject
segment are elderly (Le., 65 years of age or older), an average fra"-flow walking
speed of 4.4 ft/s is recommended for segment evaluation. If more than 20% of
pedestrians are elderly, an average free-flow walking speed of 3.3 ft/s is
recommended. In addition, an upgrade of 10% or greater reduces walking speed
by 0.3 ft/s.

Pedestrian Methodology Chapter 18/Urban 5treet 5egment5

Page 18-48 V~6.0
 Highway Capacity Manual: A Guide tor Multimodal Mobility Analysis

Step 2: Determine Average Pedestrian Space

Pedeslrians are sensitive to the amount of space separating them from other
pedestrians and obstades as they walk along a sidewalk. Average pedestrian
space is an indieator of segment performance for travel in a sidewalk. It depends
on the effective sidewalk width, pedestrian flaw rate, and walking speed. This
step is nat applicable when the sidewalk does not exist.

A. Compute Effective Sidewa/k Width

The effective sidewalk width equals the total walkway width less the
effective width of fixed objects located on Ihe sidewalk and 11..'55 any shy distanee
associated with the adjaeent street or a vertical obstruction. Fixed objeets can be
eontinuous (e.g., a fence or a building faee) or discontinuous (e.g., trecs, poles, or
benehes).
1111..' effective sidewalk width is an average value for Ihe length of the link. It
is eomputed by using Equation 18-23to Equation 18-27.
WE = WT - WO,i - Wo.o - Ws,t - ~,o 2: 0.0 Equation 18-23

with
Ws.i = max (Wbuf.1.5) Equation 18-24

~,o = 3.0 Pwindow + 2.0 Pbuilding + 1.5 Pfence Equation 18-25

WO•i = wO,i - Ws.i 2: 0.0 Equation 18-26

Wo,o = wO.o - ~.o ~ 0.0 Equation 18-27

where
Wr effective sidewalk width (ft),
Wr total walkway width (ft),

Wo) adjusted fixed-object effective width on inside of sidewalk (fl),

Wo.c adjusted fixed-object effeetivc width on oulside of sidewalk (ft),
W•.i shy distanee on inside (curb side) of sidewalk (ft),
W... shy distance on outside of sidewalk (fl),
W~"f buffer width between roadway and sidewalk (ft),
Pwirodow proportion of sidewalk length adjacent to a window display (decimal),
Pbuilding proportion of sidewalk length adjacenl lo a building faee (decimal),
Pl~ proportion of sidewalk length adjacent tú a fence or low wall
(decimal),
U'o) cffective width oí fixed objects on inside of sidewalk (ft), and
U'o.. effective widlh of fixed objeets on outside of sidewalk (ft).
The relationship between the variables in thesc equations is iIlustrated in
Exhibit18.18.

Chapter lB/Urnan 5tTeet 5egments Pedestrian Methodology

VersiOn 6.0 Page 18-49
Highway Capacity Manual: A Guide for Multimodal MobUity Analysis

Exhibit 18-18 Street

Widtt1Adjustments for Fixed
Objects

ToMI w(llkway wldth, W,

.¡
Effe-ctlve walkway wldtl'l,WE

1.S n! .-.-.-. • • ¡"~O

ObJect lIne (fence 07 low wall) Building race Building tace wlth
wlndow display
•• Shy dlstance •• Flxed-object effectlve width

The variables Wr, Wbul' Pwindow' Pbuildin¡;' Pknc-t< wO.j, and wO.o are input variables.
They represent average, oc typical, values for the length of the sidewalk.
Chapter 24, Off-Street Pcdestrian and Bicyde Facilities, provides guidance for
estimating the effective width of many common fixed objects.
Typical shy distances are shown in Exhibit 18-18. Shy distance on the inside
(curb side) of the sidewalk is measured from the outside edge of the paved
roadway (or face of curb, if prescnt). It is generally consideced to equall.5 ft. Shy
distance on the outside of the sidewalk is 1.5 ft if a fence or a low wall is prescnt,
2.0 ft if a building is present, 3.0 ft if a window display is present, and 0.0 ft
othcrwise.

B. Compute Pedestrian Flow Rate per Unit Width

The pedestrian flow per unit width of sidewalk is computed by using
Equation 18-28 foc the subject sidewalk. Thc variable vrtd is an input variable.
1Jped
Equation 18-28 1Jp = 60W
E

where
vp .., pedestrian flow per unit width (p/ft/min),
vp<d pedestrian flow rate in the subject sidewalk (walking in both
directions) (pfh), and
Wr = effective sidewalk width (ft).

C. Compute Average Wa/king Speed

The average walking speed Sp is computed by using Equation 18-29. This
equation is derived from the relationship between flow rate and average walking
speed described in Exhibit 24-1 oí Chapter 24.
Equation 18-29 Sp = (1 - 0.00078 vi) Sp! 2: 0.5 Sp!
where Sp is the pedestrian walking speed (ft/s), Spfis the free-fIow pedestrian
walking speed (ft/s), and vp is the pedestrian flow per unit width (p/ft/min).

Pedestrian Methodology Olapter 18/Urban 5treet 5egments

Page 18-50 Version 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

D. Compute Pedestrian Space

Finally, Equation 18-30is used to compute average pedestrian spaee.
Sp
Ap = 60- Equation 18-30
vp
where Ap is the pcdestrian space (ft2/p)and all other variables are as previously
defined.
The pcdestrian space obtained from Equation 18-30can be compared with
the ranges provided in Exhibit 18.15 to rnake sorne judgrnents about the
performance of the subject intersection comer.

Step 3: Determine Pedestrian Delay at Intersection

Pedestrian delay at three locations along the segment is determined in this
step. Each of these delays is an input variable for the rnethodology and is
described in the previous subsedion titled Required Data and Sources.
The first deJay variable dpp represents the delay incurred by pedestrians who
travel through the boundary intersection along a path that is paralleJ to the
segment centerline. The second delay variable dr< represents the deJay incurred
by pcdestrians who cross the segment at the ncarest signaJ-controlled crossing.
The third deJay variable dp'" represents the deJay incurred by pedestrians waiting
for a gap to cross the segment at an uncontrolled location.

Step 4: Determine Pedestrian Travel Speed

Pedestrian traveJ speed represents an aggregate measure of speed along the
segmento lt combines the deJay incurred at the downstream boundary
intersection and the time required to walk the length of the segmentoThus, it is
typically slower than thc average walking speed. The pedestrian travel speed is
computed by using Equation 18-31.
L
Srv.seg = L Equation 18-31
S+d
p
pp

where
STP.~ travel speed oí through pedestrians for the segment (ft/s),
L scgment length (ft),
Sp pedestrian walking speed (ft/s), and
dpp pedestrian delay incurred in walking parallel to the segment (s/p).
In general, a traveJ speed of 4.0 ft/s or more is considered desirable and a
speed of 2.0 ft/s or less is considered undesirable.

Step 5: Determine Pedestrian lOS Seore tor Intersection

The pedestrian LOS seore for thc boundary intersection Ip•in, is determined in
this step. If the boundary intersection is signalizcd, the pedestrian methodology
described in Chapter 19 is used for this determination.lf the boundary
intersection is two-way STOPcontrolled, the seore is equal to 0.0.

Chapter 18/Urban Street 5egments Pedestrian Methodology

Version 6.0 Page 18.51
Highway Capacify Manual: A Guide for Multimodal Mobility Ano/ysis

Step 6: Determine Pedestrian LOS Score for Link

The pcdcstrian LOS seore far the link l¡dink is calculated with Equation 18-32.
Equation 18-32 lp.link = 6.0468 + Fw + Fv + Fs
with
Equation 18-33

Equation 18-34

Equation 18-35

where
lp.link pedestrian LOS score for link;
F IV cross-section adjustment factor;
F~ motorized vehicle volume adjustment factor;
Fs motorized vehicle spced adjustment factor;
In(x) naturallog of x;
W~ effective total width of outside through lane, bicyele lane, and
shoulder as a function of traffic volume (see Exhibit 18-19)(ft);
W¡ total width of shoulder, bicycle lane, and parking lane (see Exhibit 18-
19) (ft);
PpI< = proportion of on-street parking occupied (decimal);
W••¡ = buffer width between roadway and available sidewalk (= 0.0 if
sidewalk does not exist) (ft);
lb = buffer area coeffident = 5.37 far any continuous barrier at least 3 ft
high that is located bctween the sidewalk and the outside edge of
roadway; otherwise use 1.0;
W,., available sidewalk width = 0.0 if sidewalk does not exist or WT- Wb,,¡if
sidewalk exists (ft);
WT total walkway width (ft);
WoA adjusted available sidewalk width = min(W,." 10) (£t);

1"" sidewalk width coefficient = 6.0 - 0.3 W.A;

v.. midscgment demand f10wrate (direction nearest to the subject
sidewalk) (veh/h);
N1h number oi through lanes on the segment in the subject direction oi
travel (In); and
SR = motorized vehiele running speed = (3,600 L)/(5,280 tR) (mi/h).
The value used for several of the variables in Equation 18-33to Equation 18-
35 is dependent on various conditions. These conditions are identified in
Column 1 of Exhibit 18-19.1£the condition is satisficd, the equation in Column 2

Pedestrian Methodology Olapter 18/Urban 5treet 5e9ments

Page 18-52 Version 6.0
 Highway Capacify Manual: A Guide tor Mu/fimodal Mobility Analysis

is used to compute the variable value. If it is not satisfied, the equation in

Column 3 is used. The equations in the first h\'o rows are considered in sequenee
to determine the effeetive width of the outside lane and shoulder W~.

Variable When Condition Variable When Condition Exhibit 18-19

Condition Is Satisfied 15 Not Satisfied Variables
fO( Pedestrian
LOS
Vm > 160 veh/h or WA > O ft W. = Wa/+WI>' + W,,;+ W,,* W,. =(W",+WI>'+W": +Wpl)x 5core for Link
(2 - 0.005 v",)
p,,* > 0.25 Of w",+W,,/+ W,,*:S 10 W, = WI>'+ W",' +W"" Wi=lO
Notes: W•• = widlh of the outside lhrough lane (ft);
W",' = adjustoo width of paved outside shoolder; jf curo is present W",' = W",- 1.5 O!:0.0, otherwise W,,;
= W",(ft);
W", = widlh of paved outside shoulder (ft);
W". = widlh af the bkyde lane = 0.0 if bicyde lane not provided (ft); ar,d
W", = wKntl of striped parking lane (ft).

The buffer width coeffident determination is based on the presence oí a

eontinuous barrier in the buffer. In making this determination, repetitive vertical
objects (e.g., trces or bollards) are eonsidered to represent a continuous barrier if
they are at least 3 ft high and have an average spacing oí 20 ft or less. For
example, the sidewalk shown in Exhibit 18-18does not have a eontinuous buffer
beeause the street trees adjaeent to the curb are spaced at more than 20 ft.
The pedestrian LOS score is sensitive to the separation between pedestrians
and moving vehicles and to the speed and volume of these vehicles. Physical
barriers and parked cars between moving vehicles and pedestrians effeetively
increase the separation distance and the perceived guality of service. Higher
vehicle speeds or valumes lower the pereeived guality oí service.
If the sidewalk is not eontinuous for the length of the segment, the segment
should be subdivided into subsegments and eaeh subsegment separatcly
evaluated. For this application, a subsegment is defined to bcgin or end at eaeh
break in the sidewalk. Eaeh subsegment is then separately evaluated by using
Eguatian 18-32. Eaeh eguation variable is uniquely quantified to represent the
subscgment to whieh it applies. The buffer width and the effeetive sidewalk
width are each set to 0.0 ft for any subsegment without a sidewalk. The
pedestrian LOS seore Ir.link is then computed as a weighted average of the
subsegment scores, where the weight assigned to each seore eguals the portian of
the segment length represented by the eorresponding subsegment.
The motorized vehicle running speed is computed by using the motorized
vehicle methodolagy, as described in Section 3.

O1apter18/Urban5treetSegments PedestrianMethoclology
VersiOn 6.0 Page18-53
Highway Capacify Manual: A Guide tor MultimodaJ Mobility Analysis

Step 7: Determine Link LOS

The pedestrian LOS for the link is determined by using the pedestrian LOS
score from Step 6. This score is compared with the link-based pedestrian LOS
thresholds on the right side of Exhibit 18-2 to determine the LOS for the specified
direction of travel along the subject link.

Step 8: Determine Roadway Crossing Difficulty Factor

The pedestrian roadway crossing difficulty factor measures the difficulty of
crossing the street between boundary intersections. Segment performance from a
pedestrian perspective is reduced if the crossing is perccived to be difficult.
The roadway crossing difficulty factor is based on the deJay incurred by a
pedestrian who crosses the subject segmento One crossing option the pedestrian
may consider is to alter his or her travel path by diverting to the nearest signal-
conlrolled crossing. This crossing location may be a midsegment signalized
crosswalk or a signalized intersection.
A second crossing option is to conUoue on the original travel path by
completing a midsegment cros-<;ingat an uncontrolled Iocation. If this type of
crossing is legal along the subject segment, the pedestrian crosses when there is
an acceptable gap in the motorized vehicle stream.
Each of these two crossing options is considered in this step, with the option
requiring the least delay used as the basis for computing the pedestrian roadway
crossing difficulty factor. The time to walk across the segment is common to both
options and therefore is not included in the delay estimate for either option.

A. Compute Diversion De/ay

The delay incurred as a consequence of diverting to the nearest signal-
controlled crossing is computed first. It ¡neludes the del ay involved in walking to
and from the midsegment crossing point to the nearest signal-controlled crossing
and the delay waiting to cross at the signa!. Hence, calculation of this delay
requires knowledge of the distance to the nearest signalized crossing and its
signal timing.
The distance to the nearest crossing location D< is based on one of two
approaches. The first approach is used if there is an identifiable pedestrian path
(a) that interseets the segment and continues on beyond the segment and (b) on
which most crossing pedestrians travel. The location of this path is shown for
two cases in Exhibit 18-20. Exhibit 18-20(a)iIlustrates the distance Dc when the
pedestrian divcrts to thc ncarest signalizcd interscction. This distance is
measured frorn thc crossing location to thc signalized intersection.
Exhibit 18-20(b) ilIustrates the distance D<when a signalized crosswalk is
provided at a midsegment location. In this situation, the distance is mcasured
from the pedestrian crossing location to the location of the signalized crosswalk.
In eithcr case, the distance Dcis an input value provided by the analyst.
The second approach is used if crossings occur somewhat uniforrnly along
the length of the segmento In this situation the distance Dccan be assumed to

Pedestrian Methodology Chapter IB/Urban 5treet Segments

Page 18-54 Versoo 6.0
 Highway Capacity Manual: A Guide tor MultimodaJ MobWty Analysis

equal one-third of the distance bctween the nearest signaI-controlled crossings

that bracket the subject segment.

1 Exhibit 18.20

,. Iv,
•
Diversion Distance
Components

••
A

•
TypiUlI Pedestri¡m Peth

(a) Divert to Nearest Boundary Intersection

(b) Divert to Midsegment Signalized Crosswalk

The diversion distance to the nearest crossing is computed with Equabon

18.36.
Equation 18.36

where
Dd diversion distance (ft), and
De distance to nearest signaI-controlled crossing (ft).
lf the nearest crossing localion is at the signalized intersection and the
crossing is at Location A in Exhibit 18-20(a),Equation 18-36applies directly.1f
the nearesl crossing location is at the signalized intersection but the crossing is al
Location B, the distance obtained from Equation 18-36should be increased by
adding two increments of the intersection width W¡.
The del ay incorred due to diversion is calculated by using Equation 18-37.
Vd
dpd =S+dpc Equation 18-37
P

where
dpd pedestrian diversion delay (s/p),
Dd diversion distance (ft),
Sp '" pedestrian walking speed (ft/s), and

O1apter 18/Urban 5treet 5egments Pedestrian Methodology

Version 6.0 Page 18-55
Highway Capacity Manual: A Gujde fer Mu/tjmedal Mebility Analysis

d".. = pedestrian deJay incurred in crossing the segrnent at the nearest signaJ-
controned crossing (s/p).
The pedestrian delay incurred in crossing at the nearest signal-controlled
crossing was deterrnined in Step 3.

B. Compute Roadway Crossing Difficu/ty Fador

The roadway crossing difficulty factor is computed by using Equation 18-38.
0.10 dpx - (0.318 Ip.link+ 0.220 lp.int + 1.606)
Equation 18-38 Fcd = 1.0 + 7.5

where
F(,j roadway crossing difficulty factor,

dpx crossing delay = min(d¡w dI"<' 60) (s/p),

drd pedestrian diversion delay (s/p),

dI"" pedestrian waiting deJay (s/p),

Ip.~nk pedestrian LOS score for link, and
Ip•in, pedestrian LOS score for intersection.
If the factor obtained from Equation 18-38 is less than 0.80, the factor is set
equal to 0.80.lf the factor is greater than 1.20, it is set equal to 1.20.
The pedestrian waiting delay was determined in Step 3. If a midsegrnent
crossing is illegal, the crossing delay determination does not inelude
consideration of the pedestrian waiting deJay dI"" [i.e., dpx = min(dp4<60)].

Step 9: Determine Pedestrian LOS SCore tor Segment

The pedestrian LOS score for the segment is computed with Equation 18-39:
,
(FCd lp.link + 1)3~ + (lp,int + 1)3 dpp]'
Equation 18-39 Ip,seg = 0.75 L + 0.125
[
,
r+dpp

where Ip..-g is the pedestrian LOS score for the segment and aH other variables are
as previously defined.

Step 10: Determine Segment LOS

The pedestrian LOS for the segrnent is deterrnined by using the pedestrian
LOS score from Step 9 and the average pedestrian space from Step 2. These two
performance measures are compared with their respective thresholds in Exhibit
18-2 to determine the LOS for the specified direction of travel along the subject
segment. lf a sidewalk does not exist and pedestrians are relegated lo walking in
the slreet, LOS is determined by using Exhibit 18-3 because the pedestrian space
concept does nol apply.

Pedestrian Methodology Olapter 18/Uman 5treet 5egments

Page 18-56 Vt"lSiOn 6.0
 Highway Capacity Manual: A Guide tar Multimodol Mobility Analysis

5. BICYCLE METHODOLOGY

This scction describes the mcthodology for evaluating the qua lity of service
provided to bicyelists traveling along an urban street segment.

SCOPE OF THE METHODOLOGY

The overall scope of the four methodologics was providcd in $ection 2. This
section identifies the additional condinons for which the bicyele methodology is
applicable.
• Target travelmodes. Thc bicyele mcthodology addresses travel by bicyele in
the urban street right-of-way.lt is not designcd to evaluate the
performance of other travel means (e.g., motorized bicyele, rickshaw).
• Shared or exclusive lanes. The bicyele methodology can be used to evaluate
the service provided to bicyclists when thcy share a lane with motorized
vehieles or when they travel in an exclusive bicyele lane.

Spatial Limits
Travel Directions to Be Evaluated
Urban street segment performance fram a bicyclist perspective is separately
evaluated for each travel direction along the street. Ullless otherwise stated, a1l
variables idelltijied iI, this ser/ion are sperific to the sllbject directiotl o/ travel. Thc
bicyele is assumed to travel in the street (possibly in a bicyele lane) and in the
same direction as adjacent motorized vehicles.
The typical evaluation will focus on the performance of thc segment (i.e., the
link and boundary intcrscction combined). However, in sorne situanons, an
evaluation of just the link is appropriate. Each approach is discussed in this
subscction.

Segment-Based Evaluation
For a segment-based evaluation, the bicycle methodology considers the
performance of the link and the boundary intersection. lt is applied through a
series of eight steps that culminate in the determinanon of the segment LOS.
The methodology supports the analysis of a segmcnt with either signal-
controlled or two-way STOI'-controllcdboundary interscctions. Chapter 19
describes a methodology for evaluating signalized inlersection performance from
a bicyclist perspective. No methodology exists for evaluating two-way 5TOI'-
controlled intersection performance (with the cross street STOPcontrolled).
However, the influence of this type of control is incorporated in the methodology
for evaluating segment performance. This edition of the HCM does nol include a
procedure {or evaluating a segment's performance whcn the boundary
intersection is an all.way STOP-controlledintersection, a roundabout, or a
signalized interchange ramp terminal.

Chapter l8¡Uman Street 5egments Bicycle Methodology

Vt'fSiOn 6. O Page 18-57
Highway Capacity Manual: A Guide lar Multimadal Mability Analysis

Unk-Based Eva/uation
Only two of the eight steps of the bicyele methodology are used for link-
based evaluation of bicyele service. This approach is regularly used by local,
regional, and state transportation agencies. 1toffers the advantage of being less
data-intensive than the full eight-step methodology and produces results that are
generally reflective of bicyelist perceptions of service along the roadway. lt can
be especially attractive when agencies are performing a networkwide evaluation
for a large number of roadway links.
The analyst should recognize that the resulting link LOS does not consider
sorne aspects of bicyele travel along a segment (e.g., intersection service). For this
reason, the LOS score for the link should not be aggregated for the purpose of
characterizing facility performance. The analyst should also be aware that this
approach preeludes an integrated multimodal evaluation because it does not
reflect all aspects of segment performance.

Performance Measures
Performance measures applicable to the bicyele travel mode inelude_bicyele
travel speed and bicyele LOS score. The LOS score is an indication of the typical
bicyelist's perception of the overall segment travel experience.
LOS is also considered a performance measure. lt is useful for describing
segment performance to elected officials, policy makers, administrators, or the
publico LOS is based on the bicyelist LOS score.

Limitations of the Methodology

This subsection identifies the known limitations of the bicyele methodology.
lf one or more of these limitations are believed to have an important influence on
the performance of a specific strt..>etsegment, the analyst should consider using
altemative methods or tools for the evaluation.
The bicyele methodology does not account for the effect of the following
conditions on the quality of service provided to bicyclists:
• Segments bounded by an all-way SfOr-controlled intersection,
roundabout. Of signalized interchange ramp terminal;
• Grades in excess of 2%; and
• Presence of railroad crossings.

REQUIRED DATA ANO SOURCES

This subsection describes the input data needed for the bicyele methodology.
The required data are Iisted in Exhibit 18-21. They must be separately specified
for each direction of travel on the scgment and for each boundary intersection.
The exhibit also lists default values that can be used if local data are not available
(2,3).

Bicycle Methodology O1apter 18/Urban 5treet Segments

Page 18-58 Vem:¥l6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Required Data and Units Potential Data Source{s) Suggested Default Value Exhibit 18-21
TrafflC CJlaracteristics RequiredInput Data, Potential
Midsegmentmotorizedvehiclef10wrate4 Fielddata, past counts, Mustbe provided Data Sources, and Default
(veh{h) forecasts Valuesfor Bicy<:le
Analysis
Heavyvehiclepercentage (%) Fie!ddata, past counts Mustbe provided
Proportiooof oo-street parkingOCC1JpiedFielddata 0.50 (if parking lane present)
(decimal)
Geometric Design
5egment length4 (It) Fielddata, aerial photo Mustbe provided
Numberof midsegment through !anes~ Fielddata, aerial photo Mustbe provided
Outside through lane width (ft) Fielddata, aeria! photo 12ft
Bícyde lane wídth (It) Fielddata, aerial photo 5.0 ft (if provided)
Palled outsíde shoulder wídth (ft) Fielddata, aerial piloto Mustbe provided
StTipedparking lane width (ft) Fielddata, aerial photo Mustbe provided
Mediantype (dividedor undivíded) Flelddata, aerial photo Mustbe provided
Curb presence (yes or no) Fielddata Mustbe provided
Numberof access point approaches Fielddata, aerial photo 5ee discussion in text
OtiIer Data
Pavement condltionC Fielddata, pavement 3.5 (good)
(FHWA5-point scale) conditioninventory
Performance Measures
Motorizedvehiclemidsegment running HCMmethod output Mustbe provided
speed' (mVh)
Bicycledelay at bounclaryint. <stbicycle} HCMmethod output Mustbe provided
BicyclelOS score at boundary int. HCMmethod output Mustbe provided
(decimal)
Notes: FHWA" Federal H ghway AdministTation; int " intersectiOll.
Bold italicindicates high sensJtivity (:!o2lOS letters) of lOS to Ihe choice of default value.
Bold indk:ates moderate sensitivity (:tIlOS Ietter) of lOS to the choice of default vall.le .
• Also used or calculated by the motorized vehide methodology.
~Sensitivity reflects pavement oonditions 2-5. Very poor pavement (Le., 1) typically results in lOS F,
reg.ardless of olher input value5.

The data elements listed in Exhibit 18-21 do not ineludc variables that are
considered to represent calibration factors. A calibration factor typically has a
relatively narrow range of rcasonable values or has a small impact on the
accuracy of the performance estima tes. The recommended value for each
calibration factor is identified at relevant points in the prescntation of the
methodology.

Traffic Characteristics Data

This subsection describes the traffic characteristics data listed in Exhibit 18.
21. These data describe the motorized vehiele and bicyele traffic streams
traveling along the scgment during the analysis periodo Midsegment flow rate is
dcfined in $ection 3 for the motorized vehiele mode. The "proportion of on-street
parking occupied" is dcfined in $ection 4 for the pedestrian mode.
A heavy vehicle is defined as any vchiele with more than four tires touching
the pavement. Local buses that stop within the intersection area are not ineluded
in the count of heavy vehieles. The percentage ofheavy vehicles is the count of
heavy vehieles that arrive during the analysis period divided by the total vehiele
count for the same periodo lhis percentage is provided for the same location on
the segment as represented by the midsegment flow rateo

O1apter 18jUrban 5treet 5egments BicycleMethodology

Ve~1On6.0 Page 18-59
Highway Capacify Manual: A Guide for Multimodal Mobility Analysis

Geometric Design Data

This subsection describes the geometric design data listed in Exhibit 18-21.
These data describe the geometric elements that influence bicyde performance.
AH input data should be representative of the segment for its entire length. An
average value should be used foc each element that varies along the segmento
Most of the geometric design input data are defined in previous sections.
Segment length, number of through lanes, and number of access point
approaches are defined in SecUon 3 for the motorized vehide mode. The
following variables are defined in Section 4 foc the pedestrian mode: width of
outside through lane, width of bicyde lane, width of paved outside shoulder,
width of striped parking lane, and curb presence.
Median type is designated as "undivided" or "divided." A street is indicated
to have a divided median type if it has a nonrestrictive median (e.g., two-way
left-tum lane) or restrictive (e.g., raised curb) median; otherwise, it is undivided.

Other Data
This subsection describes the data listed in Exhibit 18-21 that are categorized
as "othcr data."
The pavemellt COllditioll rating describes the road surface in terms of ride
quality and surface defects. It is based on the present serviceability rating, a
subjective rating system based on a scale of Oto 5 (5). Exhibit 18-22 provides a
description of pavement conditions associated with various ratings.

Exhibit 18-22 Pavement Motorized Vehide

Pavement CollditionRating Condition Ride Quality and
Rating Pavement Description Trame Speed
4.0 to 5.0 Newor near1ynew superior pavement. Free of Good <id,
cracks and patches.

3.0 to 4.0 Aexiblepavements may begin to show evidence of Go<xl ride

rutting and fine cracks. Rigidpavements may begin
to show evidence of minar cracking.

2.0 to 3.0 Aexiblepavements may show rutting and extensive Acceptab/eride for Iow.
patching. Rigidpavements may have a few joint speed traffic but barely
fractures, faulting, or cracking. tolerable foc high-speed
traffic
1.0 to 2.0 Distressoc:cursover 50% or more of the sulfacc. Pavement deterioration
Aexiblepavement may have large potholes and affects the speed of free--
deep cracks. Rigidpavement distress ¡ncludesjoint now traffic; ride quality
spalling, patching, and cracking. not acceptable

0.0 to 1.0 Distressoc:cursover 75% or more of the surface. Passable onty at reduced
Large potholes and deep cracks exisl speed alld considerable
rider discomfort

Performance Measures
This subsection describes the data listed in Exhibit 18-21 that are categorized
as "performance measures."

BicycleMethodology Chapter 18jUrban5treet 5egments

Page 18-60 Vet:S"1On 6. O
 Highway Capacity Manual: A Guide for Multimoda/ Mobi/ity Analysis

Motorízed Vehic/e Running Speed

The motorized vehide running speed is based on the segment running time
obtaincd from thc motorizcd vehide methodology. The running speed is equal to
the segment length divided by the segment running time.

8icyc/e De/ay
Bieyc1edeJay is the delay to bicydists who travcl through the boundary
intersection along a path that is parallel to the segment centerline. The bicyde
movement of interest is traveling on the subject side of the street and heading in
the same direction as motorized vehides. For a two-way STOP-controlled
boundary intersection, this deJay is reasoned to be negligible. For a signal-
controlled boundary intersection, the procedure described in Section 3 of
Chapter 19 is used to compute this delay.

Bicyc/e LOS 5core for lntersection

The bicyde LOS score for the signalized intersection is obtained from the
bicyde mcthodology in Chaptcr 19.

OVERVIEW OF THE METHODOLOGY

This subsection provides an overview of the methodology for evaluating the
performance of an urban street segment in terms of its service to bicyclists. The
methodology consists of eight calculatian steps. These steps are illustrated in
Exhibit 18-23.AHcight stcps are completed for the typical segment-based
evaluation. Gnly Steps 5 and 6 are needed for a link.based evaluation.

Exhibit 18-23
I Bicycle Methodology for Urban
Step 1: Determine Bi'Yde Running Step 5: Determine Bi'Yde LOS $core Street 5egments
Speed for Link

J I
Step 2: Determine Bi'Yde Delay
Step 6: Determine link LOS
at lntersection

1 j
Step 3: Determine Bi'Yde Travel Step 7: Determine Bi'Ycle LOS Score
Speed for SC9ment

I 1
Step 4: Determine Bicyde LOS Score
Step 8: Determine Segment LOS
for lntersection

I
A methodology for cvaluating off-street bicycle facilities is provided in
Chapter 24, Off-Strect Pedestrian and Bicyde Facilities.

Olapter 18/Urban Street 5egments Bicycle Methodology

VersiOn 6.0 Page 18-61
Highway Capacity Manual; A Gujde for Multimodol MobUity Anolysis

COMPUTATIONAL STEPS

Step 1: Determine Bicycle Running Speed

An estimate of the average bieyc1erunning speed SIJ is determined in this step.
The best basis for this estimate is a field measurement of midsegment bieyc1e
speed on representative streets in the vidnity of the subjeet street. In the absenee
of this information, the average running speed of bicyc1es is recommended to be
taken as 15 mi/h between signalized intersections (6). Many faetors might affeet
bieycle speed, including adjacent motor vehicle traffie, adjacent on-street parking
activity, commerdal and residential driveways, lateral obstruetions, and
significant grades. To date, researeh is nol available to make any spcdfic
recommcndations as to the effect of these faetors on speed.

Step 2: Determine Bicycle Delay at Intersection

Bicycle delay at the boundary interseetion dlJ is eomputed in this step. This
delay is incurred by bieyclists who travel through the intersection in the same
lane as (or in a bicyc1eJane that is parallel to) the lanes used by segment through
vehicles.
If the boundary intersection is two-way STOPcontrolled (where the subjeet
approach is uncontrolled), the delay is egual to 0.0 slbicyc1e.If the boundary
intersection is signalized, the deJay is eomputed by using the motorized vehicle
methodology deseribed in Chapter 19, Signalized Intersections.

Step 3: Determine Bicycle Travel Speed

Bieyc1etravel speed represents an aggregate measure of speed along the
segmento It combines the delay incorred at the downstream boundary
interseetion and the time reguircd to ride the length of the segmento Thus, it is
typicalIy slower than the average bieyc1erunning speed. The average bicycle
travel spL'€dis computed by using Eguation 1840:
3,600 L
Equatlon 18-40 STb,seg = 5,280 (tRb + db)
where
Sn."'X traveJ speed of through bicycles along the segment (mi/h),
L segment length (ft),
tRb segment running time of through bieycles = (3,600 L)/(5,280 S~)(s),
Sb bicycle running speed (mi/h), and
db bicycle control delay (s/bicyele).
In general, a travel speed of 10.0 mi/h or more is considered desirable and a
speed of 5.0 mi/h or less is considered undesirable.

Step 4: Determine Bicycle lOS Score tor Intersection

The bicyele LOS seore for the boundary intersection lb.iwt is determined in this
step. If the boundary intersection is signalized, the bicyde methodology
described in Chapler 19 is used for this determination. If the boundary
intersection is two-way STOPcontrolled, the seore is egual to 0.0.

Bicyde Methodology Chapter 18¡Urtan 5treet 5e9ments

Page 18-62 VetSon 6.0
 Highway Capacify Manual: A Gujde for Multimodal Mobility Analysis

Step 5: Determine Bicycle LOS Score for link

The bicycle LOS seore for the segment lb,li"k is ealculated by using Equation
18-41,

lb,link= 0.760 + Fw + Fv + Fs + Fp Equation 18-41

with
Fw = -0.005 We2 fquation 18-42

Fv = 0.507 In --
4Nth
("m. ) Equation 18-43

F, = 0.199 [1.1199In(5R• - 20) + 0.8103](1 + 0.1038?Hv.)' fquation 18-44

7.066
Fp =---p2 Equation 18-45
<
where
lUn\; bicycle LOS seore for link,
FUI cross-seetion adjustment factor,
F~ motorized vehicle volume adjustment factor,
Fs motorized vehicle speed adjustment factor,
Fp pavement condition adjustment factor,
In(x) naturallog of x,
W. effective width of outside through lane (see Exhibit ]8-24) (U),

v"'" adjusted midsegment demand flow rate (see Exhibit 18.24) (veh/h),
N1h number of through lanes on the segment in the subject direction of
travel (In),
SR4 adjusted motorized vehicle running speed (see Exhibit ]8.24) (mi/h),
PHV• adjusted percent heavy vehicles in midsegment demand flow rate (see
Exhibit ]8-24) (o,{,),and
Pe = pavement condition rating (see Exhibit ]8-22).
The value used for several of the variables in Equation 18-42to Equation 18-
45 is dependent on various conditions. These conditions are identified in
Column 1 of Exhibit 18-24. If the condition is satisfied, the eguation in Column 2
is used to compute the variable value. If it is not satisfied. the equation in
eolumn 3 is used. The equations in the first three rOW5are considered in
sequence to determine the effective width of the autside through lane W,.
The motorized vehicle running speed is computed by using the motorized
vchicle methodology described in a previous subsection.

Chapter lB/Urban Street 5egments Bicycle Methodology

Ver.;ion 6.0 Page 18-63
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Exhibit 18-24 Variable When Condition Variable When Condition

Variables for Bicyde LOS Condition 15satisfied 15 Not satisfjed
SCore for link W,= W,..+ Wb'+ W",' +Wpk Wr= W,..+ WIt+W",'
Ppk '" 0.0
v'" > 160 veh/h or street is divided W,= W, W, = m(2 - O.QOSvm)
W¡< 4.0ft W~= W. - 10ppk ~ 0.0 W~=W,+W¡ -20ppk~ 0.0
v'" (1 - 0.01 P"",) < 200 veh/h P""., = 50% P""., = P"",
and Pw> 50%
S" < 21 mí/h SR.> = 21 mi/h 5110 = S"
vm>4N,,,, v••• = v", v"" = 4N",
Notes: W; = total wi(!th of the outside through laoe, bicyi:1e laoe, alld paved shoulder (ft);
W••.= width of out:side through Iaoe (ft);
W,,:= adJusted widltl of paved outside shoulder; if curo is present w",,' = W",,- 1.5 ~ 0.0, otherwise
W",,' = W",,(ft);
W",,= width of paved outside shoulder (t't);
WI>'= width of bicycle Iaoe = 0.0 if bicyde lane not provided (ft);
w,.,= width of striped parking tane (ft);
w,. total width of shoulder, bicycle laoe, and par1<.ingIane = W,. + W",,' + W'" (ft);
w,= effective total widltl of outside ltlrough Iane, bicycle Iane, and shoulder as a function of traffi<;
vOlume (t't);
p",= proportion of O""5treet paoong oa.:upied (decimal);
v,. = midsegmef1t demalld Ilow rate (vell/h);
PIN= perceot heavy vehides in the midsegment demand fIow rate; and
SR = motorized vehide running speed (mi/h).

Step 6: Determine Link LOS

The bicycle LOS for the link is determined by using the bieycle LOS score
from Step 5. This score is compared with the Iink-based bicycle LOS thresholds
in Exhibit 18-3 to determine the LOS for the specified direction of travel along the
subject link.

Step 7: Determine Bicycle LOS Score tor Segment

The bicycle LOS score for the segment is computed by using Equation 18-46:
3 3
,
-
Equation 18-46
lb.sen = 0.75 [(F" + Ib,li,' + 1) 'R,b: (lb,in, + 1) dbr + 0,125
tRb + b
with
Equation 18.47 5,280 N )
Fe = 0.035 L ap.s - 20
(

whcre
I~."X bicycle LOS score for segment;
I~.li"" bicycle LOS score for link;
Fe unsignalized conflicts factor;
I~.iw' bicycle LOS score for intersection; and
N""" numbcr of access point approaches on the right side in the subject
direction of travel (points).
The count oí access point approaches used in Equation 18-46 includes both
public street approaches and driveways on the right side of the segment in the
subject direction oí travel.

Bicyde Methoclology Olapter 18/Urban Street 5egments

Page 18-64 Version 6.0
 Highway Capacity Manual: A Guide for Mulfimodal Mobility Analysis

Step 8: Determine Segment LOS

The bieycle LOS far the segment is detennined by using the segment bicycle
LOS scare from Step 7. This scare is compared with the segment-based bicyc1e
LOS thresholds in Exhibit 18-3 to determine the LOS far the spedfied direction of
travel along the subject segmento

O1apter 18¡Urban 5treet 5egments Bicycle MethodoIogy

VefSiOn 6.0 Page 18-65
Highway Capacify Manual: A Guide for Mulfimodal Mobilify Analysis

6. TRANSIT METHODOLOGY

This section describes the methodology for evaluating the capacity and
quality of servicc provided to transit passengers on urban street segments.

SCOPE OF THE METHODOLOGY

The overall scope of the four methodologies was provided in Section 2. In
addition, thc transit methodology is limited to the evaluation of public transit
vehicles opcrating in mixcd or exclusive traffic lanes and stopping along the
street. lt is not designed to evaluate the performance of other travel means (e.g.,
grade-separated rail transit).

Spatial Limits
Trave/Directions to Be Eva/vated
Urban street segment performance from a transit passenger perspective is
separately evaluated for cach travel direction along the street. U"less otherwise
sfafed, all variables ide"tijied in this section are specific fo the subject direcfion o/ fmvd.
Route-Based Eva/uation
The methodology is used to evaluate a single transit mute on thc segmento If
multiple routes exist on the segment, each route is eva!uated by using a separate
app!ication of the methodology.

Performance Measures
Performance measures applicab!e to the transit trave! mode includeJransit
vehicle travel speed, transit wait-ride score, and transit LOS score. The LOS
score is an indication of the typical transit rider's perception of the overall trave!
experience.
LOS is also considered a performance measure. It is useful for describing
segment performance to elected officials, poliey makers, administrators, or the
public. LOS is based on the transit LOS score.

Limitations of the Methodology

In general, the methodology can be used to evaluate the performance of most
urban street segments. However, jt does not address all conditions or types of
control. The inability to replicatc thc infiucnce of a condition or control type in
the methodology is a limitation.
This subsection identifies the known limitations of the transit mcthodology.
lf one or more of these Iimitations are believed to have an important influence on
the performance of a specific street segment, the analyst should consider using
alternative methods or tools for the evaluation.
The transit methodology does not account for the eHect of the following
conditions on the quality of service provided to transit passengers:
• Prcsence of railroad crossings, and
• Transit vehicles on grade-separated or non-public-street rights-of-way.

Transit Methodology Chapter 18/Urban Street 5egments

Page 18-66 Version 6.0
 Highway Capacity Manual: A Guide far Multimodal Mobility Analysis

Procedures for estimating transit vehide performance on grade-separated or

non-public-street rights-of-way, along with procedures for estimating origin-
destination service quality, are provided in the Trmlsít Capacity alld Q1Jalityo/
SeroiceMallllal (7).

REQUIRED DATA ANO SOURCES

This subsedion describes thc input data needed for the transit methodology.
The required data are listed in Exhibit 18-25. They must be separately specified
for each dircction of travel on the segment and for each boundary intersection.
The exhibit also lists default values that can be used if local data are not available
(2,3).

Reguired Data and Units Potential Source(s) Suggested Default Value Exhibit 18-25
Trame Characterlstics Required Input Data, Potential
Dwell time (s) Data Sources, and Default
Field data, AVL data 60 s (downtown stop, transit
Values for Transit Analysis
eenter, major on-line transfer
point, major park-and-ride)
30 s (major outlying stop)
15 s (typical outlying stop)
Exeess wait time (min) Field data, AVl data 5ee discussion in text
Passenger trip length (mi) National Transit Database 3.7 mi
Transit frequency (vehjh) Transit schedules Must be provided
Passenger load factor (pfseat) Field data, APC data 0.80 pfseat
Geometric Data
5egment len9th. (ft) Fleld data, aerial photo Must be provided
Other Data
CBD of 5-million-plu5 metro area (ensus data Must be provided
(yeslno)
Traffie signal effedive green-to-cyele- Field data or HCM method Must be provided (if present)
length ratio (decimal) output
Traffie signal cyele length (s) Field data or HCM method Must be provided (if present)
output
Transit stop location (nearsidejother) Fietd data, aerial photo Must be provided
Transit stop position (on.lineloff-line) Field data, aerial photo Must be provided
Proportion af transit stops with Fiejd data, transit faeility Must be provided
shelters (decimal) inventory
Proportion of transit stops with Field data, transit faeility Must be provided
benches (decimal) inventory
Performance Measures
Motorized vehicle running time. (s) HCM method output Must be provided
Pedestrian lOS score far link (decimal) HCM method output Must be provided
Reentry delay (slveh) HCM method output Must be provided
Roundabout volume-to-capacity ratio HCM method output Must be provided (if present)
(decimal)
Notes: AVL = alltomatic vehicle locabclrI, APC = automabc paS5el1ger counter, CBD = central business district .
• Also used o. calculated by the motOriled vehicle methodology.

Thc data elements Usted in Exhibit 18.25 do not indude variables that are
considered to represent calibration factors. A calibration factor typically has a
relatively narrow rangc of reasonable values or has a small impad on the
accuracy of the performance estimates. The recommended value for each
calibration factor is identified at relevant points in the presentation of the
methodology.

Qlapter lBjUrban 5treet 5egments Transit Methodology

Version 6.0 Page 18-67
Highway Capacity Manual: A Guide tor Multimodal Mobility Analysis

Traffie Characteristics Data

This subsection describes the traffic eharacteristics data listed in Exhibit 18-
25. These data describe the transit traffic streams traveling along thc segment
during the analysis periodo If there are multiple transit routes 00 the segment, the
transit-related variables are needed for each route.

Dwell Time
Dwell time is the time that the traosit vehiele is stopped at the curb to serve
passenger movements, ineluding the time required to open and elose the doors.
It docs not inelude time spent stopped after passenger movements have ceased
(e.g., waiting for a traffie signal or waitiog for a gap in traffie to reenter the travel
lane). Dwell times are typically in thc range of 10 to 60 s, depending on boarding
and alighting demando Proeedures for measuring and estimating dwell time are
provided in the Transit Capacity and Qualify ofService Manual (7).

Excess Wait Time

Transit reliability is measured by excess toait time, the average number of
minutes passengers must wait at a stop past the scheduled departure time. lt is
measured in the field as the sum of the differenees between the scheduled and
actual departure times at the preceding time point, divided by the number of
transit vehiele arrivals. Eariy departures from the preceding time point are
treated as the transit vehiele being one headway late, sinee a passenger arriving
at the stop by the scheduled departure time would have to wait one headway for
the next transit vehiele. If time point-specific excess wait time information is not
available but oo-time performance (e.g., percentage of departures from a time
point Oto 5 min late) data are available for a route, the methodology provides a
procedure for estimating excess wait time from on-time performance.
The scheduled departure time from a stop and the scheduled travel time for
a trip set the baseline for a passenger's expeetations for how long a trip should
take. If the transit vehiele departs late-or worse, departs before the scheduled
time (Le., before al1 the passengers planning to take that vehiele have arrived at
the stop)-the trip willlikely take longer than planned, which negatively affects
a passenger's perceptions of the quality of service.

Passenger Trip Length

For most purposes, the average trip length can be determined from National
Transit Database data for the transit agency (8) by dividing total passenger miles
by total unlinked trips. However, if an analyst has reason to believe that average
trip length on a route is substantially different from the system average, a route-
specific value can be determined from automatic passenger counter data or
National Transit Database count sheets for the route by dividing total passenger
miles by the total number of boarding passengers.
The impact of a late transit vehiele departure on the overall passenger speed
for a trip (as measured by using scheduled departure time to actual arrival time)
depends on the length of the passenger's trip. For example, a departure 5 min
late has more of a speed impact on a l-mi-Iong trip than on a 10-mi-long trip.

Transit Methodology Chapter lBJUman 5treet 5egments

Page lB-68 Version 6.0
 Highway Capacify Manual: A Guide tor Multimodal Mobitify Analysis

Average passenger trip length is used to determine the impact oi late departures
on overaH trip speed.

Transit Frequency
Transit frequency is defined as the count of scheduled fixed-route transit
vehicles that stop on or near the scgment during the analysis periodo It is
expresscd in units of transit vehicles per hour.
Scheduled transit vehides can be considered "local" or "nonlocal." Local
transit vehicles make regular stops along the street (typically every 0.25 mi or
less), although they do not necessarily stop within the analysis segment when
segment lengths are short or when stops alternate between the near and far sides
of boundary intersections. They are always counted, regardless of whether they
stop within the subject segmento Nonlocal transit vehicles operate on routes with
longer stop spadng than local routes (e.g., limited-stop, bus rapid transit, or
express routes). They are only counted when they stop within the subject
segment.

Passenger Load Fador

The load factor is the number of passengers occupying the transit vehicle
divided by the number of seats on the vehide. lf the number oi passengers
equals the number of seats, the load factor equals 1.0.This factor should be
measured in the field or obtained from the ageney serving the transit route. 1tis
an average value for aHof the scheduled fixed.route transit vehicles that travel
along the segment during the analysis periodo

Geometric Design Data

This subsection describes the geometric design data listed in Exhibit 18-25.
These data describe the geometric elements that influence the service provided to
transit passengers. Al! input data should be representative of the segment for its
entire length. An average value should be used for each element that varies
along the segment.
Segment length is the onIy variable in this category. Jt is defined in a
similarly titled section for the motorized vehicle methodology.

Other Data
This subsection describes the data listed in Exhibit 18-25that are categorized
as "other data."

Area Type
Area type describes the environment in which the subject segment is located.
This data element is used in the transit methodology to set a baseline for
passenger expectations of typical transit travel speeds. For this application, it is
sufficicnt to indicate whether the area type is a "central business district of a
metropolitan area with over 5 miHion persons" or "other."

Chapter 18/Urban Street Segments Transit Methodology

Vt'lWn 6.0 Page 18-69
Highway Capacify Manual: A Guide for Mulfimodal Mobilify Analysis

£ffeetive Green-to-Cycle-Length Ratio and eye/e Length

The eycle length and the effective green-to-eycle-Iength ratio for the through
movement are used in the transit methodology when the boundary intersection
is a traffic signa!. If the signal is actuated, the motorized vehicIe methodoJogy in
Chapter 19 can be used to estimate the average green-to-cycle-Iength ratio and
cycle length.

Transit Stop Location

This input describes whcther a transit stop is located on the near side of a
boundary intersection or elsewhere. A portian of the time required to serve a
near-side transit stop at a boundary intersection may overlap with the control
deJay incurred at the intersection.

Transit Stop Position

Transit stops can be either on-line, where the bus stops entirely or mostly in
the travellane and does not have to yield to other vehicles on exiting the stop, or
off-lille, where the bus puUs out of the travellane to serve the stop and may have
to yield to other vehicles on exiting.

Proportion of Stops with Shelters and with Benches

These two input data elements describe the passenger amenities provided at
a transit stop. A sheltered stop provides a structure with a roof and three
enclosing sides that protect occupants from wind, rain, and sun. A shelter with a
bench is counted twice, once as a shelter and a second time as a bench.

Performance Measures
This subsection describes the data Jistcd in Exhibit 18-25that are categorized
as "performance measures."

Motorized Vehie/e Running Time

The motorized vehicle running time for the segment is obtained from the
motorized vehicIe methodology that is described in Scction 3.

Pedestrian LOS SCore far Link

The pedestrian LOS score for the link is obtained from the pedestrian
methodology that is described in Section 4.

Reentry De/ay
The final component of transit vehicle stop delay is the reentry delay, the
time (in seconds) a transit vehicIe spends waiting for a gap to reenter the
adjacent traffic stream. Reentry deJay is estimated as follows (7):
• Reentry deJay is zero at on-line stops.
• At off-line stops away from the influence of a signalized intersection
queue, reentry delay is estimated from the procedures of Chapter 20,
Two-Way STop-Controlled Intersections, as if the bus were making a right
turo onto the link, but a critical headway of 7 s is used to account for the
sJower acceleration of buses.

Transit Methodology O1apter 18/Urnan Street 5egments

Page 18-70 VersiorI 6. o
 Hjghway Capadty Manual: A Guide tor Mu/tjmodal MobUity Anolysis

• At an off-line bus stop Joeated within the influenee of a signalized

intersection queue, reentry deJay is estimated from the queue service
time, g" by using the motorized vehiele methodology in Chapter 19,
Signalized Intersections.
Reentry deJay can be reduccd by the presence of yield-to-bus Jaws or
placards (and motorist compliance with thcm), the existence of an acceJeration
lane or queue jump departing a stop, or a higher-than-normal degree of bus
driver aggressiveness in forcing buses back into thc traffie stream. Analyst
judgment and local data can be used to make appropriate adjustments to reentry
deJay in these cases.

Volume-to-capaCity Ratio (If Roundabout)

If the boundary intersection is a roundabout and it has a near-side transit
stop, thc volume-to-capacity ratio for thc rightmost Jane of the segment approaeh
to the roundabout is needed. It is obtained from the Chapter 22 methodology.

OVERVIEW OF THE METHODOLOGY

This subsedion provides an overview of the methodology for cvaluating the
performance of an urban street segment in terms of its service to transit
passengers.
The transit methodology is applied through a series of seven steps that
eulminate in the determination of segment LOS. Thesc steps are iIlustrated in
Exhibit 18-26. Performance measures that are estimated inelude transit travcl
speed along thc strcet, transit wait-ride score, and a LOS seore reflective of al!
transit service stopping within or near thc segmento

J Exhibit 18-26
TransitMethodology
forUrtan
Step 1: DetermineTransitVehide Step 4: DetermineTransitWaít-Ride 5treet 5egments
RunningTime seo,.
1 I
Step 2: DetermineDelayat Step 5: DeterminePedestrianLOS
Intersectien Score fer Link
j I
Step 6: DetermineTransit LOS
Step 3: DetermineTravelSpeed
Scorefer Segment
I j
Step 7: Dete"mineLOS

COMPUTATlONAL STEPS

Step 1: Determine Transit Vehicle Running Time

There are two principal eomponents of the transit vehide's segment running
time. Onc is the time required to traveJ the segment without stopping. (To allow
direct comparison with automobile segment speeds, transit vehides are treated
as if they travel the cntire segment, even if they join midlink.) The seeond is the

Chapter18fUrbanStreet5egments TransitMethodology
Version 6.0 Page18-71
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

deJay incurred at the transit stops that are provided on the link. The following
subparts to this step describe procedures that are used to calculate these
components. They culminate with a subsection that describes the calculation of
transit vehiele segment running time.

A. Compute Segment Running Speed

Transit vehiele segment running speed is the speed reached by the vehiele
when it is not influenced by the proximity oi a transit stop or traffie control
device. This speed can be eomputed with Equation 18-48, whieh is derived from
tables given in a Transit Cooperative Researeh Program report (9).

Equatlon 18.48 SR( = min (SR' 1 + e-LOO :~1,185 Hui L) )

where
511.1 transit vehicle running speed (mi/h),
L segment length (ft),
N" number of transit stops on the segment for the subjcct route (stops),
SR motorized vehiele running speed ""(3,600 L)/(5,280 tR) (mi/h), and
tR segment running time (s).
The segment running time is computed by using Equation 18-7 in Step 2 oi
the motorized vehiele methodology.

8, Compute De/ay due to a Stop

The delay due to a transit vehicle stop for passenger pickup ineludes the
iollowing components:
• Acceleration-deceleration deJay,
• DeJay due to serving passengers, and
• Reentry delay.
This procedure is applied once for each stop on the segmento The deJay due
to eaeh stop is added (in a subsequent step) to compute the total delay due to aH
stops on thc segmento

Acceleration-Deceleration De/ay
Aceeleration-deeeleration delay is the additional time required to deceJerate
to stop and then accelerate back to the transit vehiele running speed SRI' It is
eomputed with Equation 18-49 and Equation 18.50.
5.280(5,,) ( 1 1)
Equatlon 18-49 daa = 3600""2
,
r+r
at
fad
dt
with
LOO (staps nat on the near side of a boundary intersection)
0.00 (near.side stops at all-way and major-street two-way STOP-
Equation 18-SO fad = eontrolled intersections)
{ 1-x (near.side stops at roundabouts)
g/e (near.side stops at trame signals)

Transit MetI1oc1ology Chapter 18/Urban 5treet 5egments

Page 18-72 V~6.0
 Highway Capacity Manual: A Guide tor Multimodal Mobility Analysis

where
dod transit vehicle acceleration--deceleration deJay due to a transit stop (s),

'.1 transit vehicle acceleration rate = 3.3 (ft/s2),

'dI transit vehicle deceleration rate = 4.0 (ftfs2),

fOIl proportion of transit vehicle stop acceleration-deceleration deJay not

due to traffic control,
x =- volume-to-capacity ratio af the Iink's rightmast Jane on a roundabout
approach,
g effective green time (s), and
C cycle length (s).
Acceleration--deceJeration deJay represents travel time that is in excess 01 that
required to traverse the equivalent distance at the running speed. It is incurred
when the transit vehicle stops solely because of a transit stop. When a transit stop
is 10cated on the near side of a boundary intersection, a transit vehicle might
need to stop anyway due to the traffic control. In this situation, acceJeration-
deceleration delay is already included in the through delay estimate (addressed
in a subsequent step) and should not be included in d.,¡. Equation 18-50 is used to
determine the proportion of dad incurred solely beeause of a transit stop.
If representative acceleration and deceleration rates are known, they should
be used in Equation 18-49.lf these rates are unknown, an acceleration rate of 3.3
ft/s2 and a deccleration rate of 4.0 ft/s2 can be used (7).

De/ay due to 5erving Passengers

The delay due to serving passengers is based on the average dwell time,
which is an input to this procedure. At signalized intersections, a portion of the
d\\'elI time may overlap time the transit vehicle would have spent stopped
anyway due to the traffie control. Equation 18-51 is used to compute the deJay
due to serving passengers.
Equation 18-51

where
dp' transit vehicle delay due to serving passengers (s),
Id average dwell time (s), and
fd' proportion of dwell time occurring during effective green (= g/C at
near-side stops at signalized intersections and 1.00 otherwise, where g
and C are as previously defined),

Reentry De/ay
The final eomponent of transit vehicle stop deJay i5 the reentry delay d",
which is an input to this procedure. Guidance lor estimating reentry delay is
provided in the Required Data and Sourees section.

Olapter l8/Urnan 5treet 5egments Transit Methodology

Version 6.0 Page 18-73
Highway Capacity Manual: A Guide for Multimodal MobUity Analysis

De/ay due to a Stop

Oelay due to a transit stop is the sum of acceleration-deceleration delay,
passenger service time delay, and reentry delay. lt is eomputed with Equation
18-52.

dts = dad + dps + dre

where dI> is the delay due to a transit vehicle stop (s), d" is the reentry delay (s),
and aH other variables are as previously defined.

e Compute Segment Running n"rne

Equation 18-53is used to compute transit vehicle running time, which is
based on segment running speed and delay due to stops on the segmento
H"
3.600 L "
Equation 18-53 tRt = 5 280 S + L dts.i
, Rt i=l

where tRI is the segment transit vehicle running time (s), d••.¡ is the delay due to a
transit vehide stop for passenger pickup at stop j within the segment (s), and aH
other variables are as previously defined.
If there are no stops on the segment, the seeond term of Equation 18-53
equals zero.

5tep 2: Determine Delay at Intersection

The through delay dI incurred at the boundary interseetion by the transit
vehicle is determined in this step. This delay is equal to the control deJay
incurred by through vehicles that exit the segment at the downstream boundary
intersection. Guidanee for dctermining this delay is provided in Step 5 of the
motorized vehicle methodology.
Alternatively, Equation 18.54 can be used to estimate the through delay due
to a traffie signal (9). This estimate is suitable for a planning-level analysis.

Equation 18-54 dt = tj 60(5,~80)

where
d¡ through delay (s/veh),
ti transit vehicle running time loss (min/mi), and
L segment length (ft).
The running time loss tI used in Equation 18-54is obtained from Exhibit 18-27.

Transit Methodology Chapter lB/Uman 5treet 5e9ments

Page 18-74 Version 6.0
 Highway Capadty Manual: A Guide for Multimodal Mobi/ity Anolysis

Runninq Time Loss by Signal Co;o;'~dWit~¡O~,g(~m~;~'~/m~ili"

Exhibit 18.27
Transit Vehicle Running Time
Signals More
Transit Frequent L=
Area la,e Traffic Signals 5et Than Transit
T Allocation Condition T ical for Transit StODS
No right
1.2 0.6 1.5-2.0
tum,
central With right-
Exclusive 2.0 lA 2.5-3.0
business tum delay
district Blocked by
2.5-3.0 Not available 3.0-3.5
traffie
Mixed traffie A, 3.0 Not available 3.5-4.0
Exclusive A, 0.7 0.5-1.0 Not available Not available
OOe,
Mixed traffie A, 1.0 0.7-1.5 Not available Not available
Sollrte: St. JKQues and Levinsoo (9).

Step 3: Determine Travel Speed

Transit travel speed is .ln aggregate me.lsure of speed along the street. lt
combines thc dclay incurred at the downstream intersection with the segment
running time. Thus, it is typically slower than the running specd. The transit
travel spced is computed by using Equation 18-55.
3.600 L
Equation 18.55
Sn.seg = 5,280 (tRI + dt)
where Sn."i: is the travel speed of transit vehicles .llong thc scgment (milh), tRt is
the segment running time of transit vehicles (s), and all other variables are as
previously dcfined.

Step 4: Determine Transit Wait-Ride Score

The transit wait-ride score is a performance measure that combines
perceived time spent waiting for the transit vehicle and perceived travel time
rate. If transit servicc is not provided for the subject direction of travcl, this score
equals 0.0 and the analysis continues with Step 5.
The procedurc for calculating the wait-ride score is descr¡bed in this step. It
consists of the separate calculation of the headway factor and the perceived
travel time factor. The following subsections describe these two calculations,
which culmina te in the calculation of the wait-ride score.

A. Compute Headway Factor

The headway factor is the ratio of the estimatcd patronage at the prevailing
average transit headway to the estimated patronage at a base headway of 60 mino
The patronage values for the two hcadways (i.e., the input hcadway and thc base
headway of 60 min) are computed from an assumed set of patronage elasticities
that relate the percentage change in ridership to the pcrcentage change in
headway. The he.ldway factor is computed by using Equation 18-56.
Fh = 4.00 e-1.434/(lIs+0.001) Equation 18.56

where
Fh headway factor, and

v. transit frcqucncy for thc segment (veh/h).

Chapter 18{Urtan $treet Segments Transit Methodology

verwn6.0 Page 18-75
Highway Capacity Manual: A Guide for Multimodal Mobilify Analysis

The transit frequency v, is an input to this procedurc. Guidance for

estimating this input is provided in the Reguired Data and Sources section.

8. Compute Perceived Trave! Time Factor

Segment performance, as measured by the wait-ride score, is influenced by
the travel time rate provided to transit passengers. The perceptibility of this rate
is further influenced by the extent to which the transit vehicle is late, crowded, or
both and whether the stop provides passenger amenities. In general, travel at a
high rate is preferred, but travel at a lower rate may be nearly as acceptable if the
transit vehicle is not late, the bus is lightly loaded, and a shelter (with a bench) is
provided at the transit stop.
The perceived travel time factor is based on the perceived travel time rate
and the expected ridership elasticity with respect to changes in the perceived
travel time rateoThis factor is computed with Eguation 18-57.

Ce - 1) Tbtt - Ce + 1) Tptt
Equation 18-57 Ftt = ----------
Ce - 1) Tptt - Ce + 1) Tbtt
with

Equation 18-58 Tptt = (a, S 60 ) + (2 Tex) - Tat

Tt,seg

1.00 F, S; 0.80

4 (F, - 0.80)
1 + ----- 0.80 S; F¡ S; 1.00
al = 4.2
4 (F, - 0.80) + (F, -1.00)[6.5 + 5 (F, - 1.00)]
1 + 4.2 F F¡ > 1.00
l

1.3 Psh
Tat = -------
+ 0.2 Pbe
Equation 18-60
Lp'
where

F" perceived travel time factor;

, = ridership elasticity with respect to changes in the travel time rate
= -0.40;

base travel time rate = 6.0 for the central business district of a
metropolitan area with 5 million persons or more, otherwise = 4.0
(min/mi);
perceived travel time rate (min/mi);
excess wait time rate due to late arrivals (min/mi) = t.jLp,;
excess wait time due to late arrivals (min);

T., amenity time rate (min/mi);

a, passenger load weighting factor;
travel speed of transit vehicles along the segment (mi/h);

Transit MethodoIogy O1apter 18/Urban 5treet Segments

Page 18-76 v~6.0
 Highway Capacity Manual: A Gu;de for Multimodal Mobility Analysis

F, average passenger load factor (passengers/seat)¡

Lp, average passenger trip length = 3.7 typically (mi);
P_<II proportion of stops on segment with shelters (decimal)¡ and
PI>< proportion of stops on segment with benches (decimal).
The perceived travel time rate is estimated according to three components, as
shown in Equation 18-58.The first component reflects the avcrage travel speed of
the transit service, adjusted for the degree of passenger loading. The second
component reflects the average excess wait time for the transit vehiele (Le.,the
amount of time spent waiting for a late arrival beyond the scheduled arrival
time). The third component reflects the ability of passengers to tolerate longer
travel time rates when amenities are pravided at the transit stops.
The first term in Equation 18-58ineludes a factor that adjusts the transit
vehiele travel time rate by using a passcnger load weighting factor. This factor
accounts for the decreasc in passenger comfart when transit vehieles are
crawded. Values of this factor range fram 1.00when the passenger load factor is
less than 0.80 passengers/seat to 2.32 when the load factor is 1.6 passengers/seat.
The second term in Equation 18-58represents the perceived excess wait time
rate. It is based on the excess wait time fa associated with late transit arrivals.
The multiplier of 2 in Equation 18-58is used to amplify the exeess wait time rate
bccause passcngers perceive excess waiting time to be more oneraus than actual
travel time.
The excess wait time la reflects transit vehiele rcliability. It is an input to this
pracedure. If excess wait time data are not available for a stop but on-time
performance data are available for rautes using the stop, Equation 18-61may be
used to estimatc the average excess wait time.
tex = [t¡ate(1- Pot)]2 Equation 18-61

where
ta excess wait time due to late arrivals (min),
ti.!, threshold late time = 5.0 typical (min), and
POI proportion of transit vehicles arriving within the threshold late time
(default = 0.75) (decimal).
The third term in Equation 18-58represents the amenity time rate reduction.
This rate is computed in Equation 18-60as the equivalent time value of various
transit stop improvements divided by the average passenger trip length. If
multiple transit stops exist on the scgment, an average amenity time rate should
be uscd for the segment, based on the average value far all stops in the segment.
The average passenger trip ¡ength is used to convert time values for excess
wait time and amenities into distance-weighted travel time rates that adjust the
perceived in-vehicle travel time rateoThe shorter the trip, the greater the
influence that late transit vchielcs and stop amenitics have on thc oYerall
perceived speed of the trip.

O1apter lB/Urban Street 5egments Transit Methodology

Version 6.0 Page 18.77
Highway Capacity Manua/.' A Guide for Multimodal Mobility Analysis

The average passenger trip length should be representative of transit routes

using the subjeet segment. A value of 3.7 mi is considered to be national1y
representa tive. More accurate local values can be obtained from the National
Transit Database (8). Specifieally, this database provides annual passenger miles
and annual unlinked trips in the profile of most transit agencies. The average
passenger trip length is eomputed as the annual passenger miles divided by the
annual unlinked trips.

C. Compute Wait-Ride SCOre

The wait-ride seore is eomputed with Equation 18-62.A larger seore
eorresponds to better performance.
Equation 18-62 Sw.r = Fh Frt

where
sU'-, transit wait-ride seore,
Fh headway factor, and
Fu pereeived travel time factor.

Step S: Determine Pedestrian LOS SCore for Link

The pedestrian LOS seore for the link 11',1in~ is computed by using the
pedestrian methodology, as deseribed in Seetion 4.

Step 6: Determine Transit LOS Score for Segment

The transit LOS seore for the segment is eomputed by using Equation 18-63.
Equation 18-63 lt,seg = 6.0 - 1.50 Sw.r + 0.15 lp,1ink
where 11.~is the transit LOS seore for the segment and aHother variables are as
defined previously.

Step 7: Determine LOS

The transit LOS is determined by using the transit LOS score from Step 6.
This performance measure is eompared with the thresholds in Exhibit 18-3 to
determine the LOS for the specified direction of travel along the subjeet street
segment.

Transit Methodology Chapter 18jUrnan 5treet 5egments

Page 18-78 Verron 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

7. APPLICATIONS

EXAMPlE PROBlEMS
Chapter 30, Urban Strect Segments: SupplemcntaL describes the application
of each of the four methodologies through the use of example problems. There is
one example problem associated with each methodology. Thc examples iIIustrate
the operational analysis type.

GENERAUZEO OAllY SERVICE VOLUMES

Generalized daily service volume tables provide a means of quickly
assessing one or more urban street facilities to determine which facilities need to
be more carefully evaluated (with operational analysis) to ameliorate existing or
pending problems. Their application in practice is typically at the facility level
rather than at the segment leve!. For this reason, service volumc tables are
provided in Chapter 16, Drban Street Facilities.

ANAL YSIS TYPE

The four methodologies described in this chapter can each be used in three
types of analysis. The analysis types are described as operational, design, and
planning and preliminary engineering. The selected analysis type applies to thc
methodology described in this chapter and to all supporting methodologics. The
characteristics of each analysis type are described in the subsequent paragraphs.

Operational Analysis
The objective of an opcrational analysis is to determine the LOS for current
or near-term conditions when details of traffic volumes, geometry, and traffic
control conditions are known. AH the methodology steps are implemented aod
aHcalculation procedures are applied for the purpose of computing a wide range
of performance measures. The operational analysis type will provide the O1ost
reHable results because it uses no (or minimal) default valucs.

Design Analysis
The objective of the design analysis is to identify the alternatives that operate
at the target level of thc specified performance measures (or provide a beUer
level of performance). The analyst may then recommend the "best" design
alternative after consideration of the full range of factors.
The nature of the design analysis type depends on whether the boundary
intersections are unsignalized or signalized. When the segment has unsignalized
boundary intersections, the analyst spccifies traffic conditions and target levcls
for a set of performance measures. The methodology is then applicd by usiog an
iterative approach in which alternativc geometric conditions are scparately
cvaluated.
When the segment has signalized boundary intersections, the design analysis
type has two variations. 80th variations require the specification of traffic
conditions and target levels for a set of performance measures. Dne variation
requires thc additional specification of the signalization conditions. The

Chapter 18jUrban 5treet Segments Apptications

V~ion 6.0 Page 18-79
Highway Capadty Manual: A Guide for Multimodal Mobility Analysis

methodology is then applied by using an iterative approach in which alternative

geometric conditions are separately evaluated.
The second variation of the design analysis requires the additional
specification of geometric conditions. The methodology is then applied by using
an iterative approach in which alternative signalization conditions are evaluated.

Planning and Preliminary Engineering Analysis

The objective of a planning and preliminary engineering analysis can be (a)
to determine the LOS for either a proposed scgment or an existing segment in a
future year or (b) to size the OYerallgeometrics of a proposed segmento
The level of precision inherent in planning and preliminary engineering
analyses is typically lower than for operational analyses because default values
are often substituted for field.measured values of many of the input variables.
Recommended default values for this purpose were described previously in the
section associated with each methodology.
The requirement for a complete description of the signal timing plan can be a
burden for sorne planning analyses involving signalized intersections. The
intersection planning-Ievel analysis application described in Chapter 31,
Signalized Intersections: Supplemental, can be used to estimate a reasonable
timing plan, in conjunction with the aforementioned default values.
For some planning and preliminary engineering analyses, the segrnent
planning-Ievel analysis application described in Chapter 30, Urban Street
Segments: Supplemcntal, may provide a better balance between accuracy and
analysis effort in the evaluation of vehicle LOS.

USE OF ALTERNATlVE TOOLS

General guidance for the use of alternative traffic analysis tools for capacity
and LOS analysis is provided in Chapter 6, HCM and AIternative Analysis Tools,
and Chapter 7, Interpreting HCM and Alternative Tool Results. This subscction
contains specifie guidance for the application of aIternative tools to the analysis
oí urban street segments. Additional information on this topie is provided in the
Technieal Reference Library in Volurne 4. The forus of this subsection is the
application oí alternative tools to evaluate motorized vehicle operation.

Comparison of Motorized Vehicle Methodology and Alternative Tools

Motorized Vehiele Methodology
The motorized vehicle methodo!ogy rnodels the driver-vehicle-road system
with reasonable accuracy for most applications. It accounts for signal
coordination, platoon dispersion, the origin-destination patterns of all segment
traffic flows, driveway impacts on traffic flow, and the influence of volume on
speed.
The motorized vehic1e methodology offers severa! advantages over
alternative analysis tools. One advantage is that it has an empirically calibrated
procedure íor estimating saturation flow rateo Alternative tools often require
saturation flow rate as an input variable. A second is that it produces a direct
estimate oí capacity and vo!ume-to-capacity ratio. These measures are not

Applications O1apter 1a/Uman 5treet Segments

Page 18.80 Vt'lSOn 6.0
 Highway Capacity Manual: A Guide far Multimodal Mobility Analysis

directly available from simu1ation tools. A third advantage is that it produces an

expectcd value for each of several performance measures in a single application.
Simulation tools require mulliple runs and manual calculations to obtain an
expected value for a given performance measure. A fourth is that its analytic
procedures are described in the HCM so that analysts can understand the driver-
vehicle-road interactions and the means by which they are modeled. Most
proprietary alternative tools operate as a "black box," providing Hule detail
describing the intermediate calculations.

Alternative Tools
80th deterministic tools and simulation tools are in common use as
altematives to the motorized vehicle methodology offered in this chapter.
Deterministic tools are often used for the analysis of urban street segments. The
main reasons for their popularily are found in the user interface, optimization
options, and output presentation features. Sorne also offer additional
performance measures such as fuel consumption, air quality, and operating costo

Conceptual Differences
Alternative deterministic tools apply traffic models that are conceptually
similar to those described in this chapter. While their computational details will
usually produce different numerical results, there are few major conceptual
differences that would preclude comparison of commonly defined performance
measures.
Simulation tools, on the other hand, are based on entircly different modeling
concepts. A general discussion of the conceptual differences is presented in
Chapters 6 and 7. Some specific examples for signalized intersections, which also
apply to urban street segments, are presented in Section 7 of Chapter 19.
One phenomenon that makes comparison difficult is the propagation of
platoons along a segment. Deterministic tools, including the model presented in
this chapter, apply equations that spread out a platoon as it progresses
downstream. Simulation t001screate platoon dispersion implicitly from a
distribution of desired speeds among drivers. Both approaches will produce
platoon dispersion, but the amount of dispersion will differ among tools.
Simulation tools may also exhibit platoan compression because of the effect
of slower-moving vehicles that cause plataons to regenerate. For this and other
reasons, comparability of platoon representation along a segment between these
tools and the motorized vehicle methodology is difficult to achieve.

Alternative Tool Application Guidance

Development of HCM-Compatible Performance Measures
Altemative tools generally define travel spt-"Cdin the same way that it is
defined in this chapter. However, these tools may not compute dela)' and
running speed by using the procedures presented in this chapter. Therefore, care
must be taken in comparing spt->edand delay estimates from this chapter with
those from other tools. lssues related to the comparison of speed (or delay)
among different tools are discussed in more detail in Chapter 7. In general, the

Chapter 18/Urban 5treet 5egments Applications

Vers~6.0 Page 18-81
Highway Capadfy Manual: A Guide for Multimodal Mobilify Analysis

travel speed from an altcmative tool should not be used for LOS assessment
unless the tool is confirmed to apply the definitions and proccdurcs described in
this chapter.

Adjustment of Parameters
For applications in which either an altemative too! or the motorized vchiclc
methodology can be used, sorne adjustment will generally be required for the
a!temative tool if consistency with the motorized vehicle methodology is desired.
For examp!e, the parameters that determine the capacity of a signalized approach
(e.g., saturation flow rate and start-up lost time) should be adjusted to ensure
that the !ane group (or approach) capacitics from thc altemative tool match those
estimated by the motorized vehicle methodo!ogy.
Adjustment of the altemative tool parametcrs that affcct the travel time
along the segment might also be necessary to produce comparable results. The
motorized vehicle methodology is based on a free-now speed that is computed
as a function of demand flow rate, median type, access point density, parking
presence, and speed limito Most altemative tools typical1y require a user-
specified free-flow speed, which could be obtained from the motorized vehicle
methodology to maintain comparability. Adjustment of the platoon modeling
parameters may be more difficult. Thus, if comparability is desired in
representing the platoon cffect, it is prefcrable to adjust the free-flow speed
specified for simu!ation so that the actual trave! specds are similar to those
obtained from the motorized vehicle methodology.

Sample Calculations IlIustrating Alternative Tool Applications

Chapter 29, Urban Street Facilities: Supplemental, includes a set of examples
illustrating the use of a!temative too!s to address the stated limitations of this
chapter and Chapter 16, Urban Street Facilities. Specifically, the examples
illustrate (a) the application of deterministic tools to optimize signal timing, (b)
the effect of platooned arriva!s at a roundabout, (e) the effect of midsegment
parking maneuvers on facility operation, and (d) the use of simulated vehicle
trajectories to evaluate the proportion of time that the back of the queue on the
minor-street approach to a two-way STOP-control1cdintcrscction exceeds a
specified distance from the stop lineo
Chapter 31, Signalized lntersections: Supplemental, includes examp!e
problems that address left-tum storage bay overflow, righHurn-on-red
operation, short through lanes, and c10sely spaced intersection5.

Applications O1apter 18/Urban 5treet 5egments

Page 18-82 Version 6.0
 Highway Capacity Manual; A Guide for Mulfimodal Mobility Analysis

8. REFERENCES

1. Bonneson, J. A, M. P. Pratt, and M. A. Vandehey. Predicting the Performance oJ Some of tfrese references CiJn
be found in the Tedmic<1f
Automobi/e Traffic 011 Urball Streets: Filial Report. National Cooperative Reference Library in VoIume 4.
Highway Research Program Project 03.79. Texas Transportation lnstitute,
Texas A&M University, College Station, Jan. 2008.
2. Dowling, R. G., D. B. Reinke, A Flannery, P. Ryus, M. Vandehey, T. A
Petritsch, B. W. Landis, N. M. Rouphail, and J. A. Bonneson. NCHRP Report
616: Multimodal Leve! oJ Service Analysis Jor Urban Sfreets. Transportation
Research Board of the National Academies, Washington, D.C., 2008.
3. Zegeer, J. D., M. A. Vandehey, M. Blogg, K. Nguyen, and M. Ereti. NCHRP
Report 599: DeJault Values Jor Highway Capacity and Level oJ Service Analyses.
Transportation Research Board of the National Academies, Washington,
D.e" 2008.
4. RouphaiJ, N., J. Hummer, J. Milazzo, and D. AlIen. Capacity Analysis oJ
Pedesfrian and Bicycle Facilities: Recommeuded Procedures for the "Pedestrians"
Chapter oJ the Highway Capacity Manual. Report FHWA-RD-98-107. Federal
Highway Administration, Washington, D.C., 1998.
5. U.S. Department of Transportation. 2004 Status oJ the Natioll's Highways,
Bridges, alld Transit: Conditions alld Performance. Washington, D.C., 2004.
http://www.fhwa.dot.gov/policy/olsp/reportspubs.htm. Accessed Sept. 4,
2008.
6. Rouphail, N., J. Hummer, J. Milazzo, and D. AlIen. Capacity Analysis oJ
Pedestrian alld Bicycle Facilities: Recomme"ded Procedures for the "Bicycles"
Chapter oJ the Highway Capacity Manual. Report FHWA-RD-98-108. Federal
Highway Administration, Washington, D.C., 199B.
7. Kittelson & Associates, lne.; Parsons Brinckerhoff; KFH Group, lne.; Texas
A&M Transportation Institute; and Arup. TCRP Report 165: TrQ/lsit Capacity
and Quality oJServiee Manual, 3rd cd. Transportation Research Board of the
National Academies, Washington, O.c., 2013.
8. Federal Transit Administration. Anllual Data Publicafions. National Transit
Databasc. http://www.ntdprogram.gov/ntdprogram/. Accessed Sept. 1, 2008.
9. StoJacques, K, and H. S. Levinson. TCRP Report 26: Operatiollal Analysis oJ
Bus Lmes 0/1 Arferials. Transportation Research Board, National Research
Council, Washington, D.C., 1997.

Chapter lB/Urban 5treet 5egments References

Version 6.0 Page 18-83
 Highway Capacify Manual: A Guide tor Multimodal MobWty Analysjs

CHAPTER 19
SIGNALIZED INTERSECTIONS

CONTENTS

1. INTRODUCTION 19-1
Overvicw 19.1
Chapter Organization 19-2
Related HCM Content 19.2

2. CO NCEPTS 19.4
Traffic Signal Concepts 19-4
Analysis Type 19-14
Spatial and Temporal Limits 19.14
LOS Criteria 19-14
Scope of thc Methodologies 19-16

3. eORE MOTORIZED VEHIClE METHODOLOGY 19-18

Scope of the Methodology 19-18
Required Data and Sources 19-22
Overview of thc Mcthodology 19.40
Computational Steps 19-42

4. EXTENSIONS TO THE MOTORIZED VEHICLE METHODOLOGY ..... 19.58

Critical Intersection Volume-to-Capacity Ratio 19-58
Uniform Oelay Calculation Using Queue Accumulation Polygon 19-62
Initial Queue Delay Calculation 19-65

5. PEDESTRIAN METHODOlOGY 19-70

Scope of the Methodology 19-70
Required Data and Sources 19-72
Overview of the Methodology 19-76
Computational Steps 19.77

6. BICYCLE METHODOLOGY 19.86

Scope of the Methodology 19-86
Required Data and Sources 19-86
Overview of the Methodology 19-88
Computational 5teps 19-89

O1apter 19j5ignalized Intersections Contents

VersiOn 6.0 Page 19-i
Highway Capacity Manual: A Guide tar Multimadal Mobility Analysis

7. APPLlCA TlO NS .•••.••••••••.•••..••..•••.•••.••••••••.•••.••.......•••....•••••...........••••..•••............•

19-92
Example Problems 19-92
Generalized Daily Service Volumes 19-92
Analysis Type 19-93
Use of Altemative Tools 19-94

8. R EFER EN CES.................................................................................................. •••••19-97

Conteots Olapter 19j5ignalized Intersectlons

Page 19-11 V~6.0
 Highway Capacity Manual: A Guide far Mu/timodal Mobi/ity Analysis

LIST OF EXHIBITS

Exhibit 19-1 Intersection Traffie Movements and Numbering Scheme 19-5

Exhibit 19-2 Dual-Ring Structure with IIIustrative Movement
Assignmen ts 19-6
Exhibit 19-3 IIIustrative Protected Lag-Lag and Permitted-Only Phasing 19-8
Exhibit 19-4 lliustrative Protected Lead-Lag and Split Phasing 19-9
Exhibit 19-5 Fundamental Attributes of Traffie Flow at Signalized
Intersections 19-10
Exhibit 19-6 Fundamental Variables of Traffie Flow at Signalized
Interseetions 19-11
Exhibit 19-7 Time Elements Influencing Actuated Phase Duration 19-13
Exhibit 19-8 LOS Critena: Motorized Yehicle Mode 19-16
Exhibit 19-9 LOS Criteria: Pedestrian and Bicycle Modes 19-16
Exhibit 19-10 Three Alternative Study Approaehes 19-19
Exhibit 19-11 Required Input Data, Potentiai Data Sourees, and Default
Yalues for Motorized Yehicle Analysis 19-23
Exhibit 19-12 Required Additionallnput Data, Potential Data Sources,
and Default YaJucs for Motorized Yehicle Analysis with
Coord inated Signal Control.. 19-24
Exhibit 19-13 Relationship Between Arrival Type and Progression
Quali ty 19-26
Exhibit 19-14 Arrival Type Selection Guidelincs 19-28
Exhibit 19-15 DefauIt Lane Utilization Adjustment Factors 19-31
Exhibit 19-16 DefauIt Parking Maneuver Rate 19-32
Exhibit 19-17 Default System Cycle Length 19-38
Exhibit 19-18 Motorized Vehicle Methodology for Signalized
Interseetions 19-42
Exhibit 19~19Typical Movement Groups and Lane Groups 19-43
Exhibit 19-20 Lane Width Adjustment Factor 19-45
Exhibit 19-21 Cumulative Arrivals and Departures During an
Oversaturated Analysis Period 19-53
Exhibit 19-22 Critical Path Determination with Protected Left-Tum
Phases 19-60
Exhibit 19-23 Critical Path Determination with Protected.Permitted Left-
Tu m Operanon 19-61
Exhibit 19-24 Decomposition of Queue Accumulation Polygon 19-63
Exhibit 19-25 Initial Queue Delay with Inereasing Queue Size 19-66
Exhibit 19-26 Initial Queue Delay with Decreasing Queue Size 19-66
Exhibit 19-271nitial Queue Dclay with Queue Clearing 19-66

O1apter 19j5ignalized Intersections Contents

V~6.0 Page 19-iii
Highway Capacity Manual: A Guide for Mu/timodal Mobility Analysis

Exhibit 19-28 Qualitative Description of Pedestrian Space 19-71

Exhibit 19-29 Condition 1: Minor-Street Crossing 19-72
Exhibit 19-30 Condition 2: Major-Street Crossing 19-72
Exhibit 19-31 Required Input Data, Potential Data Sources, and Default
Values for Pedestrian Analysis 19.73
Exhibit 19-321ntersection Comer Geometry and Pedestrian Movements 19-75
Exhibit 19-33 Pedestrian Methodology for Signalized lntersections 19.77
Exhibit 19-34 Required Input Data, Potential Data Sources, and Default
Values for Bieycle Analysis l9-87
Exhibit 19.35 Bieycle Methodology for Signalized Intersections 19-89
Exhibit 19-36 Il1ustrative Generalized Service Volumes for Signalized
Intersections (veh/h) 19-92
Exhibit 19.37 Effect of Cycle Length on Delay 19-95

Contents Chapter 19j5ignalized Intersections

Page 19-iv version 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

1. INTRODUCTION

OVERVIEW VOLUME J: iNTERRUPTED flOW

16. Urban Street Facilities
This chapter describes separate methodologies for evaluating the operation 17. Urban Street Reliability and
ATDM
of each of the following intersection travel modes: motorized vehiele, pedestrian, 18. Urban Street Segments
and bicyc1e.Each methodology is used to evaluate the quality of service 19. Signalized Intersections
20. TWSC !ntersections
provided to road users traveling through a signalized intersection. A detailed 21. AWSC intersections
description of each travel mode is provided in Chapter 2, Applications. 22. Roundabouts
23. Ramp Terminals and
The methodologies are much more than just a means of evaluating quality of Alternative intersections
24, Off-Street Pedestrian and
service. They inelude an array of performance measures that fully describe Bicycle Facilities
intersection operation. These measures serve as elues for identifying operational
issues. They also provide insight into the developmcnt of effective improvement
strategies. The analyst is encouragcd to consider the full range of performance
measures associated with each methodology.
This chapter also describes methodologics for evaluating interscction
performance from the perspective of motorists, pedestrians, and bicyc1ists.These
methodologies are referred to as the motorized vehiele methodology, pedestrian
methodology, and bicyc1emethodology. Collectively, they can be used to
evaluate the intersection operation from a multimodal perspective.
Each methodology in this chapter focuses on the evaluation of a signalized
intersection. Chapter 18, Urban Street Segments, provides a methodology for
quantifying the performance of an urban street segment. The methodology
described in Chapter 16, Urban Street Facilities, can be used to combine the
performance measures (for a spccified travel rnode) on successive segments into
an overall measure of facility performance for that mode.
An intersection' s performance is described by the use of one or more
quantitative measures that characterize sorne aspect oE the service provided to a
specific road-user group. Performance measures cited in this chapter inelude
volume-to-capacity ratio, rnotorized vehiele control delay, pedestrian comer
circulation area, pedestrian del ay, pedestrian level-of-service (LOS) score, bicyele
delay, and bicyele LOS score.
The motorized vehic1e methodology has evolved and reflects the findings
from a large body of research. It was originally based, in part, on the results of a
National Cooperative Highway Research Program (NCHRP) study (1, 2) that
formalized the critical movement analysis procedure and the rnotorized vehicle
delay estimation procedure. The critical movement analysis procedure was
developed in the United States (3, 4), Australia (5), Great Britain (6), and Sweden
(7). The motorized vehic1cdelay estimation procedure was developed in Great
Britain (8), Australia (9), and the United States (10). Updates to the original
methodology were developed in a series of rescarch projccts (11-24).
The procedures for evaluating pedestrian and bicyelist perception of LOS are
documented in an NCHRP report (25).The procedures for evaluating pedestrian
delay, pedestrian circulation arca, and bicyelist delay are documented in two
Federal Highway Administration reports (26, 27).

Qlapter 19{5ignalized intersections Introduction

Verskrn 6,0 Page 19-1
Highway Capacity Manual: A Guide for Multimodal Mobility Ana/ysis

CHAPTER ORGANIZATION
Section 2 oi this chapter presents concepts used to describe signalized
intersection operation. It provides an overview oi traffic signal-timing and
phasing concepts. It also ineludes guidance for establishing the intersection
analysis boundaries and the analysis period duration. It coneludes with a
discussion of the service measures and LOS thresholds used in the methodology.
Section 3 presents the core methodology for evaluating motorized vehiele
service at a signalized intersection. The presentation describes the scope of the
methodology and its required input data. It coneludes with a description oi the
computational steps that are followed for each application of the methodology.
Section 4 describes extensions to the core motorized vehiele methodology,
ineluding the calculation of intersection volume-to-eapacity ratio, uniform delay
calculation, and initial queue delay calculation.
$ection 5 presents the methodology for evaluating pedestrian service at a
signalizcd intersection. The presentation ineludes a discussion of methodology
scope, input data, and computational steps.
$ection 6 presents the methodology for evaluating bicyele service at a
signalized intersection. The presentation ineludes a discussion of methodology
scope, input data, and computational steps.
Section 7 presents guidance on using the rcsults of the intersection evaluation.
The presentation ineludes example results from each methodology and a
discussion of situations in which altemative evaluation tools may be appropriate.

RELATED HCM CONTENT

Other Highway Capacity Manual (HCM) content related to this chapter
includes
• Chapter 18, Urban Street Segments, which describes concepts and
methodologies for the evaluation of an urban street segment;
• Chapter 29, Urban Street Facilities: SupplementaL which provides details
of the reliability methodology, a procedure for sustained spillback
analysis, information about the use oi alternative evaluation tools, and
example problems demonstrating both the urban street facility
methodologies and the reliability methodology;
• Chapter 30, Urban Street Segments: Supplemental, which describes
procedures for predicting platoon flow, spillback, and delay due to tums
from the major street; a planning-Ievel analysis application; and example
problems demonstrating the urban street segment methodologies;
• Chapter 31. Signalized intersections: Supplemental, which describes
procedures for predicting actuated phase duration, lane volume
distribution, queue length, and saturation flow adjustment factors for
pedestrian, bicyele, and work zone presence; a planning.level analysis
application; and example problems demonstrating the signalized
intersection methodologies;

Intnxluctlon O1apter 19/5ignalized Intersections

Page 19-2 Ver5kln 6.0
 Highway Capacity Manual: A Guide tor Multimodal Mobility Analysis

• Case 5tudy 1, U.5. 95 Corridorj Case 5tudy 2, Route 146 Corridorj and
Case Study 3, Krome Avenue, in the HCM Applicafions Cuide in Volume 4,
which demonstrate how this chapter's methods can be applied to the
evaluation of actual signalized intersections; and
• Section L, Signalized lntersections, in Part 2 of the Planning and
Preliminary Engineering Applicalions Cuide lo ¡he HCM, found in Volume 4,
which describes how to incorporate this chapter's methods and
performance measures into a planning effort.
A procedure for detennining intersection saturation flow rate when a work
zone is present upstream (or downstream) of the intersection is provided in the
final report for NCHRP Project 03-107, Work Zone Capacity Methods for the
HCM. This report is in online Volume 4.

OIapter 19j5ignalized Intersections Introduction

Ver.sion 6.0 Page 19-3
Highway Capadty Manual: A Guide for Multimodal Mobility Analysis

2. CONCEPTS

This section presents concepts used to describe signalized intersection

operation. The first subsection describes basic signalized intersection concepts,
induding traffic signal control, the traffic movement numbering scheme, phase
sequence and operational modes, intersection traffic flow characteristics, and
phase duration components. The second identifies the types of analysis that can
be conducted. The third indudes guidance for establishing the intersection
analysis boundaries and the analysis period duration. The fourth discusses the
service measures and LOS thresholds used in the methodology, and the last
subsection identifies the scope of the collective set of methodologies.

TRAFFIC SIGNAL CONCEPTS

Types of Traffie Signal Control
In general, two types of traffic signal controller are in use today. They are
broadly categorized as pretimed or actuated according to the type of control they
provide. These two types of control are described as follows:
• Prefimed COlltrol consists of a fixed sequence of phases that are displayed
in repetitive order. The duration of each phase is fixed. However, the
green interval duration can be changed by time of day or day of week to
accommodate traffic variations. The combination of a fixed phase
sequence and fixed duration produces a constant cyde length.
• Actuated colltrolconsists of a defined phase sequence in which the
presentation of each phase depends on whether the phase is on recall or
the associated traffic movement has submitted a call for service through a
detector. The grecn interval duration is determined by the traffic demand
information obtained from the detector, subject to preset minimum and
maximum Iimits. The termination of an actuated phase requires a eaH for
service from a conflieting traffic movement. An actuated phase may be
skipped if no demand is detected.
Most modero control1ers have solid-state components and use software to
implement the controllogic. This architeeture is sufficiently flexible to provide
either actuated control or pretimed control.
The operation of a pretimed controller can be described as eoordinated or not
coordinated. In contrast, the operation of an actuated controller can be described
as fully actuated, semiactuated, or coordinated-actuated. These actuated control
variations are described as follows:
• Ful1y actuated cOlltrolimplies that aH phases are actuated and all
interseetion traffie movements are detected. The sequence and duration of
each phase are determined by traffic demando Hence, this type of control
is Ilot associated with a constant cyde length.
• Semiactuated control uses actuated phases to serve the minor movements at
an intersection. Only these minor movements have deteetion. The phases
associated with the major movements are operated as "nonactuated." The
eontroller is programmed to dwell with the nonaetuated phases

Coocep" Chapter 19/5ignalized Intersections

Page 19-4 Version 6.0
 Highway Capacity Manual: A Guide for Mulfimodal MobUity Analysis

displaying green for at least a specified minimum duration. The sequence

and duration of each actuated phase are determined by traffic demando
Hence, this type of control is 'lOt associated with a constant cyele length.
• Coo,dinated-actuated control is a variation of semiactuated operation. It uses
the controller's force-off settings to constrain the noncoordinated phases
associated with the minor movements so that the coordinated phases are
served at the appropriate time during the signal cyele, and progression
for the major movements is maintained. This type of control is associated
with a constant cyele length.
Signalized interscctions located c10seto one another on the same street are
often operated as a coordinated signal system, in which specific phases at each
intersection are operated on a common time schedule to pcrmit the continuous
flow of the associated movements at a planned speed. The signals in a
coordinated system typically opcratc by using pretimed or coordinated-actuated
control, and the coordinated phases typicalIy serve the major-street through
movements. Signalized intersections that are not part of a coordinated system are
characterized as "isolated" and typicalIy operate by using fully actuated or
semiactuated control.

Intersection Traffie Movements

Exhibit 19-1 iIIustrates typical vehicle and pedestrian traffic movcments at a
four-leg intersection. Three vehicular traffic movements and one pedestrian
traffic movement are shown for each intersection approach. Each movement is
assigned a unique number or a number and letter combination. The letter "P"
denotes a pedestrian movement. The number assigned to each left-turn and
through movement is the same as the number assigned to each phase by
National Electrical Manufacturers Association (NEMA) specifications.

Minor 5treet Exhibit 19-1

Intersection Traffic
- Vehicle Movements 14 4 7
Movements and
- Pedestrian Movements

______
jll
J -,----- L _
Numbering 5cheme

Intersection traffic movements are assigned the right-of.way by the signal

controller. Each movement is assigned to one or more signal phases. A phase is
defined as the green, yellow change, and red clearance intervals in a cycle that
are assigned to a specified traffic movement (or movements) (28). The

Chapler 19j5ignalized Intersections Concepts

VersiOO 6.0 Page 19.5
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

assignment of movements to phases varies in practice, depending on the desired

phase sequence and the movements present at the intersection.

Signal Phase Sequence

Modem actuated conrrollers implement signal phasing by using a dual-ring
structure that allows for the concurrent presentation of a green indication to two
phases. Each phase serves one or more movements that do not conflict with each
other. Early controllers used a single-ring structure in which all nonconflicting
movements were assigned to a common phase, and its duration was dictated by
the movement needing the most time. Of the hvo struetures, the dual-ring
structure is more efficient bccause it allows the controller to adapt phase
duration and sequence to the needs of the individual movements. The dual-ring
structure is typically used with eight phases; however, more phases are available
for complex signal phasing. The eight-phase dual-ring structure is shown in
Exhibit 19-2.The symbol el> represents the word phasc, and the number following
the symbol represents the phase number.

Exhibit 19-2 Majar 5treet Phases Minol' 5treet Phases

Dual-Ring Structure with
r------------------------------------------l
IlIustrative Movement
Assignments
I
I
I
I
L__
•., •.,, ..•....• ••• ••3
<\p 14 <\
I
I
I
I
_ __ 1
í'
" 1) !
2-
12 •••••••• ,
•., "- •., ,
ro--
I
I
I
,./ -
416 ~6P
..•..•

•
..•....
,,
16

...-
\-.
•••
1(1
," .. '"
---1
I
I
I

---
l
1 BalTler Barner /' 1I

••••••••
-
I'emlitted Mo\oement
Pedestrian Movement
--------~.~n~
Exhibit 19-2 shows one way traffic movements can be assigned to each of the
eight phases. Thcse assignments are illustrative, but they are not uncommon.
Each left-turn movement is assigned to an exclusive phase. During this phase,
the left-turn movemcnt is "protected," so it receives a green arrow indication.
Each through, right-turn, and pcdestrian movement combination is also assigned
to an exclusive phase. The dashcd arrows indicate turo movements that are
served in a "permitted" manner so that the turn can be completed only after
yielding the right-of-way to conflicting movements.
Two rings and two barriers are identified in Exhibit 19-2. A ring consists of
two or more sequentially timed conflicting phases. Ring 1 consists of Phases 1, 2,
3, and 4. Ring 2 consists of Phases 5, 6, 7, and 8. A barrier is used when there are
two or more rings. lt represents a reference point in the cycle at which one phase
in each ring must reach a common point of terminabon. In Exhibit 19-2, a barrier
is shown following Phases 2 and 6. A second barrier is shown following Phases 4
and 8. Behveen barriers, only one phase can be active at a time in each ringo

Con<epts Olapter 19¡5ignarized Intersections

Page 19-6 Vef5k¡n 6.0
 Highway Capacity Manual; A Guide tor Multimodal Mobility Analysis

The ring structure dictates the sequence of phase presentation. Sorne

common rules are provided in the following list:
• Phase Pairs 1-2, 3-4, 5--6,and 7-8 typically occur in sequence. Thus, Phase
Pair 1-2 begins with Phase 1 and ends with Phase 2. Within each phase
pair, it is possible to reverse the arder of the pair. Thus, Pair 1-2 could be
set to begin with Phase 2 and end with Phase 1 if it is desired to have the
lefHum Phase 1 lag through Phase 2.
• Phase Paír ]-2 can operate concurrently with Phase Pair 5--6.That is,
Phase 1 or 2 can time with Phase 5 or 6. Similarly, Phase Pair 3-4 can
operate concurrently with Phase Pair 7-8. These phase pairs are also
known as concurreney groups.
• For a given concurrency group, the last phase to occur in one phase pair
must end at the same time as the last phase to occur in the other pair (Le.,
they end together at the barrier).
• Phascs between two barriers are typically assigned to the movements on a
common street. For example, the four phases bctween the first and second
barriers shown in Exhibit 19-2are assigned to the minar street.

Operational Modes
There are three operational modes for the tum movements at an intersection.
The names used to describe these modes refer to the way the tum movement is
served by the controller. The three modes are as follows:
• Permitted,
• Protected, and
• Protected-permitted.
The permitted mode requires tuming drivers to yield to conflicting traffic
streams before completing the tumo Permitted leH-turning drivers yíeld to
oncoming vehicles and conflicting pedestrians. Permitted right-tuming drivers
yield to conflicting pedestrians. The efficiency of this mode depends on the
availability of gaps in the conflicting strcams. An exclusive tum lane may be
provided, but it is not required. The permittcd tum movement is typically
prcscnted with a circular green indication (although sorne agencies use other
indications, such as a f1ashing yellow arrow). The right-tum movements in
Exhibit 19-2 are operating in the permitted mode.
The protected mode gives turning drivers the right-of-way during the
associated turn phase, while all conflicting movements are required to stop. This
mode provides for efficient turn-movement service; however, the additional turn
phase typically results in increased delay to the other movements. An exclusive
turn lane is typically provided with this mode. The tum phase is indicated by a
green arrow signal indication. Left-tum Movements 3 and 7 in Exhibit 19-2 are
operating in the protected mode.
The protected-permitted mode represents a combination oí the permitted and
protected modes. Turning drivers have the right-of-way during the associated
turn phase. Turning drives can also complete the turn "permissively" when the
adjacent through movcment receives its circular green (or when the turning

Chapter 19/5ignalized Intersections Concepts

VetWn6.0 Page 19.7
Highway Capacify Manual: A Guide lor Mulfimodal Mobility Analysis

driver receives a flashing yellow arrow) indication. This mode provides for
efficient tum-movement service, often without causing a signifieant inerease in
the deJay to other movements. Lcft-tum Movements 1 and 5 in Exhibit 19-2 are
operating in the protected-permitted mode.
The operational mode used for one left-tum movement is often also used for
the opposing left-tum movement. For example, if one left.tum movement is
permittcd, fuen so is the opposing left.tum movement. Howevcr, the modes for
opposing left-tum movements are not required to be the same.

Left- Turn Phase Sequence

This subsection describes the sequence of scrvice provided to left-tum
rnovements relative to the other intersection rnovernents. The typical options
inelude the following:
• No left-tum phase (Le., pennitted only),
• Leading left-tum phase,
• Lagging left-tum phase, or
• Split phasing.
The permitted-only option is used when the left-tum rnovement operates in
the permitted mode. A left-tum phase is not provided with this option. An
illustrative implementation of permittcd-only phasing for left- and right-tuming
traffic is shown in Exhibit 19-3 foc the minor strcet.

Exhibit 19-3 Major Street Phases Mil"lOf Street Phases

II1ustrative Protected Lag-lag r------------------------------------------l

,- ..
and Permitted-Qnly Phasing I I
I I
I
I
••
2
••• "3 ••• , , I
I
L__
í' " _ __ 1

,,-
12 .•••••••• "'"'"
\./ ! \.
..,
-
••• "5 •••
r--
I
I
-'",
-...... 16
,../ "'."' '\ 1 (\ ---1
I
I
..,.'"
3 8 18
I I
lI Barnef./ Barner 1I
••••• ProIected Movement
.••••••• Permilted Movement
--------~.~ TI~
- Pedeslriiln Movement

A leading, lagging, or split-phase sequence is used when the left tum

operates in the proteeted mode or the protected.permitted mode. Leading and
lagging indicate the order in which the left-tum phase is presented relative to the
eonflicting through movement. The leading left-tum sequence is shown in
Exhibit 19-2 foc the left-tum movements on the major and minor streets. lOe
Jagging left-tum sequence is shown in Exhibit 19-3 foc the left-tum movements
00 the major street. A mix of leading and lagging phasing (ealled lead-lag) is
shown in Exhibit 19-4 for the left-tum movements on the major street.

Concepts O1apter 19/5ignali2ed Intersections

Page 19-8 Version 6.0
 Highway Capacity Manual: A Guide tor Multimodal Mobility Analys;s

Split phasing describes a phasc scquencc in which one phase serves aH

movements on one approach and a second phase serves aH movements on the
opposing approach. Split phasing requires that al! approach movements
simultaneously receive a green indication. Split phasing is shown in Exhibit 19-4
for the minor street. Other variations of split phasing exist and depend on the
treatment of the pcdestrian movements. The left-turn movement in a split phase
typicaHy opera tes in the protected mode (as shown), provided there are no
conflicting pedestrian movements.
Majar Street Phases Minar 5treet Phases Exhibit 19-4
r------------------------------------------I IIIustrative Protected Lead-

,- '\ t (/
I I Lag and Split Phasing
I I
I ~1 ~1 ~3 ~4 I
I
•• 1~ ~ 7 I
L__
í' _ __ 1

~.
_.-., /./ 1"
12 •••••• ~

"-
4>'
3 8 18

~7 " ~8

r--
I
I
I
••••••• 16
,../ "o_ ".- ---,
I
I
I
I Barrier ••...• Barrier .....- I
L I
••••• Prol:ected Movemeflt
.••••••• Permitted Movemeot
--------~ •• Tome
- Pedestri<ln ~ent

Traffie Flow Characteristics

This subsection describes several fundamental attributes of flow at
signalized intersections. Exhibit 19-5provides a reference for much of the
discussion. The diagram represents a simple situation of vehicles on one
approach to a signalized intersection during one signal eycle. The red clearance
interval is not used in the example phase sequence shown in the exhibit.
Exhibit 19-5 is dividcd into thn.'Cparts. Part 1 shows the time-space
trajectory of several vehicles on the approach as they travel to (and through) the
intersection. The horizontal bar represents the signal display (or "indication")
over time. lt is located in the figure at a position that coincides with the
intcrsection stop lineoPart 2 shows the durations of the displayed red, green, and
yellow intervals. It also shows the effective green, effective red, and lost time
durations. The terms and symbols shown in Part 2 of Exhibit 19-5are used
throughout this chapter and Chapter 31, Signalized lntersections: Supplemental.
Part 3 shows a flow profile diagram of the discharge flow rate (measured at thc
stop line) as a function of time.
The motorized vehide mcthodology describcd in this chapter disaggregates
thc signal eyde into an effective grecn time and an effective red time for each
phase to facilitatc the evaluation of intersection operation. These two times are
shown in Part 2 of Exhibit 19-5. Effective grecn time is the time that can be used
by vehides to proceed effectively at the saturation flow rate. Effective red time
for a phase is egual to the cyde length minus the effective green time. Formal
definitions for the effective grecn and red times are provided in Exhibit 19-6.

Chapter 19j5ignalized Intersections Concepts

Version 6.0 Page 19-9
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Exhibit 19-5
Fundamental Attributes of
Traffic Aow at Sigrlalized
Intersections
J L
¡J

2 I i I I I,
• 1/,1

ITI
! ! !
,-- ,
~
!
!
I
I I
i

-,-
I ! •
------------~-~-._--------------~- ----------¡...--i-;--
I i ! ........,..¡-
I1 l:..--= :
i
3 ,I I
I I!
,
S II
I ~-- I
I

Lost Time
As shown in Part 2 of Exhibit 19-5, two increments oElost time are associated
with a phase. At the beginning oEthe phase, the first few vehides in the queue
depart at headways that exceed the saturation headway. The longer headway
reflects the additional time the first few drivers require to respond to the change
in signar indication and accelerate to the running speed. The start-up losses are
caBed start-up lost time 1¡.

At the end of the phase, the yellow indication is presented, and approaching
drivers prepare for the signal to change to red. An initial portion of the yellow is
consistently used by drivers and is referred to as the extension oE the effective
green e. The latter part oEthe change period (Le., the yellow change interval and
the red dearance interval), which is not used, is reEerred to as dearance lost time
12, Phase lost time 1,equals the sum of the start-up and dearance lost times.
Formal definítions Eorthese terms are provided in Exhibít 19-6.

Con<epts O1apter 19j5ignalized Intersections

Page 19-10 Vemm6.0
 Highway Capacify Manual: A Guide for Multimodal Mobility Analysis

Term Symbol Definition Exhibit 19-6

Fundamental Variables of
Terms Used as Variables
Traffk Flow at Signalized
Green interval The duration of the green intervaC'-,-"-oaC,Cted~W-iC,,,-,-pCh'-"'-.
CA-9-"-,,o- Intersections
G
duratiOn (s) indication is displayed fo{ this duration.
This interval follows the green interval. It is used to wam drivers of
Yellow change y the impending red indication. A yellow indicatiOll is displayed for this
interval (s)
duration.
This interval tollows the yellow change interval and is optionally
Red elearance
used to provide additional time before conflicting movements
¡nterval (S)
receive a green indication.
The sum of the yellow change interval and red clearance interval for
Change pericxl CP
a given phase.
The time in the signal cyele during which the signal indication ís red
Red time (s) R
for a given phase.
Cycle length (s) e The total time fof a signal to complete one cyele.
Effective green The time during which a combination of traffic movements is
time (s) 9 considered to pt"oceed effectively at the saturation flow rate.

Effective red time

(,)
, The time during which a combination of traffic movements is IlOt
considered to proceed effectively at the saturation flow rate. It is
equal to the cycle length minus the effective green time.
The initial portion of the yellow change interva1 during which a
Extensioo of
effective green (s)
e combination of traffic movements is considered to proceed
effectively at the saturation flow rate.
The additiOnal time consumed tiy the first few vehicles in a queue
Start-up lost time
whose headway exceeds the saturation headway because of the
(,)
need to react to the initiatian of the green interval and accelerate.
The latter part of the change pericxl that is not typicalty used by
Oearance lost time
drivers te proceed through the intersection (i.e., they use this time
(,)
to stop in advance of the stop line).
Phase lost time (s) The sum of the clearance lost time and start-up lost time.

Cyele lost time (s)

"L The time Iost during the cyele. rt represents the sum of the lost time
for each critical phase.
The equivalent hourly rate at which previously queued vehieles can
Adjusted saturation traverse an intersection approach under prevailing conditions,
fIow rate (veh/h/ln)
s assuming the green signal is available at all times and no lost times
are experienced.
The component of delay that results when a traffic control device
Control delay
d causes a traffic movement to reduce speed or to stop. It represents
(s/veh)
the increase in travel time relame to the uncontrolled condition.

Terms Not Used as Variables

Cyele
'--:--:--:-----
The time te complete one sequence of signal indications.
Interval A pericxl of time during which all signa! indications remain constant.
The green, yellow change, and red clearance intervals assigned to a
Phase
spedfied movement (or movements) .
.....:.._-----
The relationship betwt.>enphase 10st time and signa] timing is shown in
Equation 19-1.
le=ll+lz Equation 19-1

le = II + Y + Re - e
where
II phase 105ttime (s),
11 start-up lost time = 2.0 (s),

Chapter 19/5ignalized Intersections Concepts

Version 6.0 Page 19-11
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

12 c1earance lost time = Y + R, - e (s),

e = extension of effective green = 2.0 (s),
y yellow ehange interval (s), and
Re red c1earance interval (s).

Research (29) has shown that start-up lost time is about 2 s and the extension
of effeetive green is about 2 s (longer times may be appropriate for eongested
conditions, higher speeds, or heavy vehicles). If start-up lost time eguals the
extension of effective green, then phase lost time is egual to the change period
(Le., 1/ = Y + Re)'

Saturation Flow Rate

Saturation f10w rate is the equivalent hourly rate at which previously queued
vehicles can traverse an intersection approach, assuming the green signal is
avaiJable at aH times and no lost times are experieneed. lt is expressed as an
expeeted average hourly rate in units of vehicles per hour per lane. The concept
of saturation f10w rate is discussed in more detail in Chapter 4, Traffie
Operations and Capacity Coneepts.
The base saturation f10w rate represents the expeeted average f10w rate for a
through-traffie lane for exeeptionally favorable geometrie and traffic eonditions
(e.g., no grade, no trucks, and so forth). The adjllsted saturation f10w rate
represents the saturation f10w rate for prevailing geometric and traffie
conditions. Prevailing conditions typically result in the adjusted saturation f10w
rate being smaller than the base saturation f10w rate.
A procedure for estimating the adjusted saturation f10w rate for a Jane group
is provided in $eetion 3. The procedure consists of a base saturation f10w rate
and a series of adjustment factors. The factors are used to adjust the base rate to
refleet the prevailing conditions associated with the subjeet lane group.
The saturation f10w rate for prevailing conditions can be determined direetly
from field measurement. A teehnique for measuring this rate is described in
$ection S of Chapter 31, Signalized lntersections: SupplementaL

capaCity
Capacity is defined as the maximum number of vehicles that can reasonably
be expeeted to pass through the interseetion under prevailing traffic, roadway,
and signalization eonditions during a lS-min periodo Capacity is eomputed as
the product oí adjusted saturation f10w rate and effective green-to-cycle length
ratio. Capacity is expressed as an expceted average hourly rate in units oí
vehicles per hour.

Phase Duration
This subseetion describes the eomponcnts oí phase duration. The discussion
is íocused on an actuatcd phasc; however, some elements of the discussion are
egually applicable to a pretimed phasc.
The duration oí an aetuated phase is composed oí five time periods. The first
period represcnts the time lost while the queue reacts to the signal indication

Concepts Chapter 19j5igrlalized Intersections

Page 19-12 V~6.0
 Highway Capacity Manual: A Guide for Multimodal MobMy Analysis

ehanging to grecn. The second interval represents the time required to clear the
queue of vehicles. The third period represents the time the green indication is
extended by randomly arriving vehicles. It ends when there is a gap in traffic
(Le.,gap-out) or the green extends to the maximum limil (Le.,max-out). The
fourth period represents the yellow change interval, and the fifth period
represents the red clearance interva1. The duration of an actuated phase is
dcfined by Eguation 19-2.
Dp = l1 + B5 + Be + y + Re Equation 19-2

where
Dp phase duration (s),
1\ start-up lost time'" 2.0 (s),
g; gueue service time (s),
g, green extension time (s),
y yellow change interval (s), and
Re red clearanee interval (s).
The relationship bctween the variables in Eguation 19.2 is shown in
Exhibit 19-7by using a gueue accumulation polygon.

Exhibit 19-7
Red Phase Duration, Dp
•, .,
Time Elements Influencing
• 9 Actuated Phase Ouration
& 9s g~ Y.f.Re
II I

.5
•
-1l
:c
~
'S
• Area '" Delay
.1i
,
E
z O
O Time es)

Exhibit 19-7 shows the relationship between phase duration and queue size
far the average signal cycle. During the red interval, vehicles arrive at arate of qr
and form a queue. The queue reaches its maximum size /\ seconds after the red
interval ends. At this time, the queue bcgins to diseharge at arate equal to the
saturation flow rate s less the arrival rate during green qg' The queue clears y<
scconds after it first bcgins to discharge. Thereafter, random vehicle arrivals are
deteeted and cause the green interval to be extended (provided the headway
between vehicles remains below a specified value). Eventually, a gap occurs in
traffic (or the maximum greco limit is reached), and the green interval ends. The
end of the green interval coincides with the end of the extension time y~.
The effective grecn time for the phase is computed as shown in Equation 19-3.
g=Dp-l1-i2 Equation 19-3

9 = B5 + Be + e

Chapter 19j5ignalized Intersections Concepts

Vef3ion 6.0 Page 19-13
Highway Capacify Manual; A Guide for Mulfimodal MobiHty Analysis

where
12 elearancc lost time"" Y + Re - e (s), and
e '= extension of effective green" 2.0 (s).

ANALYSIS TYPE
The term analysis type is used to describe the purpose for which a
methodology is used. Each purpose is associated with a different level of detail
as it relates to the predsion of the input data, the number of default values used,
and the desired accuracy of the results. Three analysis types are recognized in
this chapter:
• Operational,
• Design, and
• Planning and preliminary engineering.
These analysis types are discussed in more detail in Chapter 2, Applications.

SPATlAL ANO TEMPORAL L1MITS

Analysis Boundaries
The intersection analysis boundary is defined by the operational influence
area of the intersection. It is not defined as a fixed distance for a1lintersections.
Rather, it extends upstream from the interscction on each intersection leg. The
size of this area is leg specific. It ineludes all geometric featUfes and traffic
conditions that influence intersection operation during the srudy periodo For
these reasons, the analysis boundaries should be established for each intcrscction
on the basis of the conditions present during the study periodo

5tudy Period and Analysis Period

The study period is the time interval represented by the performance
evaluation. It consists of one or more consecutive analysis periods. An analysis
period is the time interval evaluated by a single application of the methodology.
The methodology is based on the assumption that traffic conditions are
steady during the analysis period (Le., systematic change over time is negligible).
For this reaso", the duration of the analysis period is in the range of 0.25 to 1 h.
The longer durations in this range are sometimes used for planning analyses. In
general, the analyst should use caution with analysis periods that exceed 1 h both
because traffic conditions typically are not steady for long time periods and
because the adverse impact of short peaks in traffic demand may not be detected
in the evaluation.

LOS CRlTERIA
This subsection describes the LOS crHeria for the motorized vehiele,
pedestrian, and bicyele modes. The criteria for the motorized vehiele mode are
different from those for the other modes. Specifically, the motorized vehiele-
mode criteria are based 00 performance measures that are field measurable and
perceivable by travelers. The criteria for the other modes are based 00 scores
reported by travelers indicating their perception of service quality.

Concepts Chapter 19j5ignalized Intersections

Page 19.14 Version 6.0
 Highway Capacify Manual: A Guide for Mulfimodal Mobility Analysis

Motorized Vehicle Mode

AfI usesofthe tennsvolume ()("
LOS can be characterized for the entire interscction, each interscction
volume-to-capacity ratio in this
approach, and each lane group. Contro! del ay alone is used to characterize LOS chapter refer lo demand
voIume or demi1nd voIume-to-
for the entire intersection or an approach. Control delay and volume-to-capacity
capacity ratio.
ratio are used to characterize LOS for a lane group. Oelay quantifies the increase
in travel time due to traffic signal control. It is also a surrogate measure of driver
discomfort and fuel consumption. The volume-to-capacity ratio quantifies the
degree to which a phase's capacity is utilized by a lane group. The following
paragraphs describe each LOS.
LOS A describes operations with a control delay of 10 s/veh or less and a
volume.to.capacity ratio no greater than 1.0. This level is typically assigned
when the volume.to-capacity ratio is low and either progression is exceptionally
favorable or the cyele length is very short.1f LOS A is the resuH of favorable
progression, most vehieles arrive during the green indication and trave1 through
the intersection without stopping.
LOS B describes operations with control dclay between 10 and 20 s/veh and a
volume-to-capacity ratio no greater than 1.0. This level is typically assigned
when the volume.to-capacity ratio is low and either progression is highly
favorable or the cyele length is shorL More vehieles stop than with LOS A.
LOS e describes operations with control delay between 20 and 35 s/veh and
a volume-to-capacity ratio no greater than 1.0. This level is typically assigned
when progression is favorable or the eycle length is moderate. Individual cycle
failures (Le., one or more queued vehicles are not able to depart as a result of
insufficient capacity during the cyele) may begin to appear at this leve!. The
number of vehicles stopping is significant, although many vehicles stiB pass
through the intersection without stopping.
LOS O describes operations with control delay betwecn 35 and 55 s/veh and
a volume-to-capacity ratio no greater than 1.0. This level is typically assigned
when the volume-to.capacity ratio is high and either progression is ineffective or
the cyele length is long. Many vehicles stop and individual cycle faHures are
noticeable.
LOS E describes operations with control del ay between 55 and 80 s/veh and a
volume-to-capacity ratio no greater than 1.0. This level is typically assigned
when the volume-to-capacity ratio is high, progrcssion is unfavorable, and thc
eycle length is long. Individual cycle faHures are frequent.
LOS F describes operations with control delay exceeding 80 s/veh or a
volumc-to-capacity ratio greater than 1.0. This leve1is typically assigned when
the volume-to-capacity ratio is very high, progression is very poor, and the cycle
length is long. Most cycles fail to clear the queue.
A lane group can incur a delay less than 80 s/veh when the volume-to-
capadty ratio exceeds 1.0. This condition typically occurs when the cyc1elength
is short, the signal progression is favorable, or both. As a resuH, both the delay
and volume.to-capacity ratio are considered whcn lane group LOS is established.
A ratio of 1.0or more indica tes cycle capacity is fully utilized and represents

Chapter 19/5ignalized intersections Concepts

Version 6.0 Page 19-15
Highway Capacify Manual; A Guide tor MuJtimodal Mobilify Analysis

faUure fram a capacity perspective (just as delay in excess of 80 s/veh represents

faUure fram a delay perspective).
Exhibit 19-Blists the LOS thresholds established for the motorized vehide
made at a signalized intersection.

Exhibit 19-8 LOS by Volume-to-CaDacityRatjo"

LOS úiteria: Motorized Control Delay(s{veh) ~1.0 >1.0
Vehicle Mode
S10 A F
>10-20 B F
>20-35 e F
>35-55 D F
>55-80 E F
>80 F F
Note: •. For approact>.based and ifltersectioowide assessmef1ls, LOS is defined solely by control delay.

Pedestrian and Bicycle Modes

Historically, the HCM has used a single performance measure as the basis
for defining LOS. However, research documented in Chapter 5, Quality and
Level.of-Service Concepts, indicates travelers consider a wide variety of factors
in assessing the quality of sen-ice pravided to them. Sorne of these factors can be
described as performance measures (e.g., speed) and others can be described as
basic descriptors of the intersection character (e.g., crosswalk width). The
methodologies for evaluating the pedestrian and bicycle modes combine these
factors to determine the corresponding mode's LOS.
Exhibit 19-9lists the range of scores assaciated with each LOS for the
pedestrian and bicycle travel modes. The association between score value and
LOS is based on traveler perception research. Travelers were asked to rate the
quality of service associated with a specific trip through a signalized intersection.
The letter "A" was used to represent the best quality of service, and the letter "F"
was used to represent the worst quality of service. Best and worst were left
undefined, allowing respondents to identify the best and worst conditions on the
basis of their traveling experience and perception of service quality.

Exhibit 19-9 LOS LOSScore

LOS Criteria: Pedestrian and A S 1.50
Bicyde Modes
B >1.50-2.50
e >2.50-3.50
O >3.50-4.50
E >4.50-5.50
F >5.50

SCOPE OF THE METHODOLOGIES

This subsection identifies the conditions for which each methodology is
applicable.
• Intersectioll geometry. Al! methodologies apply to three- and four.leg
intersections of two streets or highways.
• Flow cOllditiolls. The three methodologies are based on the analysis of
steady traffic conditions and, as such, are not well suited to the evaluation
of unsteady conditions (e.g., congestion, queue spillback, signal
preemption).

Concepts O1apter 19¡5ígnalízed Intersections

Page 19.16 VersiQn 6.0
 Highway Capacify Manual: A Guide far Multimodal Mobility Analysjs

• TaTget Toad usas. Collectively, the three methodologies were developed to

estimate the LOS perceived by motorized vehicle drivers, pedestrians,
and bicyclists. They were not developed to provide an estimate oi the
LOS perceived by other road users (e.g., commercial vehicle drivers,
automobile passengers, delivcry truck drivers, recreational vehicle
drivers). However, it is likely the perceptions oi these other road users are
reasonably well represented by thc road uscrs for whom the
methodologies were developed.
• lnjluellces iJl the Tight-of-way. A road user's perception oi quality of servicc
is influenced by many factors inside and outside the urban street right-oi.
way. However, the methodologies in this chapter were spedfical1y
constructed to exclude factors that are outside the right-of-way (e.g.,
buildings, parking lots, scenery, landscaped yards) that might influence a
traveler's perspective. This approach was iollowed because factors
outside the right-of-way are not under the direct control of the agency
operating the street.

Chapter 19j5ígnalízed Intersections Concepts

Version 6.0 Page 19-17
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

3. CORE MOTORIZED VEHICLE METHODOLOGY

This section describes the methodology for evaluating the capacity and
quality of service provided to motorized vehides at a signalized intersection.
Basic extensions of this methodology to address critical intersection volume.to-
capacity ratio and initial queue presence are provided in Section 4. Extensions to
this methodology for evaluating more complex intersection operational elements
(e.g., queue length) are described in Chapter 31, Signalized Intersections:
Supplemental.

SCOPE OF THE METHODOLOGY

The overall scope of the three methodologies is provided in Section 2. This
section identifies the additional conditions for which the motorized vehide
methodology is applicable.
• Upstream intersectiOlls. The influence oi an upstream signalized
intersection on the subject intersection's operation is addressed by input
variables that simply, and subjectively, describe platoon structure and the
uniiormity oi arrivals on a cydic basis. Chapter 18, Urban Street
Segments, extends the methodology described in this chapter to the
evaluation oi an interscction that is part of a coordinated signal system.
• Target travel modes. The motorized vehide methodology addresses mixed
automobile, motoreyde, truck, and transit traffic streams in which the
automobile represents the largest percentage oi a1l vehides. The
methodology is not designed to evaluate the performance oi other types
of vehides (e.g., golf carts, motorized bicydes).

Spatial and Temporal Limits

Analysis Boundaries
The intersection analysis boundary is defined by the operational influence
area oi the intersection. It should indude the most distant extent oi any
intersection-related queue expected to occur during the study periodo For these
reasons, the ¡nfluence area is Iikely to extend al least 250 £tback from the stop
line on each intersection leg.

5tudy Perlod and Analysis Pen"od

The concepts of study period and analysis periad are defined in Section 2 with
general terms. They are defined more precisely in this subsection as they relate to
the motorizcd vehide methodology.
Exhibit 19-10 demonstrates three altemative approachl's an analyst might
use for a given evaluation. Other altematives exist, and the study period can
exceed 1 h. Approach A is the approach that has traditionally been used and,
unless otherwise justified, is the approach that is recommended ior use.

Core Motorized Vehide Methoclology Chapter 19j5ignalized lntersections

Page 19-18 Vt'1Wn 6.0
 Highway Capacify Manual: A Guide for Multimodal Mobility Analysis

AOIXcachA Aporcad! B AWoach e Exhibit 19-10

Three Altemative Study
Stu:ly Period = 1.0 h Study Period '" 1.0 h Study Period '" 1.0 h Approaches

Single analysis period Single analysis period Mulliple analysis periods

T=0.25h T=l.Oh T =0,25h

Time

~ - analysis period

Approaeh A is based on evaluation of the peak lS-min period during the The use of a peak 15-min
traffie count multiplied by four
study periodo The analysis period T is 0.2S h. The equivalent hourly flow rate in is preferred for exi5ting
vehicles per haur used for the analysis is based on either (a) a peak 1S-min traffic conditions when trafflC counts
are avallable. Tñe use of a l-h
count multiplied by four or (b) a 1-h demand volume divided by the peak hour ciernand voIume divided by a
factor. The former option is preferred for existing conditions when traffie counts peak hour Setor is preferred
when projected voIumes are
are available; the laUer option is preferred when hourly projected volumes are used or when voIumes are
used or when hourly projected volumes are added to existing volumes. used that llave been added to
exiSóng voIumes.
Additional discussion on use of the peak hour factor is provided in the
subsection titled Required Data and Sources.
Approaeh B is based on evaluation of one 1.h analysis period that is
coincident with the study periodo The analysis pcriod T is 1.0 h. The flow rate
used is equivalent to the 1-h demand volume (Le., the peak hour factor is not
used). This approach implicitly assumes the arrival rate of vehicles is eonstant
throughout the period of study. Therefore, the effects of peaking within the hour
may not be identificd, and the analyst risks underestimating the delay actually
incurred.
Approach e uses a 1.h study period and divides it into fourO.2S-h analysis
pcriods. This approaeh accounts for systematic flow rate variation among
analysis periods. It also accounts for queues that carry over to the next analysis
period and produces a more accurate representation of delay. This approach,
which is called a multip!e time-period analysis, is dcscribed in the next subsection.
Regardless of analysis period duration, a single-period analysis (i.e.,
Approach A or B) is typical for planning applications.

Multiple Time.Period Analysis

If the analysis period's demand volume exceeds capacity, then a multipJe
time-period analysis should be undertaken whcn the study period indudes an
initial analysis period with no initial queue and a final analysis period with no
residual queue. The initial queue for the second and subsequent periods is equa!
to the residual queue from the previous periodo This approaeh provides a more
accurate estimate of the deJay associated with the congestiono

Chapter 19/5ignalízed Intersections Core Motorized Vehicle Methodology

Vef5ion 6.0 Page 19-19
Highway Capacity Manual: A Guide fer Multimedal Mebility Analysis

If evaluation of multiple analysis periods is determined to be important, then

the performance estima tes for each period should be separately reported. In this
situation, reporting an average performance for the study perlod is not
encouraged because it may obscure extreme values and suggest acceptable
operation whcn in reality sorne analysis periods have unacceptable operation.

Performance Measures
Performance measurcs applicable to the motorized vehiele travel mode
inelude volume-to-capacity ratio, control delay, and queue storage ratio. The
queue storagc ratio describes the ratio of the back-of-queuc size to the available
vehicle storage length. The back of queue represents the maximum backward
extent of queued vehieles during a typical cyele.
LOS is also considered a performance measure. It is useful for describing
intersection performance to eJected officials, policy makers, administrators, or
the publico LOS is based on control delay.

limitations of the Methodology

This subsection identifies the known limitations oí the motorized vehiele
methodology. If one or more of thcse limitations are believed to have an
important influence on the performance of a spt.."Cific
street segment, then the
analyst should consider using altemative methods or tools for the evaluation.
The motorized vehiele methodology docs not account for the effect of the
following conditions on intersection operation:
• Turn bay overflow;
• Multiple advance detectors in the same lane;
• Demand starvation due to a closely spaced upstream intersection;
• Queue spillback into the subject intersection from a downstream
intersection;
• Queuc spillback fram the subject intersection into an upstream
intersection;
• Premature phase termination due to short detection Jength, passage time,
or both;
• Right-turn-on-red (RTOR) volume prediction or resulting right-turn
deJay;
• Tum movements servcd by more than two exclusive lanes;
• Oelay to traffic movements that are not under signal control;
• Through lane (or lanes) added just upstream of the intersection or
drapped just downstream of the intersection; and
• 5torage of shared-Iane left-turning vehicles within the intersection to
permit bypass by through vehicles in the same lane.
In addition to the aboye conditions, the methodology does not directly
account for the following controller functions:

(ore Motorized Vehicle Methodology O1apter 19j5ignalized Intersections

Page 19-20 Vetsion 6.0
 Highway Capacity Manual: A Guide for Mulfjmodal Mobility Analysis

• Rest-in-walk mode for actuated and noncoordinated phases,

• Preemption or priority modes,
• Phase overlap (see discussion that follow5), and
• Gap reduction or variable ininal settings for actuated phases.
Two control strategies that use phase overlap are addressed by the
methodology. One strategy is "right-turn overlap with the complementary left.
turn phase." This strategy uses the overlap featurc to providc a protccted signal
indication for the right-turn movement that is concurrent with the
complementary left-turn phase. Procedures for evaluating this operation are
described in Chapter 31, Signalized Intersections: Supplemental. They address
the case in which the right-turn movement has an exclusive tum lane.
The second strategy, "Ieft-tum overlap with the opposing through phase,"
uses the overlap fcature to provide the signal indication for permissive left-turn
operation that is concurrent with the opposing through phasc. This overlap
feature is referred to as "Dalias lefHum phasing." It is typically used with
flashing yellow operation. Procedures for evaluating this opcration are described
in Chapter 31.

lane Groups and Movement Groups

The motorized vehicle methodology uses the concepts of talle grOUl'S and
movemc"t grOl/ps to describe and evaluate intersection operation. These two
group designations are very similar in meaning. In fact, their differences emerge
onl)' when a shared lane is present on an approach with two or more lanes. Each
designa non is defined in the following paragraphs. Guidelines for establishing
lane groups and movement groups are described in the subscction titled
Computational Steps.

Lane Groups
The motorized vehicle methodology is designed to analyze the performance
of designated lanes, groups of lanes, an intersection approach, and the entire
inter5ection. A lane or group of lanes designated for separate analysis is referred
to as a lanc grol/p. In general, a separate lane group is established fOf (a) each lane
(or combination of adjacent lanes) that exclusively serves one movement and (b)
each ¡ane shared by two or more movements. A lane group can include one or
more ¡anes.
The concept of ¡ane groups is useful when a shared lane is present on an
approach that has two or more lanes. Several procedures in the methodology
require sorne indication of whether the shared lane serves a mix of vehicles or
functions as an exclusive tum lane. This issue cannot be resolved until the
peoportion of turns in the shared ¡ane has becn cornputed. If the cornputed
proportion of turos in the shared lane equals 1.0 (Le., 100%),the shared lane is
considered to operate as .lO exclusive tuen I¡,me.

Movement Groups
The concept of lIIovemellf groups is established to facilitate data entry to the
rnethodology. In this regard, input data describing intersection traffic are

Qlapter 19{5ignalized Intersections Core Motorized Vehícle Methodology

Version 6.0 Page 19.21
Highway Capacity Manual: A Guide tor Multimodol Mobility Anolysis

traditionally specific to the movement (e.g., lefHurn movement volume); the

data are not specific to the lane (e.g., analysts rarely have the volume for a lane
shared by left-turning and through vehides). Thus, most traffie eharacteristie-
related and geometric design-related input data are specific to a movement group.
The basie principIe for establishing movement groups is that no traffie
movement can be assigned to more than one movement group. Thus, a separate
movement group is established for (a) each tum movement with one or more
exclusive tum lanes and (b) the through movement (inclusive of any tum
movements that share a lane). A movement group can include one or more lanes.

REQUIRED DATA ANO SOURCES

This subsection describes the input data needed for the motorized vehicle
methodology. These data are Usted in Exhibit 19-11. The additional data needed
for coordinated control are Usted in Exhibit 19-12. The second column (Iabeled
Basis) of both Exhibit 19-11 and Exhibit 19-12 indicates whether the input data
are needed for eaeh traffic movement, a specifie movemcnt group, eaeh signal
phase, eaeh interseetion approaeh, or the intersection as a whole. The exhibits
also list potential data sources and default values that can be used if local data
are not available (30).
The data elements listed in Exhibit 19-11 and Exhibit 19-12 do not indude
variables that are considered to represent calibration faetors (e.g., start-up lost
time). A ealibranon factor typically has a relatively narrow range of reasonable
values or has a small impact on the accuracy of the performance estimates. The
recommended value for eaeh calibranon factor is idcntified at relevant points in
the presentation of the methodology.

Traffie Characteristics Data

This subsection describes the traffic characteristics data listed in Exhibit
19-11. These data describe the motorized vehicle traffic stream that lravels
through the intersection during the study period.

Demand F/ow Rate

The demand flow rate for an intersection traffic movement is defined as the
count of vehides arriving at the interscction during the analysis period divided
by the analysis period duration. It is expressed as an hourly flow rate but may
represent an analysis period shorter than 1 h. Demand flow rate represents the
flow rate of vehides arriving at the intersection. The "count of vehicles" can be
obtained from a variety of sourees (e.g., froro the field or as a fore<:ast from a
planning model).
When measured in the ficId, the demand flow rate is based on a traffie count
takcn upstream of the queue associated with the subjed interse<:tion. This
distinetion is important for counts during congested periods because the count of
vehides departing from a eongested approach will produce a demand flow rate
cstimatc that is Im'ver than the true rate (Le., the estimate will equal capacity
rather than true demand).

Core Motonzed Vehicle Methodology Qlapter 19/5ignalized Intersections

Page 19-22 Vet'Sion 6.0
 Highway Capacify Manual: A Guide tor Multimodal Mobility Analysis

Potential Data
Exhibit 19-11
Required Data and Units Basis SOurce($) Suggested Default Value Required Input Data, Potential
Data Sources, and Default
Traffic Characteristics
Values for Motorized Vehicle
Demand fIow rate (veh/h) M Fteld data, past counts Mus! be provided
Analysis
Right-turn-on-red fIow rate (veh/h) A Fteld data, past counts 0.0 veh,lh
Percent<lge heavy vehides ('lb) MG Fteld data, past counts 3%
Hourly data aM 0.2S-h analysis pefiod:
Field data, Total entefing vol. ~1,000 veh,lh: 0.92
Peak hour factor (decimal)
analyst Judgment Total entering vol. <1,000 veh,lh: 0.90
Otherwise: 1.00
Fteld data,
Platoon ratio (decimal) MG analyst judgment
UpstTeam fIItenng adJustment factor Field dat<!,
(decimal)
MG analyst Judgment
1.0

Fteld data,
Inibal queue (veh) MG analyst judgment
Must be provided

Fteld dat<!, Metro popo "2c2S0,000: 1,900 pc/h/ln

Base saturation fIow rate (pcjh/ln) MG analyst judgment Otherwise: 1,750 pc/h/ln
l.<me utilizatiOn adjustment factor Field data,
(decimal)
MG (In¡¡lystjudgment
See discussion

Pedeslri<In I\ow rate (p/h) A Field data, past counts Must be provkled
Bicycle flow r(lte (bicydes¡'h) A Fteld data, P<JStcounts Mus! be provided
On-street par1dng maroeuver rate Field data,
(veh,lh)
MG ana~t judgment
Field data, C60 bus stop: 12 buses¡h
Local bus stopping rate (buses/N A
ana~t Judgment NOI1-CSO bus stop: 2 buses,lh
Unsignalized movement delay (s) M Field data See diSCUSSK>n
Geometric Design
Number ot lanes (In) M Fteld data, aerial p/'lOtO Must be provided
Average lane width (ft) MG Field data, aeMa! photo ¡2 ft
Number ot receiving lanes (In) A Field data, aeria! photo Must be provided
Tum bay lengttl (ft) MG Field data, aerial p/'lOtO Must be provided
Presence ot on-street pali<.ing MG F.eld data, aeria! photo MuS! be provkled
Aat(lpproach:O%
Approach grade (%) A Field data Modefate grade 011approach: 3%
$teep grade 011approad!: 6%
Signal Control
Type of signal control ¡ Field data Must be provided
Phase sequeoce A Fielddata Must be provided
Left-turn operatlooal mode A Field data MuS! be provided
Dalias Ieft-wm ph<lsing opbon A Field data Dictated by local use
Passage time (s) (tf aetuated) p Field Q¡¡ta 2.05 (presence detection)
Major-street ttlrough movement: 50 s
Ma~imum green (s) (if aetuated) P Field data Minor-street ttlrough movement: 305
Left-wm movement: 20 ~
Major-street ttlrough movement: 50 s
G~ duration (5) (if pretimed) P Field data Minor""5treet ttlrough movement: 30 s
LeIt-tum movement: 20 s
Majof-street ttlrough movement: 10 s
Minimum green (5) P Field data Minor-street ttlrough movement: 8 s
LeIt-turn movement: 6 s
YeHow change + red dearance (5). P Field data 4.0 s
Actuated: 7.0 s
Walk(s) P Field data Pretimed: green interval minus
pedestrian dear
Pedestrian dear (5) P Field data 8ased en 3.S-ltIs walking speed
Ph<Jserec<lU (if actuated) P Field data No recal!
Dual ently (if aetuated) p Field data Not enabled (te., use single ently)
Simultaneous gap-oot (tf actuated) A
¡
Field data
EM"'"
Cycle lerogttl {if pretimed} Field data See discussion
Note: Exhibit COIltinues on the nelct page.

Chapter 19/5ignalized Intersections Core Motorized Vehicle Mettlodology

VerslOn 6.0 Page 19-23
Highway Capacity Manual: A Guide tor Multimodal MobiJity Analysis

Exhibit 19.11 (cont'd.) Potential Data

Required Input Data, Potential Required Data and Units Basls Source(s) Suggested Default Value
Data Sources, and Default otherData
Values for Motorized Vehicle 0.25 h
Analysis period duration (h) I Set by anatyst
Analysis
Speed limit (mlfh) A Field d<ll<I, roa<! inventory Must be provided
Stop-line detector lenglh (ft)
MG Field data 40 ft, presence deteetion mode
and detection mode (if actuated)
Alea type (eBD, non-eBD) Analyst Judgment Must be provided
Notes: M" movement: ooe valt.le for each left-turn, l:t1rotIQh, and right-turn movement
A " approach: one valt.le or condition for the intersection approach.
MG = IlI()VMlent group: Orle value for each tum movement wilh exdusive tum lanes and one valt.le for the
l:t1rough rnovement (inclusive of any tum rnovements in a stlared lane).
1 = intersection: one valt.le or COIldition for the intersection.
P " phase: one valt.le or conditiorl for eadl s¡gnal ptlase.
CBD " central business district; vol. " volume; popo " population.
'SpeciflC values of yeHow ctlange and red deararoce should be determined by local gu'delines or practice.

Exhibit 19-12 Potential Data

Required Additional Input Required Data and Units 8asis Source(s} Suggested Default Value
Data, Potential Data Sources, Cycle lengtrl (s) I Fteld d<lta See d&usslon
and Defautt Values for
Phase spj¡ts (s) (if actuated) P Fteld d<lta See discussion
Motorized Vehide Analysis
with Coordinated Signal EQual to tTavel time in l't1ase 2
Offset (s) Field data
(00,", Offset reference point (s) FteId d<lta
directioo'
End of green for Phase 2'
fQrce rnode (ir actuated) Aeld data Axed
Notes: 1 = intersection: one value or condition for l:t1e intersection.
P = phase: ooe valoo or COIldition for eadl signal pllase.
'Assumes l't1ase 2 is trie reference phase. SiJbstitute 6 if Phase 6 is the reference phase.

lf a planning analysis is being conducted in which (a) the projected demand

flow rate coincides with a l-h period and (b) an analysis of the peak 15-min
period is desired, then each movement's hourly demand should be divided by
the intersection peak hour factor to predict the flow rate during the peak 15.min
periodo The peak hour factor should be based on local traffic peaking trends.lf a
local factor is not available, then the default value in Exhibit 19-11can be used.
lf a multiple time-period analysis is conducted (Le.,Approach e in Exhibit
19-10), then the intersection's demand flow rates should be provided for each
analysis periodo
The methodology indudes a procedure for determining the distribution of
flow among the available lanes on an approach with one or more shared lanes.
The procedure is based on an assumed desire by drivers to choose the lane that
minimizcs their service time at the intersection. This assumption may not hold
for situations in which drivers choose a lane so they are prepositioned for a turn
at the downstream intersection. Similariy, it may not hold when an auxiliary
through lane is presento In either situation, the analyst will need to provide the
demand fIow rate for each lane on the approach and then combine these rates to
define the demand fIow rate (or each Jane group. Additional discussion of this
topic is provided in the subsection titled Lane Utilization Adjustment Factor.
The demand fIow rate for all signal-controlled movements must be provided.
The demand fIow rate for all unsignalized movements should be provided. If an
unsignalized movement exists but its f10wrate is not provided, then this
movement will be excluded from the ealculation ot approach deJay and
intersection deJay.

Core Motorized Vehide Methodology Chapter 19/5ignalized Intersections

Page 19-24 Version 6.0
 Hjghway Capadty Manual: A Guide tar Multimadal Mability Analysis

Right.Turn-on-Red Flow Rate

The RTOR fIow rate is defined as the count of vehicles that turo right at the
intersection when the controlling signal indication is red, divided by the analysis
period duration. It is exprcssed as an hourly fIow rate but may represent an
analysis period shorter than 1 h.
It is difficult to predict the RTOR fIow rate because it is based on many
factors that vary widely fmm intersection to intersection. These factors include
the following:
• Approach lane allocatian (shared or exclusive right-turn lane),
• Right-turn fIow rate,
• Sight distancc availablc to right-turning drivers,
• Volumc-to-capacity ratio for conflicting movements,
• Arrival patterns of righHurning vehieles during the signal cyele,
• Departure patterns of conflicting movements,
• Left-turn signal phasing on the conflicting street, and
• Conflicts with pedcstrians.
Given the difficolty of estimating the RTOR flow rate, it should be measured
in the field when possible. If the analysis is dealing with future conditions or if
the RTOR fIow rate is not known from field data, then the RTOR flow rate for
each right.turn movement should be assumed to equal O veh/h. This assumption
is conservative because it yields a slightly larger estimate of delay than may
actually be incorred by intersection movements.

Percentage Heavy VeNcles

A heavy vehicle is defincd as any vehiele with more than four tires touching
the pavement. Local buses that stop within the intersection area are not ineluded
in the count of heavy vchiclcs. The percentage of heavy vehiclcs represents the
count of heavy vehieles that arrive during thc analysis period divided by thc
total vehicle count for the samc periodo This percentage is providcd for each
intersection traffic movement; howcver, onc reprcsentative value for aH
movements may be used for a planning analysis.

Intersection Peak Hour Fador

One peak hour factor for the entire intersection is computed with Equation
19-4.

PHF=~
4 n15 Equation 19-4

wherc
PHF peak hour factor,
"no count of vehicles during a l-h period (veh), and
1/15 count of vehicles during the peak 15-min period (veh).

Chapler 19/5ignalized Intersections Core Motorized Vehide Meti1odology

Ver>ion6.0 Page 19-25
Highway Capacify Manual: A Guide for Multimodal Mobilify Analysis

The eount used in thc denominator of Eguation 19-4 must be taken during a
15-min period that occurs within the l-h period represented by the variable in
the numerator. 80th variables in this equation represent the total number of
vehiclcs entering the intersection during their respective time period. As such,
one peak hour factor is computed for the intersection. This factor is then applied
individually to each traffie movement. Values of this factor typically range from
0.80 to 0.95.
Tñe peak hour factor is used The peak hour factor is used primarily for a planning analysis when a
pn'marify for a pfanning
analysis wf/efl a forecast hourly forecast hourly volume is provided and an analysis of the peak 15-min period is
voIume is provided and an sought. Normally, the demand flow rate is computed as the eount of vchiclcs
analysisofthepeak 15-min
period is sought. arriving during thc period divided by the length of the period, expressed as an
hourly flow rate, and without the use of a peak hour factor.
If the peak hour factor is used, If peak hour faetors are used, a single peak hour factor for the entire
a single intersectionwide factor
shouid be used rather than intcrsection is generally preferred because it will decrease the likelihood of
mcvement-spedfic or creating demand seenarios with conflieting volumes that are disproportionate to
approach-spedfic fadots. If
individual appt()élChes (Ir the actual volumes during the 15-min analysis period.1f peak hour factors for
mcvements peak at different each individual approach or movement are used, lhey are Iikely to generate
times, a series of 15-min
analysis periods that dcmand volumes from one 15-min period that are in apparent conflict with
encompass the peaking shoufd demand volumes from another 15-min period, but in reality thesc peak volumes
be considered.
do not oecur at the same time. Furthcrmore, to determine individual approaeh or
movement peak hour faetors, actual 15-min eount dala are Iikely available,
permitting the determination of actual 15-min demand and avoiding the need to
use a peak hour factor. In the event individual approaches or movements are
known to have substantially differcnt peaking characteristics or peak during
different 15-min periods within the hour, a series of 15-min analysis pcriods lhat
encompass lhe peaking should be considered instead of a single analysis period
using a single peak hour factor for the interseetion.

Platoon Ratio
Platoon ratio is used to describe the quality of signal progression for the
corresponding movcment group. lt is computed as the demand flow rate during
the green indication divided by the average demand flow rate. Values for the
platoon ratio typically range frOID0.33 to 2.0. Exhibit 19-13 provides an indication
of the guality of progression associated with seleded platoon ratio values.

Exhibit 19-13 Platoon Ratio ArrivalType Progression Quallty

Relationship Between Arrival 0.33 1 Very peor
Type and Progressioo Quality 0.67 2 Unfavorable
1.00 3 Random arrivals
1.33 4 Favorable
1.67 5 Highly favorable
2.00 6 Exceptionally favorable

For proteeted or protected-permitted lcft-tum movements operating in an

exclusive lane, platoon ratio is used to describe progrcssion guality during the
associated tum phase (Le., the proteeted period). Hence, the platoon ratio is
based on the flow rate during the green indication of the Icft-tum phase.

Core Motorized Vehide Methodology Chapter 19/5ignalized Intersections

Pilge 19-26 V~6.0
 Highway Capacity Manual: A Guide for Multimoda/ Mobility Analysis

For permittcd left-tum movements operating in an exclusive lane, platoon

ratio is uscd to describe progression quality during the permitted periodo Hencc,
the platoon ratio is bascd on the left-tum tlow rate during the green indication of
the phase providing the permitted opera non.
For permitted or protected-pcrmitted right-tum movements operating in an
exclusive lane, platoon ratio is used to describe progression quality during the
permitted pcriod (even if a protected right-tum opcration is provided during the
complementary left-tum phase on the cross street). Hence, the platoan ratio is
based on the right-turn flow rate during the green indicatian oi the phase
providing the permitted apcration.
For through movements served by exclusive lanes (no shared lanes on the
approach), the platoon ratio for thc thraugh mavement group is based an the
through £low rate during thc green indicatian of the associated phase.
For aH movements served by split phasing, the plataan ratio far a mavement
group is based an its £Iow rate during the green indication af the cammon phase.
Far intersection approaches with one or more shared lanes, ane platoan ratio
is computed for the shared movemcnt group on the basis af the £law rate of aH
shared lanes (plus that of any exclusive through lanes that are also served)
during the green indication of the common phase.
The platoon rano for a movcmcnt group can be estimated {rom {ieIddata by
using Equation 19-5.

Equation 19-5

where
Rp platoon ratio,

P proportion of vehieles arriving during the green indication (decimal),

g effective green time (s), and
e cyele length (s).
Pis computed as the count oi vehicIes that arrive during the green indication
divided by thc count of vchicles that arrive during the entire signaI cyele. It is an
average value representing conditions during the anaIysis periodo

oetermining Platoon Ratio

If the subject intersection is part of a signal system, then the procedure in
Section 3 of Chapter 30, Urban Street Segments: SupplementaI, can be uscd to
estimate the arrival flow proiilc for any approach that is evaluatcd as part of an
urban street segment. The proccdure uses thc flow profile to compute thc
proportion oi arrivals during the green indication. If this procedure is used, then
pIatoon ratio is not an input foc the traffic movements on the subject approach.
If the subject intersection is not part of a signaI system and an existing
intersection is being evaluated, then it is recommended that anaIysts use fieId.
measured values for the variables in Equation 19-5in estimating the pIatoon ratio.

Chapter 1915ignalized intersections (ore Mot:orized Vehicle Methodology

Version 6.0 Page 19-27
Highway Capacity Manual: A Guide tar Multimadal MabiJjty Analysis

1£the subject intersection is not part of a signal system and the anaIysis dea!s
with future conditions, or if the variables in Equation 19-5 are not known from
Held data, then the platoon ratio can be estimated by using guidance provided in
the next subsection.

Guidance for Estimating Platoon Ratio

This subsection provides guidance for estimating arrival type. The platoon
ratio can then be determined from this arrival typc by using Exhibit 19.13.
Exhibit 19-]4 provides guidance for selecting the arriva! typc far through
movements. The guidance considcrs typical signal spacing and the provision of
signal coordination. Arriva! Type 3 is typically assumed for tum movements.

Exhibit 19-14 Arrival TypicalSignal Conditions UnderWhichArriva'Type

ArrivalType5election Type Spadng (ft) 15 Likelyto Occur
Guidelines Coordinatedopel1ltionon a two-waystreet wherethe subject
1 ~1,6oo directiondoesnot receivegoodprogression
2 >1,600-3,200 AlessextremeversionofArrivalType1
J >3,200 lsolatedsignafsor widelyspacedcoordinatedsignals
>1,600-3,200 Coordinatedoperationon a two-waystreetwherethe subject
4
directionreceivesgoodprogression
Coordinatedoperationon a two-waystreetwherethe subject
5 s:1,600 directionreceivesgoodprogression
Coordinatedoperationon a one-waystreet indense network$and
6 s:800 centra!businessdistricts

A description of each arrival type is provided in the following paragraphs to

provide additiona! assistance with the selection of arrival type.
Arrival Type 1 is characterized by a dense platoon of more than 80% of the
movement group volume arriving at the start of the red interva1. This arriva!
type is often associated with short segments with very poor progression in the
subject direction of travel (and possibly good progression for the other direction).
Arrival Type 2 is characterized by a moderately dense p!atoon arriving in the
middle of the red interval or a dispersed plataon containing 40% to 80% of the
movement group valume arriving throughout the red interva1. This arrival type
is oHen associated with segments of average length with unfavorable progression
in the subject dircction of trave!.
Arrival Type 3 describes one oí two conditions.l£ the signals bounding the
segment are coordinated, then this arrival type is characterized by a pIatoon
containing less than 40% oí the movement group volume arriving partly during
the red interval and partly during the green intervaL 1£the signals are not
coordinated, then this arrival type is characterized by plataons arriving at the
subject intersection at different points in time over the course of the analysis
period so that arrivals are effectively random. I
Arrival Type 4 is characterized by a moderately dense plafoon arriving in the
middle of the green interval or a dispersed platoon containing 40% to 80% oí the
movement group volume arriving throughout the green intervaL This arrival

CoreMotorized
VehideMethodology Chapter19/5ignaUzed
Intersections
Page19-28 v~6.0
 Highway Capacity Manual: A Guide tor Multimodal Mobility Analys;s

type is often associated with segments of average length with favorable

progression in the subject direction of travel.
Arrival Type 5 is characterized by a dense platoon of more than 80'}i, of the
movement group volume arriving at the start of the green interva!. This arrival
type is olten associated with short scgments with highly favorable progression in
the subject direction of travel and a low-to-moderate number al side street
entries.
Arrival Typc 6 is characterized by a dense platoon of more than 80% of the
movement group volume arriving at the start of the green interva!. This arrival
type occurs only on very short segments with exceptionally favorable
progression in the subject dircction of travel and negligibJe side street entries. lt
is reserved for routes in dense signal networks, possibly with one-way streets.

Upstream Filtering Adjustment Fador

The upstream filtering adjustment factor 1 accounts lar the eHect of an
upstream signal on vehicle arrivals to the subject movement group. Specifically,
this factor renects the way an upstream signal changes the variance in the
number of arrivals per cyele. The variance decreases with increasing volume-to-
capacity ratio, which can reduce cycle failure frequency and resulting deJay.
The filtering adjustment factor varies from 0.09 to 1.0. A vaJuc of 1.0 is
appropriate lar an isolated interscction (i.e., one that is 0.6 mi or more from the
nearest upstream signalized intersection). A value al less than 1.0 is appropriate
for nonisolated intersections. Equation 19~6is used to compute 1 lar nonisolated
intersections.
1 = 1.0 - 0.91 X~.6B ~ 0.090 Equation 19-6

where
1 upstream filtering adjustment factor, and
X. weighted volume-to-capacity ratio for a11upstream movements
contributing to the volume in the subject movement group.
The variable X. is computed as the weighted volume-to-capacity ratio of aH
upstream movements contributing to the volume in the subject movement
group. This ratio is computed as a 'weighted average with the volume-to-capacity
ratio oi each contributing upstream movement weighted by its discharge
volume. For planning and design analyses, X. can be approximated as the
volume.to-capacity ratio oi the contributing through movement at the upstream
signalized intersection. The value of X"used in Equation 19-6cannot exceed 1.0.

Initia! Queue
The initial queue represents the queue present at the start oi the subject
analysis period for the subject movement group. This queue is created when
oversaturation is sustained for an extended time. The initial queue can be
estimated by monitoring queue count continuously during each of the three
conserutive cycles that occur just befare the start of the analysis periodo The
smallest count ohserved during cach cycle is recorded. Thc initial queue estimate

Chapter 19/5ignalized Intersections Core Motorized Vehicle Methodology

Version 6.0 Page 19-29
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

equals the average of the three counts. The initial queue estimate should not
include vehicles in the queue due to random, cycle-by-cycle fluctuations.
Initial queue has a significant eHect on delay and can vary widely among
intersections and traffic movements. 1fit is not possible to obtain an initial queue
estimate, then the analysis period should be established so the previous period is
known to have demand less than capacity and no residual queue.

Base Saturation Row Rate

The saturation flow rate represents the maximum rate of flow for a traffic
lane as measured at the stop line during the green indication. The base saturation
flow rate represents the saturation flow rate for a traffic lane that is 12 ft wide
and has no heavy vehicles, a flat grade, no parking, no buses that stop at the
intersection, even lane utilization, and no tuming vehicles. TypicalIy, one base
rate expressed in passenger cars per hour per lane is sclected to represent aH
signalized intersections in the jurisdiction (or area) within which the subject
intersection is located. Chapter 31 describes a Held measurement technique for
quantifying the local base saturation flow rate.

Lane Utilization Adjustment Fador

The lane utilization adjustment factor accounts for the unequal distribution
of traffic among the lanes in those movement groups with more than one
exclusive lane. This factor provides an adjustment to the base saturation flow
rate to account for uneven use of the lanes. It is not used unless a movement
group has more than one exclusive lane. The lane utilization adjustment factor is
calculated with Equation 19-7.
Vg
Equation 19-7 tW=fi1)
, g'
where
fw adjustment factor for lane utilization,
vg demand flow rate for movement group (veh/h),
vg1 demand flow rate in the single exclusive lane with the highest flow
rate of all exclusive lanes in movement group (veh/h/1n), and
N, = number of exclusive lanes in movement group (In).
Lane flow rates measured in the Held can be used with Equation 19-7 to
establish local defauIt values of the lane ulilization adjustment factor.
A lane utilization factor of 1.0 is used when a movement group has only one
lane or a uniform traffic distribution can be assumed across all exclusive lanes in
the movement group. Values less than 1.0 apply when traffic is not uniformly
dislributed. As demand approaches capacity, the lane utilization factor is often
closer to 1.0 because drivers have less opportunity to select their lane.
At sorne intersections, drivers may choose one through lane ayer another
lane in anticipatioo af a tuen downstream. When this type of prepositioning
accurs, a more accurate evaluation will be obtained when the actual flaw rate for

Cofe Motorized Vehide Met:hodology Chapter 19/5ignalized Intersectioos

Page 19-30 Vet5kvl6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

each approach lane is measured in the Held and provided as an input to the
methodology (in which case a lane utilization factor of 1.0 is used).
Sorne intersections have an auxiliary through lane (Le., a through-Iane
addition on the approach to an intersection combined with a through-Iane drop
exiting the intersection). This type of lane can be underutilized if it is relatively
short. When it is present on an approach, a more accurate cvaluation will be
obtained when the actual flow rate for each approach lane is provided as input to
the methodology. A procedure for estimating approach lane volumes for this
situation is provided in NCHRP Report 707 (31).
Default Value. The default lane utilization factors described in this subsection
apply to situations in which drivers randomly choose among the exclusive-use
lanes on the intersection approach. The factors do not apply to sredal conditions
(such as short lane drops or a downstream freeway on-ramp) that might cause
drivers intentionaIly to choose their lane position on the basis of an antidpatcd
dawnstream mancuvcr. Exhibit 19-15provides default lane utilization
adjustmcnt factors for different movement groups and numbers of lanes.

No. of Lanes in Traffic in Most Heavily LaneUtilization Exhibit 19-15

Movement Group Movement Group (In) Traveled Lane (%) Adjustment Factor 'tu Default LaneUtilization
1 100.0 1.000 AdjustmentFactors
Exclusivethrough 2 52.5 0.952
3' 36.7 0.908
1 100.0 1.000
ExclusiveIeft tum
2' 51.5 0.~1
1 100.0 1.000
Exclusiveright tum
2' 56.5 0.885
Note: 'If a movement groop has more lanes than shown in ttlis exhiblt, it iS recommeJ'ldedthat f~ld surveys be
condlXted or the smallest l'iu ~alue shown I'or that type of movement groop be used.

As demand approaches capadty, thc analyst may use lane utilization factors
that are doscr to 1.0 than thosc offered in Exhibit 19-15.This re£incment to the
factor value recognizes that a high valume-to-capadty ratio is assodated with a
more uniform use of the available lanes because of rcduccd opportunity far
drivers to select their lane freely.

Pedestrian Flow Rate

The pcdestrian flow rate is bascd on the count of pedestrians traveling in the
crosswalk that is crosscd by vehides tuming right from thc subject approach
during the analysis periodo For example, the pedestrian f1~wrate for the
westbound approach describes thc pedestrian flow in the crosswalk on the north
legoA separate count is taken for each direction of travel in the crosswalk. Each
caunt is divided by the analysis period durabon to yicld a directional hourly
flow rateoThese rates are thcn added to obtain the pedestrian flow rate.

Bicycle Flow Rate

The bicycIe flow rate is based on the count of bicycles whose travel path is
crassed by vehides tuming right from the subject approach during the analysis
periodo These bicycIes may travel on the shaulder or in a bike lane. Any bicyde
traffic operating in the right lane with motorized vehicIe traffic should not be

Chapter 19/5ignalizedIntersections Core MotorizedVehicleMethoclology

Vers.oo 6.0 Page 19-31
Highway Capacify Manual: A Guide for Multimodal Mobility Analysis

included in this count. This interaction is not rnodeled by the methodology. The
count is divided by the analysis period duration to yield an hourly fIow rateo

On-Street Parking Maneuver Rate

The parking maneuver rate represents the eount of influential parking
rnaneuvers that occur on an intersection leg as measured during the analysis
periodo An infIuential maneuver occurs directly adjacent to a movement group,
within a zone that extends from the stop line to a point 250 ft upstream of it. A
maneuver occurs when a vehide enters or exits a parking stall. If more than
180 maneuvers/h exist, then a practical limil of 180 should be used. On a two-
way leg. maneuvers are counted for just the right side of the leg. Do a one-way
leg, maneuvers are separately counted for each side of the leg. The count is
divided by thc analysis period duration to yield an hourIy fIow rateo
Exhibit 19.16 gives default values for the parking maneuver rate on an
interscction approach with on-street parking. It is estimated for a distance of
250 ft back from the stop lineoThe calculations assurne 25 £tper parking space
and 80% occupaney. Each tumover (one car leaving and one car arriving)
generates two parking maneuvers.

Exhibit 19-16 No. of Parking Parking TIme Turnover Rate Maneuver Rate
Default Parking Maneuver Street Type Spaces in 250 ft Limit (h) (veh/h) (maneuvers/h)
Rate 1 1.0 16
Two-way 10
2 0.5 8
1 1.0 32
One-way 20
2 0.5 16

Local Bus 5topping Rate

The bus stopping rate represents the number of local buses that stop and
block traffic f10win a movernent group within 250 £tof the stop line (upstream or
downstream) as measured during the analysis periodo A local bus is a bus that
stops to discharge or pick up passcngers at a bus stop. Thc stop can be on the
near side or the far side of the intersection.1f more than 250 buses/h exist, then a
practicallimit of 250 should be used. The count is divided by the analysis period
duration to yield an hourly f10wrateo

Unsignalized Movement De/ay

The delay £orunsignalized movements at the intersection should be
provided as input to the methodology whenever these movements exist at the
intersection. If provided, they can be included in the calculation of approach
deJay and intersection delay.
The delay will need to be estimated by means extemal to the rnethodology.
These extemal means may indude direct field measurement, observation oí
similar conditions, spedal application of procedures in other HCM chapters, and
simulation. The level oí effort expended for such estimation should be
commensurate with the relevance the unsignalized delay has to the overalI
analysis. For example, high-volume or high-delay movements should be
estimated carefulIy. Free-f1ow right tums can be assumed to have zero delay.

Core Motorized Vehicle Methodology Chapter 1915ignalizedIntersections

Page 19.32 Version 6.0
 Highway Capacify Manual: A Guide far Multimodal Mobilify Analysis

Geometric Design Data

This subsection describes the geometric design data listed in Exhibit 19-11.
These data describe the geometric elements oí the intersection that influence
traific operation.

Number of Lanes
The number of lanes represents the eount of lanes provided for each
intersection traific movement. For a rum movement, this count represents the
lanes reserved for the exclusive use oí turning vehicles. Turn-movement lanes
include tum lanes that extend backward for the length of the segment and lanes
in a turn bayoLanes that are shared by two or more movements are included in
the count of through lanes and are described as s!Jared lanes. If no exclusive tum
lanes are provided, then the turn movement is indicated to have zero lanes.
The number of lanes on an approach depends on approach volume and
signal timing. A single exclusive ¡eh-tum lane is often provided when the left-
tum volume ranges between 100 and 300 veh/h. Similarly, a dual exclusive left-
tum lane is often provided when the lefHum volume exceeds 300 veh/h. An
exclusive right-turn lane is often provided when the right-turn volume exceeds
300 veh/h and the adjacent through volume exceeds 300 veh/h/ln.

Average Lane Width

The average lane width represents the average width of the lanes
represented in a movement group. The minimum average lane width is 8 ft.
Standard lane widths are 12 ft. Lane widths greater than 16 ft can be included;
however, the analyst should consider whcther the wide lane actual1y opera tes as
two narrow lanes. The analysis should reflect the way in which the lane width is
actually used or expected to be used.

Number of Receiving Lanes

The number of receiving Janes represents the count of lanes departing the
intersection. This number should be separately determined for each left-tum and
right-turn movement. Experience indicates praper turning cannot be executed at
sorne intersections because a receiving Jane is frequently bJocked by double-
parked vehicles. For this reason, the number of receiving lanes should be
determined from field observation when possibJe.

Turn Bay Length

Turn bay [ength represents the Jength oí the bay for which the Janes have full
width and in which queued vchicles can be stored. Bay length is measured
parallel to the roadway centerline. 1£there are multipJe Janes in the bay and they
have different lcngths, then the length entered should be an average value.
If a two-way left-turn lane is provided for [eft-turn vehicle storage and
adjacent access points exist, then the bay lcngth entered should represent the
"effective" storage length available to thc l('ft-tum movement. The determination
of effective length is based on consideration of the adjacent access points and the
associated lefHurning vehicles that can be stored in the two-way left-tum lane.

Chapter 19/5ignalized Intersections Core MotOlized Vehicle Methodology

Ver5km 6.0 Page 19-33
Hjghway Capacity Manual: A Guide tor Multimodal Mobility Analysis

Presence of On-Street Parking

The input for presence of on-street parking indicates whether on-street
parking is allowed along the curb Une adjacent to a movement group and within
250 ft upstream of the stop line during the analysis periodo On a two-way street,
the presence of parking is noted for just the right side of the street. On a one-way
street, the presence of on-street parking is separately noted for each side of the
street.

Approach Grade
Approach grade defines the average grade along the approach as measured
from the stop line to a point 100 ft upstream of the stop line along a line parallel
to the direction of travel. An uphill condition has a positive grade, and a
downhill condition has a negative grade.

Signal Control Data

This subsection describes the signal control data listed in Exhibit 19-11 and
Exhibit 19-}2. They are specific to an actuated signal controller that is operated in
a pretimed, semiactuated, fully actuated, or coordinated-actuated manner.

Type of Signal Control

The intersection signal control is an input to the methodology. lt can be
pretimed controlar actuated control. Pretimed control can be describcd as
coordinated (or coordinated-pretimed) if the intersection is part of a signal
system. Actuated control can be describcd as fully actuated, scmiactuatcd, or
coordinated .actuated.
Settings used for coordinated-actuatcd control are describcd latcr in this
subsection. They are used in the motorized vehicle methodology in Chapter 18.
The motorized vehiele methodology is based on the controller functions
defined in the National Transportation Communications for lTS Protocol
Standard 1202 (28). It is incumbent on the analyst to become familiar with these
functions and adapt them, if nceded, to the functionality of the controller used at
the subject intersection. Sernon 2 provides additional information about traffic
signal controller operation.

Phase 5equence
In broad context, phasc sequcnce describes the order of sen'ice provided to
each traffic movement. This definition is narrowed here to limit phase sequence
to the order in which the left-tum movements are served relative to the through
movements. The sequence options addressed in the methodology inelude no Icft-
tum phase, leading left-tum phase, lagging left-tum phase, and split phasing.

Left-Turn Operational Mode

The left-tum operational mode describes how the left-turn movcment is
served by the controller. It can be dcscribcd as permitted, protected, or
protected-permitted.

Core Motorized Vehide Mettlodology Qlapter 19/5ignalized intersections

Page 19-34 Version 6.0
 Hjghway Capacity Manual; A Guide for Multimodal Mobility Analysis

Dalias Left- Turn Phasing Option

The Dalias left-tum phasing opnon allows the left-tum movements to
operate in the protected-permitted mode without causing a "yellow trap" safety
concern. It effectively hes the !eft turn's pennitted-period signal indicanon to the
opposing through movement signal indicanon. This phasing option is also used
with a flashing yellow arrow lefHum signal display.

Passage Time: Aduated Control

Passage nme is the maximum amount of time one vehicle actuation can
extend the green interval while green is displayed. It is input for each actuated
signal phase. It is also referred to as vehiele interval, extension interval,
extension, or unít extension.
Passage time values are typically based on detection zone length, detection
zone location (relative to the stop tine), numbcr of lanes served by the phase,
vehiele length, and vehíele speed. Longer passage times are often used with
shorter detection zones, greater distance between the zone and stop line, [ewer
lanes, smaller vehieles, and higher speeds.
The objective in determining the passage time value is to make it large
enough to ensure all queued vehieles are served but not so large that it extends
for traffic arriving randomly after the queue has eleared. On high-speed
approaches, this objective is broadened to inelude not making the passage time
so long that the phase frequently extends to its maximum setting (i.c., max-out)
so that safe phase termination is compromised.

Maximum Green: Actuated Control

The maximum green setting defines the maximum amount of time a green
signal indication can be displayed in the presence of conflicting demando Typical
maximum gno-'('nvalues [or left-tum phases range fram 15 to 30 s. Typical valucs
for through phases serving the minor-street approach range fram 20 to 40 s, and
values for through phascs scrving the major-street appraach range fram 30 to 60 s.
For an analysis of coordinatcd-actuated operation, the maximum green ís
disabled thraugh the inhibit mode, and the phase splits are used to determine the
maximum length of the actuated phases.

Green Duration: Pretimed Control

For an analysis of pretimed operanon, the green interval durabon is an input
to the methodology. Typical valucs are similar to those noted aboye for thc
maximum green scUing. A procedure for estimating pretimed green interval
duranon is described in Section 2 of Chapter 31, Signalized Intersections:
Supplemental.

Minimum Green
The minimum green setting represcnts the least amount of time a green
signal indication is displayed whcn a signal phase is activated. Its duration is
bascd on consideration of driver reaction time, queue size, and driver
expectancy. Minimum green typically ranges from 4 to 15 s, with shorter values

Qlapter 19/5ignalized lntersections eore Motorized Vehicle Methodology

Version 6.0 Page 19-35
Highway Capadty Monual: A Guide tar Mulfimodal Mobilify Analysis

in this range used for phases serving tum movements and lower-volurne
through rnovements. Foc intersections without pedestrian push buttons, the
rninimum green setting may also need to be long enough to allow time for
pedestnans to reaet to the signar indieation and eross the street.

Yel/ow Change and Red C1earance

The yellow ehangc and red clearanee settings are input for eaeh signal phase.
The yellow ehange interval is intended to alert a driver to the impending
presentation of a red indication. lt ranges from 3 to 6 s, with longer values in this
range used with phases serving high-spccd movements.
The red clearanee interval can be used to allow a brief time to elapse after the
yellow indieation, during whieh the signal heads associated with the ending
phase and all eonflicting phases display a red indication. lf used, the red
clearanee interval is typieally 1 or 2 s.

Walk
The walk interval is intended to give pedestnans adeguate time to pereeive the
walk indication and depart the eurb before the pedestrian clear interval begins.
For an actuated or noneoordinatcd phase, the walk interval is typically set at
the mirumum value needed for pedestrian pereeption and curb departure. Many
agencies eonsidcr this value to be 7 s; however, sorne agencies use as little as 4 s.
Longer walk durations should be eonsidered in sehool zones and areas with
!arge numbers of elderly pedestrians. The methodology assumes the rest-in-walk
mode is not enabled foc actuated phases and noneoordinated phases.
For a pretimcd phase, the walk interval is often set at a value egual to the
green interval duration needed for vehicle service Iess the pedestrian clear
sctting (provided the resulting interva! exceeds the minimum time ncedcd for
pcdestrian pcrception and curb departure).
For a eoordinated phase, the controller is sometimes set to use a coordination
mode that extends the wa!k interval for most of the green interval duration. This
functionality is not explicitly modeled in the motorized vehicle methodology, but it
can be approximated by setting the walk interval to a value equal to the phase split
minus the sum of the pcdestrian c1ear,yellow change, and red c1earaneeinterva!s.
If the walk and pedestrian clear settings are provided for a phase, then it is
assumed a pedestrian signal head is also provided. lf these settings are not used,
then it is assumed any pedestrian aeeommodation nl."Cdedis provided in the
minimum green setting.

Pedestrian C1ear
The pedestnan clear intcrval (also referred to as the pedestrian ehangc
interval) is intended to provide time for pcdestrians who depart the curb during
the WALKindieation to reaeh the opposite curb (or the median). Sorne agencies
set the pedestrian c1earegua! to the "erossing time," where erossing time equa!s
the curb-to-curb erossing distanee divided by the pcdestrian wa!king speed of
3.5 ft/s. Other agencies set the pedestrian c1earequal to the erossing time !ess the
vehic1eehangc penod (Le.,the eombined yellow ehange and red clearanee

CUre Motorized Vehicle Methodology Chapter 19/5ignalized Intersections

Page 19-36 Vern::v, 6.0
 Highway Capacity Manual: A Guide tor Mu/timodal Mobility Ana/ysis

intervals). This choice depends on agency policy and practice. A f1ashing OON'T
WALKindication is displayed during this interval.

Phase Recal!o'Actuated Control

H used, recan causes the controller to place a call for a specified phase each
time the controller is servicing a conflicting phase. It is input for each signal
phase. Threc types of recalls are modeled in the motorized vehicle methodology:
minimum recall, maximum recan, and pedestrian recal!.
Invoking minimum recall causes the controller to place a continuous call for
vehicle service on the phase and then sen'ice thc phase until its minimum grecn
interval times out. The phase can be extended if actuations are received.
Invoking maximum recall causes the controller to place a continuous caH for
vehicle service on the phase. It results in presentation of the green indication foc
its maximum duration every cyeIe. Using maximum recall on all phases yields an
equivalent pretimed operation.
Invoking pedestrian recalI causes the controller to place a continuous call foc
pedestrian service on the phasc and then scrvice the phase foc at least an amount
of time equal to its walk and pedestrian clear intervals (Ionger if vehicle
detections are received). Pedestrian recaH is used for phases that have a high
probability of pedestrian demand every cycle and no pedestrian detection.

Dual Entry: Actuated Control

The entry mode is used in dual-ring operation to spccify whether a phase is
to be activated (grecn) even though it has not received a call for service. Two
entry modes are possibk dual entry and single entey. This mode is input for
eaeh actuated signal phase.
A phasc operating in dual entry is avaHable to be callcd by the controller,
even if no actuations have been received for this phase. A phasc operating in
single entry wil! be ealled only if actuations have been received.
During the timing of a eycle, a point is rcached at which the next phase (or
phases) to be timed is on the other side of a barrier. At this point, the eontroller
wiIl check the phases in eaeh ring and determine which phase to activate. If a call
does not exist in a ring. the controller will activate a phase designated as dual
entry in that ringo If two phases are designated as dual entey in the ring. then the
(irst phase to occur in the phase sequence is activated.

Simultaneous Gap-Out: Actuated Control

The simultaneous gap-out mode affects the way actuated phases are
terminated befare the barricr can be crossed to serve a conflicting call. This mode
can be enabled oc disabled. It is a phasc-specific setting; however, it is typically
set the same foc all phases that serve the same street. This mode is input for each
actuated signal phase.
Simultaneous gap-out dicta tes controller operation when a barrier must be
crossed to serve the next call and one phase is active in eaeh ringo If simultaneous
gap-out is enabled, both phases must reach a point of being committed to
termina te (via gap-out, max-out, or force-off) at the same time. If one phase is

Chapter 1915ignalized Intersections (ore Motorized Vehide Methodology

Verwn6.0 Page 19-37
Hjghway Capacity Manual: A Guide for Multimodal Mobjlity Analysis

able to terminate because it has gapped out, but the other phase is not able to
termina te, then the gapped-out phase will resct its extension timer and restart the
process oi timing down to gap-out.
If the simultaneous gap-out ieature is disabled, then each phase can reach a
point oi termination independently. In this situation, the first phase to eommit to
termination maintains its aetive status while waiting Eor the other phase to
commit to termination. RegardIess oE which mode is in effect, the barrier is not
crossed until both phases are committed to termina te.

Cycle Length: Coordinated-Actuated or Pretimed Control

Cyele length is the time elapscd bctwecn the endings of two sequential
presentations oi a coordinated-phase green interva1. A eyele length is needed ior
pretimed control and for coordinated-actuated control.
Default Value. The cyele length used for a coordinated signaI system often
represents a compromise value based on interscction eapacity, queue size, phasc
sequence, segment length, speed, and progression quality. Consideration of these
factors leads to the default eyele lengths shown in Exhibit 19-17.

Exhiblt 19-17 Cvcle LenQ1:hby Street CJass and Left-Turn Pha5ina (5)b'
Default 5ystem Cycle Length Minor Arterial Street
Major Arterial Street or Grid Network
Average Left-Tum Left-Tum Left-Tum Left-Turn
5egment No Left. Phases on Phases on No Left. Phases on Phases

~~~~h
Turo O". 80th Tnm O".
on 80th
Phases Street streets Phase5 Street Streets
1,300 90 120 150 60 80 120
2,600 90 120 150 100 100 120
3,900 110 120 150
NOtes: • AI'eI'age Iengttl based 00 all street segments In the signal system .
• Selected left.turn phasing column should describe the pnase sequence at the high.lIOlume Intersectlons in
!he system.

Phase Splits: Coordinated-Actuated Control

Each noncoordinated phase is provided a "split" time that represents the
suro of the greeo, yellow change, and red elearanee intervals for the phase. The
ratiooale for determining the green interval duration varies among agencies;
however, it is often related to the "optimum" pretimed grecn interval duration.
Seetion 2 oi Chapter 31 provides a proccdure for determining pretimed phase
duration.
If the phase splits are not known, they can be estimated by using the
planning-level analysis application described in Chapter 31. The method can be
used to estimate the effective green time for eaeh phase on the basis of the
established system eyele length. The phase split Dp is then computed by adding
4 s oi lost time to the estimated effective grt..'t'ntime (Le., Dp "" g + 4.0).

Offset and Offset Reference Point: Coordinated Control

The referenee phase is specified to be one of the h'lO eoordinated phases (Le.,
Phase 2 or 6). The offset entered in the controller represents the time the
referenee phase begins (or ends) relative to the system master time zero. The
offset must be specified as being refereneed to the beginning (or the end) of the

Core Motorized Vehicle Methodology Olapter 19/5ignalized Intersections

Page 19-38 Version 6.0
 Highway Capacify Manual: A Guide tor Multimodal Mobilify Analysis

green interval of the reference phase. The offset reference point is typically the
same at all intersections in a given signal system.

Force Nade: Coordinated-Actuated Control

The force mode is a controller-specific setting. It is set to "fixed" or
"floating." The controller calcula tes the phase force-off point for each
noncoordinated phase on the basis of the force mode and the phase splits. When
set to the fixed mode, each noncoordinated phase has its force-off point set at a
fixed time in the cyde rclative to time zero on the system master. This operation
allows unused spJit time to revert to the following phase. When set to the
floating mode, each noncoordinated phase has its force-off point set at the spJit
time after the phase first becomes active. This operation allows unused split time
to revert to the coordinated phase (referred to as an "early retum to green").

Other Data
This subsedion describes the data listed in Exhibit 19-11that are categorized
as "other" data.

Analysis Period Duration

The analysis period is the time interval considered for the performance
evaluation. lts duration ranges from 15 min to 1 h, with longer durations in this
range sometimes used for planning analyses. In general, the analyst should use
caution in interpreting the results from an analysis period of 1 h or more because
the adverse impact of short peaks in traffic demand may not be detected.
Any 15-min period of interest can be evaluated with the methodology;
however, a complete evaluation should always inelude an analysis of conditions
during the 15-min period that experiences the highest traffic dcmand during a
24-h periodo
Operational analyscs require a 15-min analysis period. This durabon will
accurately capture the adversc effects of demand peaks.
Most planning analyses use a 15-min analysis periodo However, a l-h
analysis period can be used, if appropriate.

Speed Umit
The methodology is based on the assumption that the posted speed limil is
(a)consistent with posted speed limits found on other streets in the vicinity of
the subject intersection and (b) consistent with agency policy regarding
specification of speed Iimits. If it is known the posted speed limit does not satisfy
these assumptions, then the specd ¡imit value that is input to the methodology
should be adjusted so it is consistent with thc assumptions.

Stop-Une Detector Length and Detection Hade: Aduated Control

The stop-Jine detector length represents the length of the detection zone used
to extend the green indication. This detection zone is typically ¡ocated near the
stop line and may have a length of 40 ft or more. However, it can be 10cated some
distance upstream of the stop tine and may be as short as 6 ft. The latter

Chapter 19/5ignalized lntersections Core Motorized Vehicle Methodology

VersiQn 6.0 Page 19-39
Highway Capacity Manual: A Guide tor Multimodal Mobility Anolys;s

configuration typically requires a long minimum green or use of the controller's

variable initial setting.
If a video-image vehicle detection system is used to provide stop-line
detection, then the input length should reflect the physicallength oí roadway
that is monitored by the video detection zone plus a length of 5 to 10 ft to account
for thc projection of the vehicle image into the plane of the pavement (with larger
values in this range used for wider intersections).
Detection mode influences thc duration of the actuation submitted to the
controller by the detcction unit. One of two modes can be used: presence or
pulse. Presence mode is typically the default mode. lt tends to provide more
reHable intersection operation than pulse mode.
In the presence mode, the actuation starts with the vehicle arriving in the
detection zone and ends with the vehicle leaving the detection zone. Thus, thc
time duration of the actuation depends on vehicle length, detection zone length,
and vehicle specd.
The presence mode is typically used with long detection zones located at the
stop lineoThe combination typically results in the necd for a small passage-time
value. This characteristic is desirable because it tends to result in effident queue
service.
In the pulse mode, the actuation starts and ends with the vehicle arriving at
the detector (actually, the actuation is a short "on" pulse of 0.10 to 0.15 s). This
mode is not used as olten as presence mode for intersection control.

Area Type
The area type input indica tes whether an intersection is in a central business
district (CSO) type of environment. An intersection is considered to be in a CBO,
or a similar type of area, when its characteristics include narrow street rights-oí-
way, frequent parking maneuvers, vehicle blockages, taxi and bus activity, small-
radius turos, Iimited use oí exclusive turo lanes, high pedestrian activity, dense
population, and midblock curb cuts. The average saturation headway at
intersections in areas with CBO-like characteristics is significantly longer than at
intersections in areas that are less constrained and less visually intense.

OVERVIEW OF THE METHODOLOGY

This subsection provides an overview of the methodology for evaluating the
performance of the signalized intersection in terms of its service to motorjzed
vehicles. The methodology is computationally intense and is most effidently
implemented using software. The intensity stems partly from the need to model
traffic-actuated signal operation.
A planning-level analysis application íor evaluating interscction performance
is provided in Chapter 31, Signalized Intersections: Supplemental. This method
is not computationally intense and can be applied by using hand calculations.
The objective of this overview is to introduce the analyst to the calculation
process and discuss the key analytic procedures. This objective is achieved by
focusing the discussion on lane groups that serve one traffic movement with

Core Motorized Vehide Methodology Chapter 19/5ignalized Intersectioos

Page 19-40 VeNion 6.0
 Highway Capacity Manual: A Guide for Mu/tjmodal Mobi/ity Analysis

pretimed control and for which there are no permiUed or protected-permitted

left.tuen movements. Details on evaluation of actuated control, shared-lane lane
groups, and interscctions with permitted or protected-pcrmitted left-turn
operation are provided in Chapter 31.
The computational engine developed by the Transportation Research Board
Committee on Highway Capacity and Quality of Service represcnts the most
detailed dcscription of this methodology. Additional information about this
engine is provided in Chapter 31.

Framework
Exhibit 19-18illustrates the calculation framework of the motorized vehicle
methodology. H identifies the sequence of calculations needed to estimate
selected performance measures. The calculation process is shown to £Iowfrom
top to boUom in the exhibit. These calculations are described more fully in thc
next scction.
The methodology adopts the same principies of critical movement analysis
used in prior editions of the HCM. The first stcp of the methodology is used to
determine the lane groups associated with each intersection approach. These lane
groups represent the basic unit of analysis. Each lane group is separately
evaluated, and the results are aggregated to the approach and intersection levels.
The second and third steps are used to determine how the left-tuen, through, and
right-tum drivers on each intersection approach distribute themselves among the
lane groups. The fourth step is used to predict the saturation flow rate for each
lane group based on prevailing conditions. The fifth step is uscd to quantify thc
eHect of upstream signals on the arrival flow rate for each Jane group. If a phase
is actuated, the sixth step is needed to estimate the average duration of this
phase. In the seventh step, lane group capacity is evaluated in terms of the ratio
of flow rate to capacity. This ratio is used in 5tep 8 to estimate the control delay
for each lane group. This estimated control dclay is used in Step 9 to estimate the
LOS for cach lane group, approach, and interscction. The tenth step can be
optionally used to estimate lane group queue length and storage ratio.
For actuated control, the methodology is shown to be iterative within Steps 3
to 6, with convergence achieved when the predicted phase duration and capacity
from successive iterations are eHectively in agreement. Before the first iteration,
an initial rough estimate of the phase duration is made to support the
calcuJations in 5teps 3 to 5. A reviscd estímate of phase duration is produced in
Step 6. The revised estimate is compared with the previous estimate and, if they
are not in agreement, the process is repeated until convergence is achieved.
Several iterations are typically needed.
Although not shown in Exhibit 19.18, 5tep 3 includes an iterative procedure
that is used when one or more ¡ane groups have a shared lane. This procedure
allocates the through volume among the available shared and exclusive through
lanes to determine thc lane group volume assignment that produces the 10west
service time. It is implemented for both pretimed and actuated control.

Chapter 19/5ignalized Intersections Core Motorized Vehícle Methodology

VeFSiOn 6.0 Page 19-41
Highway Capadty Manual; A Guide tor Multimodal Mobility Analysis

Exhibit 19.18 Pretimed Aetuated

Motorized Vehide
Methodology for Signalized
intersections
J
Step 1. Determine Movement Groups and lane Groups

Step 2. Determine Movement Group Aow Rate

Step 3. Determine Lane Group Aow Rate

Step 4. Determine Adjusted Saturation Flow Rate

Step S. Determine Proportion Arriving During Green

Step 6. Determine Signal Phase

Duration

No

Ves
Step 7. Determine Capadty and Volume-to-Capadty Ratio

Step 8. Determine Delay

5tep 9. Determine lOS

Step 10. Determine Queue Storage Ratio

COMPUTATlONAL STEPS

Step 1: Determine Movement Groups and Lane Groups

The movement groups and lane groups are established during this step.
They are established separately for each intersection approach. Rules for
establishing these groups are described in the subsequent paragraphs. Exhibit 19-
19 shows some common movement groups and lane groups. A discussion of the
need for, and difference between, movement groups and lane groups is providcd
in a previous subsection titled Lane Groups and Movement Groups.

Core Motorized Vehicle Methodology Chapter 19/5ignalized Intersections

Page 19-42 Verslon 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

"'""'.
of Lanes Moyements by tanes MoyementGrou~(~ tane Grou ,fLGl
Exhibit 19-19
Typical Movement Groups and
Lane GfOlJpS
1 len. ttlrough, and right: MG1: LG 1:
~ ~ ~

2
bdusi>e left: ---' MG 1: ~ lG 1: ---'
Through and right: MG2: lG 2:
"" "" ""
left and lt1mugh: ¿ ¿ LG 1:
¿
2 MG 1:
Ttvough and r;ght: "" LG 2:
"" ""
bdusi>e left:
Exclusi>e lett:
~
~
MG1: ~ lG 1:
=:3
3 Through: •, ~
~ lG2: ••
TtvOll9h: MG2:
~
Through and right: LGJ;
"" "'"

Determine Movement Groups

The following rules are used to determinc movement groups for an
intersection approach:
• A tum movemcnt served by one or more exclusive lanes and no shared
lanes should be designated as a movcment group.
• Any lanes not assigned to a group by the previous rule should be
combined into onc movement group.
Thcse rules result in the designation of one to thrcc movement groups for
each approach. A movement group can include one or morc lanes.

Determine Lane Groups

A lane group can include one or morc lanes. The following rules are used to
determine lane groups for an interscction approach:
• An exclusive left-turn lane or lanes should be designated as a separate
lane group. Thc samc is true of an exclusive right-turn lane.
• Any shared lane should be dcsignated as a separate lane group.
• Any lanes that are not exclusive tum lanes or shared lanes should be
combined into one lane group.
These rules result in the designation of one or more of thc following lane
group possibilities for an intersection approach:
• Exclusive left-turn lane (or lancs),
• Exclusive through lanc (or lanes),
• Exclusive right-turn lane (or lanes),
• Sharcd ¡eft-tum and through lane,
• Shared leH-turn and right-turn lanc,
• Shared right-turn and through lane, and
• Shared leh-tum, through, and right-turn lanc.

Chapter 19/5ignalized Intersections Core Motorized Vehicle Methodology

Version 6.0 Page 19-43
Highway Capacity Manual; A Guide tor Multimodal MobiJjty Analysis

Step 2: Determine Movement Group Flow Rate

The flow rate for each movement group is determined in this step. lf a tuen
movement is served by one or more exclusive lanes and no shared lanes, then
that movement's flow cate is assigned to a movement group for the exclusive
lanes. Any of the approach flow that is yet to be assigned to a movement group
(after application of the guidance in the previous sentence) is assigned to one
movement group.
The RTOR flow rate is subtracted from the right-tuen flow rate, regardless of
whether the right tuen occurs from a shared or an exclusive lane. The reduced
right-tum volume is used to compute capacity and LOS in subsequent steps.

Step 3: Determine Lane Group Flow Rate

The lane group flow rate is determined in this step. lf there are no shared
lanes on the intersection approach or the approach has only one lane, there is a
one-to-one correspondence between lane groups and movement groups. In this
situation, the lane group fIow rate equals the movement group flow rateo
If there are one or more shared lanes on the approach and two oc more ¡anes,
then the lane group flow rate is computed by the procedure described in
Section 2 of Chapter 31 in the subsection titled Lane Group Flow Rate on
Multiple-Lane Approaches. This procedure is based on an assumed desire by
drivers to choose the lane that minimizes their service time at the intersection,
where the lane volume-to-saturation fIow ratio is used to estimate relative
differences in this time among lanes. This assumption may not always hold for
situations in which drivers choose a lane so they are prepositioned for a tum at
the downstream intersection. Similarly, it may not hold when an auxiliary
through ¡ane is present. In either situation, the analyst needs to provide the fIow
cate for each lane on the approach and then combine these rates to define
explicitly the flow rate for each lane group.

Step 4: Determine Adjusted Saturation Flow Rate

The adjusted saturation flow rate for each lane of each lane group is
computed in this step. The base saturation flow rate provided as an input
variable is used in this computation.
The computed saturation fIow rate is referred to as the adjusted saturation
flow rate because it reflects the application of various factors that adjust the base
saturation flow rate to the spedfic conditions present on the subject interscction
approach.
The procedure described in this step applies to lane groups that consist of an
exclusive lane (or lanes) operating in a pretimed protected mode and without
pedestrian or bicycIe interaction. When these conditions do not hold, the
supplemental procedures described in Sections 2 and 3 of Chapter 31 should be
combined with the procedures in this step to compute the adjusted saturation
flow rateo
Equation 19-8 is used to compute the adjusted saturation flow rate per lane
for the subject lane group.
Equation 19-8 s = So fw fHVg fp fM fa fw fLT fRT fLpb fRpb fwz fms fsp

Core Motorized Vehide Methodology Olapter 19/5ignalized lntersections

Page 19-44 Version 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

wherc
s '" adjusted saturation flow rate (veh/h/ln),
s, base saturation fiow rate (pc/h/ln),
adjustment factor lor lane width,
!"
¡HV~ adjustment factor for heavy vehicles and grade,

!, adjushncnt factor for existence of a parking lane and parking activity

adjacent to lane group,

!. adjustment factor for blocking effect of local buses that stop within
intersection area,

!. adjustment factor for area type,

!w adjustment factor for lanc utilization,

fe, adjustment factor for left-turn vehicle presence in a lane group,

adjustment factor for right-turn vehicle presence in a lane group,

!" "
!,,,. pedestrian adjustment factor for left turn groups, 4

!•• pedestrian-bicycle adjustment factor for right-turn groups,

!~ adjustment factor for work zone presence at the intersection,

!•. adjustment factor for downstrcam lane blockage, and

¡sp '" adjustment factor for sustained spillback.

The adjustment factors in the list aboye are described in the following
subsections.

Base Saturation Flow Rate

Computations begin with selection of a base saturation flow rateoThis base
rate represents the expected average flow rate for a through-traffic lanc having
geometric and traffic conditions that correspond to a value of 1.0 for each
adjustment factor. Typically, one base rate is selected to represent a11signalized
intersections in the jurisdiction (or area) within which the subject intersection is
localed. Dcfault values for this rate are provided in Exhibit ]9.11.

Adjustment for Lane Width

The lane width adjustment factor ¡'" accounts for the negative impact of
narrow lanes on saturation flow rate and a110wsfor an incrcascd flow rate on
wide lanes. Values of this factor are listed in Exhibit 19-20.

Average Lane Width (ft) Adjustment Factor f", Exhibit 19-20

<1O.0~ 0.96 lane Wiclth Adjustment Factor
~10.0-12.9 1.00
>12.9 1.04
Note: • Faetors apply to a~erage lane widths of 8.0 ft or more.

Standard lanes are 12 ft wide. The lane width factor may be used with
caution for lanc widths greater than 16 ft, or an analysis with two narrow lancs

Chapter 19/5ignalized Intersections Core Motorized Vehicle Methodology

v~6.0 Page 19-45
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

may be conducted. Use of hvo narrow lanes will always result in a higher
saturation flow rate than a single wide lane, but, in either case, the analysis
should refIect the way the width is actually used or expected to be used. In no
case should this factor be uscd to estimate the saturation fIow rate of a lane
group with an average lane width that is less than 8.0 ft.

Adjustment for Heavy Vehicles and Grade

The heavy.vehicle and grade adjustment factor IINg accounts for the
combined effect of heavy vehicle and approach grade on saturation fIow rate.
The heavy-vehicle component of this factor accounts for the additional space
occupied by heavy vehicles and for the difference in their operating capabilities
compared with passenger cars. The grade component accounts for the effects of
approach grade on vehicle performance. An uphill grade has a positive value
and a downhill grade has a negative value.
If the grade is negative (Le., downhill), thcn thc factor is computed with
Equation 19-9.
100 - 0.79 PHV - 2.07 Po
Equation 19-9
fHVg = 100
If the grade is not negative (Le., level or uphill), then the factor is computed
with Equation 19-10.

Equation 19.10
100 - 0.78 PHV - 0.31 pi
fHVg = 100
where
PI/V percentage heavy vehicles in the corresponding movement group (%),
and

Pg '" approach grade for the corresponding movement group (%).

This factor applies to heavy vehicle percentages up to 50% and grades
ranging from -4.0% to +10.0%. This factor does not address local buses that stop
in the intersection area.

Adjustment for Parking

The parking adjustment factor Ir accounts for the frictional effect of a parking
lane on flow in the lane group adjacent to the parking lane.1t also accounts for
the occasional blocking of an adjacent lane by vehicles moving into and out of
parking spaces. If no parking is present, then this factor has a value of 1.00. If
parking is present, then the value of this factor is computed with Equation 19-11.

N - O1- 18 Nm
Equation 19-11 fp = ' N 3.600 ;" 0,050

where
Nm '" parking maneuver rate adjacent to lane group (maneuvers/h), and
N number oí lanes in lane group (In).

CoreMotonzedVehicleMethodology Chapter19/5ignallzedIntersections
Page19-46 Version 6.0
 Highway Capacity Manual: A Gujde for Multimodal Mobility Analysis

The parking maneuver rate corresponds to parking areas directly adjacent to

the lane group and within 250 ft upstream of the stop lineoA practical upper
limit of 180 maneuversjh should be maintained with Eguation 19-11.A minimum
value offr from this eguation is O.OSO. Each maneuver (either in or out) is assumed
lo block traffic in the lane next to the parking maneuver for an average of 18 s.
The factor applies only to the lane group adjaeent to the parking. On a one-
way street with a single-Iane lane group, the number of maneuvers used is the
total for both sides of the lane group. On a one-way street with two or more ¡ane
groups, the factor is calculated separately for each lane group and is based on the
number of maneuvers adjacent to the group. Parking conditions with zero
maneuvers have an impact different from that of a no-parking situation.

Adjustment for Bus Blockage

The bus-blockage adjustment factor f/ll¡ accounts for the impaet of local transit
buses that stop to discharge or pick up passengers at a near-side or far-side bus
stop within 250 ft of the stop line (upstream or downstream). Values of this factor
are computed with Eguation 19-12.
N _14.4Nb
,_ 3,600;" 0.050 Equation 19-12
Ibb - N
where N is the number of lanes in lane group (In), and N~is the bus stopping rate
on the subject approach (buses/h).
This factor should be used only when stopping buses block traffic flow in the
subject lane group. A practical upper limit of 2SO buses/h should be maintained
wHh Eguation 19-12. A minimum value off N! from this eguation is 0.050. The
factor used here assumes an average blockage time of 14.4 s during a green
indication.

AdJustment for Area Type

The area type adjustment factor J. accounts foc the inefficiency of
intersections in eBOs relative to those in other loeations. When used, it has a
value of 0.90.
Use of this factor should be delermined on a case-by-case basis. This factor is
not Iimitcd to designated eBO areas, noc does it need to be used for aH eBO
arcas. Instead, it should be uscd in arcas where the geometric design and the
traffic or pedestrian £lows, or both, are such that vehiclc headways are
significantly increased.

Adjustment for Lane Utilization

The input lane utilization adjustmcnt factor is used to estimate saturation
£low rate for a lane group with more than one exclusive lane. If thc lane group
has one shared lane or one exclusive lane, then this factor is 1.0.

O1apter 19j5ignalized Intersections Core Motolized Vehide Methodology

Version 6.0 Page 19-47
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Adjustment for Right Turns

The right-turn adjustment factor fn is intended primarily to reflect the effect
of right-turn path geometry 00 saturation flow rateoThe value of this adjustment
factor is computed with Equation 19-13.
1
Equatlon 19-13 fRT =-
ER
where ER is the equivalent number of through cars for a protected right-tuming
vehicle (= 1.18).
If the right-tum movement shares a lane with another movement or has
permitted operation, then the procedure described in $ection 3 of Chapter 31
should be used to compute the adjusted saturation fIow rate for the shared-Iane
lane group. The eHect of pedestrians and bicycles on right-turn saturation flow
rate is considered in a separate adjustment factor.

Adjustment for Left Turns

The lefHum adjustment factor fu is intended primarily to reflect the eHect of
left-tum path geometry on saturation f10wrateoThe value of this adjustment
factor is computed with Equation 19.14.
1
Equation 19-14 fLT =-
EL
where EL is the equivalent number of through cars for a protected left-tuming
vehicle (= 1.05).
If the left-tum movement shares a lane with another movement or has
permitted operation, then the procedure describcd in Section 3 of Chapter 31
should be used to compute the adjusted saturation fIow rate for the shared-lane
lane group. The eHect of pedestrians on left.turn saturation flow rate is
considered in a separate adjustmcnt factor.

Adjustment for Pedestrians and Hicye/es

The procedure to determine the left-turn pedestrian-bicycle adjustment
factor f/.pl> and the right-turn pedestrian-bicycle adjustment factor fRpb is based on
the concept of conflict zone occupancy, which accounts for the confliet between
turning vehicles, pedestrians, and bicycles. Relevant conflict zone occupancy
takes into account whether the opposing vehicle flow is also in conflict with the
left-tum movement. The proportion of grecn time in which the conflict zone is
occupied is determined as a function of the relevant occupancy and the number
of receiving lanes for the turning vehicles. A procedure for computing these
factors is provided in SecHon 2 of Chapter 31. These factors have a value of 1.0 if
no pedestrians or bicycles are present.

Adjustment for Work Zone Presenee

The adjustment factor for work zone presencef"" is used to evaluate the
eHect of work zone presence on saturation £low rate. This factor addresses the
case in which the work zone is located on the intersection approach. The work
zone is considered to be on the intersection approach if sorne (or all) of the work

Core Motorized Vehicle Methodology Chapter 19/5ignalized lntersections

Page 19-48 V£'fSiOn 6.0
 Highway Capacity Manual: A Gujde for Multimodal Mobility Analysis

zone is loeated betwcen the stop [ine and a point 250 ft upstream of the stop lineo
A procedure for computing this factor is provided in Section 2 ofChapter 31,
Signalized Intersections: Supplemental. The factor has a value of 1.0 if no \vork
zone is presento

Adjustment for Downstream Lane Blockage

The adjustment factor for downstream lane bloekage loo, is used to evaluate
the eHect of a downstream lane c10sure on saturation flow rateo A downstrcam
lane c10sure is a c10sure loeated downstream of the subject intersection. The
factor is applied only to those lane groups entering the segment on which the
dosure is presento The lane c10sure can be associatcd with a work zone or spedal
evento A procedure for computing this factor is provided in Section 3 of Chapter
30, Urban Street Segments: SupplementaL The factor has a value of 1.0 if no
downstream lane bloekage is presento

Adjustment for Sustained Spillback

The adjustment factor for sustained spillbaek I'P is used to evaluate the eHect
of spillback from the downstrcam intcrsection. When spillbaek oerurs, its eHect is
quantified as a reduction in the saturation flow rate of upstream lane groups
entering the segmento A proeedure is describcd in Section 3 of Chapter 29, Urban
Street Facilities: SupplementaL for evaluating urban strect facilities that
experience spillback on one or more segments during the analysis periodo The
calculation of the adjustment factor for spillback is one part of this proeedure.
The factor has a value of 1.0 if no spillback occurs.

Step S: Determine Proportion Arriving During Green

Control delay and queue size at a signalized intersection are highly
dependent on the proportion of vehic1es that arrive during the green and red
signal indications. Delay and queue size are smaller when a larger proportion of
vehides arrive during the green indication. Equation 19-15is used to compute
this proportion for each lane group.
P = Rp(g¡C) Equation 19-15

where aH variables are as previously defined.

Equation 19-15requires knowledge oí the effective green time g and cyc1e
length C. These values are known íar pretimed oper,ltion. If the intersection is
not pretimed, then the average phasc time and eyele length must be calrulated
by the procedures describcd in the next step.
A procedure is deseribed in Section 3 oí Chapter 30 th,lt can be used to
estimate the arrival flow profile for ,ln intersection approach when this ,lppro,lch
is evaluated as part of ,ln urb,ln street segmento The procedure uses the profile to
compute the proportion oí arriv,lls during the grt."Cnindication.

Step 6: Determine Signal Phase Duration

The duration of a signal phase depends on the type of control used at the
subject intersection. If the intersection has pretimed control, then the phase
duration is an input, and the evaluation continues with Step 7. If the phase

Chapter 19/5ign,llized intersections Core Motorized Vehicle Methodology

Version 6.0 Page 19-49
Highway Capacity Manual; A Guide for Multimodal Mobility Analysis

duration is unknown, then the pretimed phase duration procedure in $ection 2 of

Chapter 31 can be used to estimate the pretimed phase duration.
If the intersection has actuated control, then the actuated phase duration
procedure in $cction 2 of Chapter 31 is used in this step to estimate the average
duration of an actuated phase. This procedure distinguishes between actuated,
noncoordinated, and coordinatcd control types.

Step 7: Determine Capacity and Volume-to-Capacity Ratio

The capacity of a given lane group serving one traffic movement, and for
which there are no permitted left-tum movements, is defined by Equation 19-16.

c=Ns~
Equation 19-16 e
where e is the capacity (veh/h), and all other variables are as previously defined.
Equation 19-16 cannot be used to calculate thc capacity of a shared-lane lane
group or a lane group with permitted operation because thesc lane groups have
other factors that affect their capacity. Chapter 31 provides a procedure for
estimating the capacity of these types of lane groups.
The volume-to.capacity ratio for a lane group is defined as the ratio of the
lane group volume and its capacity.1t is computed with Equation 19.17.
v
Equation 19.17 x=-e
where
X volume.to.capacity ratio,
v = demand flow rate (veh/h), and
e = capacity (veh/h).
The critical intcrsection volume-to.capadty ratio is also computcd during
this step. Guidelines for computing this ratio are providcd in Section 4,
Extensions to the Motorized Vehicle Methodology.

Step 8: Determine Delay

The delay calculated in this step represents the average control delay
experienced by aH vehieles that arrive during the analysis periodo It ineludes any
delay incurred by thesc vehieles that are still in queue after the analysis period
ends. The control deJay for a given lane group is computed with Equation 19.18.
Equation 19-18 d = dI + d2 + d3
where
d o control delay (s/veh),
d, uniform delay (s/veh),
d, incremental delay (s/veh), and
d, initial queue delay (s/veh).
Chapter 31 describes a technique for measuring control delay in the fieid.

Cofe Motorized Vehicle Methodology Chapler 19/5ignalized Intersections

Page 19-50 VersiQn 6.0
 Highway Capacity Manual: A Guide for Multimodal MobiJjty Analysis

A. Compute Uniform De/ay

The uniform delay for a given lane group serving one traffic movement, and
for which therc are no permitted movements, is computed by using Equation 19-
19 with Equation 19-20and Equation 19-21.
0.5 C(1 - g/C)' Equation 19-19
di = PF --------
1- [min(1,X)g/C]
with
1-P 1-y [ 1-PC/9] Equation 19.20
PF = --- x ------x 1 + y----
1 - g/C 1 - min(1,X) P 1 - g/C

y = min(1,X) g/C
where
PF progression adjustment factor,
y fiow ratio,
P proportion of vehieles arriving during the green indication (decimal),
g effective green time (5), and
e cyele Jength (s).
Equation 19-19 docs not provide an accurate estimate of uniform deJay for a
shared-Jane Jane group or a lane group with permitted operatian becausc these
lane groups have other factors that affect their delay. AJso, this equation does not
provide an accurate estimate of uniform deJay when there is an initiaJ queue
prcsent for one or more intcrsection traffic movements. Section 4, Extensions to
the Motorized Vehide MethodoJogy, describes a procedure for accurately
estimating uniform deJay when any of these conditions is presento

B. Compute Initial Queue Delay

The initial queue delay term accounts for the additional uniform deJay
incurrcd due to an initiaJ queue. This queue is a result of unmet demand in the
prcvious time periodo It does /lO! inelude deJay to any vehides that may be in
qucue due to the random, cycle-by-cycle fiuctuations in demand that
occasionally exceed capacity. Thc initiaJ queue deJay equals 0.0 s/veh when there
is no initial queue present at the start of the analysis period for any intersection
lane group. A proccdure for estimating the associated delay for lane groups with
an initial queue present is provided in Section 4, Extcnsions lo the Motorized
Vehide Methodology.

C. Compute Incremental Delay Factor

The equation for computing incrementaJ deJay indudes a variabJe that
accounts for the eHect of controller type on deJay. This variable is rcferred to as
the incremental deJay factor k. It varies in value from 0.04 to 0.50. A factor value
of 0.50 is recommended for pretimed phases, coord inated phascs, and phascs set
to "recall-to-maximum."
An actuated phase has the ability to adapt its green intervaJ duration to serve
the demand on a cydc-by-cycle basis and, thercby, to minimize the frequency of

Chapter 19/5ignalized lntersections Core Motorized Vehicle Methodology

Ver5lOn 6.0 Puge 19-51
Highway Capacity Manual: A Gujde for Multimodal Mobmty Analysis

cycle faHure. Only when the green is extended to its maximum limit is this
capability curtailed. This influence of actuated operation on dclay is accounted
for in Equation 19-22through Equation 19-25.
Equation 19.22 k = (1 - 2 kmin) (v/ca - 0.5) + kmin :5 0.50
with
kmin = -0.375 + 0.354 PT - 0.0910 PT2 + 0.00889 PT3 ;::= 0.04
Bu s N
Equation 19.24 e =---
" e
Equation 19-25 Bu = Gmax + y + Re -11 - 12
whcre
k incremental delay factor,
CQ available capacity foc a lane group served by an actuated phase
(veh/h),
k",," = minimum incremental deJay factor,
PT = passage time setting (s),
Gmu maximum green setting (s), and
g. availabJe cffective green time (s).
As indieated by this series of equations, the factor vaJue depends on the
maximum green setting and thc passage time setting for the phase that controls
the subject Jane group. Research indicates shorter passagc times result in a Jower
vaJue of k (and Jower delay), provided the passage time is not so shorl that the
phase tenninates before the queue is servcd (11).

D. Compute Incremental Delay

Incremental delay consists of two de1ay components. One component
accounts for delay duc to the effect of random, cycle-by-eycle f1uctuations in
dcmand that oceasionally exceed capacity. This deJay is evidenced by the
occasional overflow queue at the end of the grcen interval (Le.,cycle failure). The
seeond component accounts for deJay due to a sustained oversaturation during
the analysis periodo This delay occurs when aggregate dcmand during the
analysis period exceeds aggregate capacity.1t is sometimes referred to as the
deterministie delay component and is shown as variable d2.d in Exhibit 19-21.
Exhibit 19.21 iIIustrates the queue growth that occurs as vehicles areive at a
demand flow rate v during analysis period T, which has capacity c. The
deterministic delay componcnt is represented by the triangular area bounded by
the thick line and is associated with an average delay per vehicle represented by
the variable du The 1ast vehicle to arrive during the analysis period is shown to
clear the queue tehours after the start of the analysis periodo The average queue
size associated with this deJay is shown in the exhibit as Q2.d. The queue present
at the end oí the analysis period [= T (v - e)J is refereed to as the residual queue.

Core Motorized Vehide Methodology Chapter 19j5ignalized IntersectiOns

Page 19-52 V£'fSiOn 6.0
 Highway Capacity Manual: A Guide for Mu/timodal Mobility Analysis

Exhibit 19-21

T----'I "--'1 Cumulative Arrivals and

Departures During an
Oversaturated Analysis Period
TI' I
C

Time

Thc incremental dclay tenn accounts for delay due to random variation in
the number of arrivals on a eyc1e-by-cyc1ebasis. It also accounts for deJay caused
by demand exceeding capacity during the analysis periodo The amount by which
demand exceeds capacity during the analysis period is referred to as unmet
demando The incremental delay equation was derived by using an assumption of
no initiaJ queue due to unmet demand in the preceding analysis period. Eguation
19-26,with Equation 19-27,is used to compute incrementaJ delay.

d, = 900 T (XA - 1) + (XA - 1)' + 8klX A 1 Equation 19-26

[ CA T

with
XA = V/CA Equation 19-27

where X.~is the average volume-to-capacity ratio, CA is the average capacity

(veh/h), and aH other variables are as previously defined. The variable c..•.is not
the same as the variable Cv the latter of which is computed in Part e of Step 8.
H no lane group at the intersection has an initial gueue, then the average lane
group capacity CA is egual to the capacity c computed in Step 7 (Le., CA = c). H one
or more lane groups have an initiaJ queue, then the procedure described in
Section 4 is used to compute capacity cA'
The incremental delay term is valid for aH values of XA, including highly
oversaturated lane groups.

E Compute Lane Group Control De/ay

The uniform delay, incremental deJay, and initial queue delay values
computed in the previous steps are added (see Eguation 19-18) to estimate the
control delay for the subject lane group.

F. Compute Aggregated De/ay Estimates

lt is often desirable to compute the average control delay for the intersection
approach. This aggregated delay represents a weighted average deJay, where

Chapter 19/5ignalized intersections Core Motorized Vehicle Methodology

VersiOn 6.0 Page 19-53
Highway Capacity Manual: A Guide for Multimodal Mobility Ana/ysis

each lane group delay is weighted by the lane group demand flow rate. The
approach control delay is compuled with Equation 19.28.
mi
d _ L¡=1 di Vi
Equation 19.28 A.j - 't"m¡
""J:1 V¡
where
d"'i approach control delay for approach j (s/veh),
di control delay for lane group i (s/veh),
ni, : number of lane groups on approach j, and
aH other variables are as previously defined. The summation terms in Equation
19-28 represent the sum for alllane groups on the subject approach.
Similariy, intersection control delay is computed with Equation 19-29.
Ld¡vJ
Equation 19.29 d ---
/ - LV¡

where d, is the interscction control delay (s/veh). The summation terms in

Equation 19.29 represent the sum for alllane groups at the subject intersection.
Unsignalized movements al the signalized intersection should also be
considered when an aggregated delay estimate is computed. Inclusion of these
movements should be handled as follows:
• Delay of unsignalized movements should be included in the approach
and intersection aggregate delay cakulations of Equation 19-28 and
Equation 19-29, except for spedal cases that are properiy annotated in the
results.
• When the delay of unsignalized movements is incJuded in the approach
and intersection averages, whether zero or nonzero, the aggregate delay
that results must be annotated with a footnote that indicates this
unsignalized deJay inelusion.
• When the delay of unsignalized movements is not ineluded in the
aggregate totals {Le.,it is not induded in either the numerator (volume x
delay) or the denominator (volume) of Equation 19-28 or Equation 19.29],
this exdusion of unsignalized delay must be deariy represented by a
footnote that indicates this unsignalized delay exdusion.

Step 9: Determine LOS

Exhibit 19-8 is used to determine the LOS for each lane group, each
approach, and the intersection as a whole. LOS is an indicabon of the
acceptability of delay levels to motorists at the intersection. It can also indicate an
unacceptable oversaturated operation for individuallane groups.

Step 10: Determine Queue Storage Ratio

A procedure is described in Section 4 of Chapter 31 for estimating the back.
of-queue size and the queue storage ratio. The back-of-queue position is the
position of the vehicle stopped farthest from the stop Jine during the cycle as a
consequence of the display of a red signal indication. The back-of-queue size

Core Motorized Vehide Methoclology Chapter 19/5ignalized Intersections

Page 19-54 v~6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

depends on the arrival pattero of vehicles and on the number of vehicles that do
not clear the intersection during the previous eycle.
The queue storage ratio represents the proportion of the available queue
storage distance that is occupied at the point in the eycle when the back.of.queue
position is reached. If this ratio exceeds 1.0, then the storage space will overflow,
and queued vehicles may block other vehicles from moving forward.

Interpretation of Results
The computations discussed in the previous steps result in the estimation of
control delay and LOS for each lane group, for each approach, and for the
intersection as a whole. They also produce a volume-to-capadty ratio for each
lane group and a critical intersection volume-to-capacity ratio. This subsection
provides sorne useful interpretations of these performance measures.

Leve/ of 5ervice
In general, LOS is an indication of the gelleral acceptability of delay to
drivers. In this regard, it should be remembered that what might be acceptable in
a large dty is not necessarily acceptable in a smaller dty or rural arca.
Intersection LOS must be interpreted with caution. It can suggest acceptable
operation of the intersection when in reality certain lane groups (particularly
those with lower volumes) are operating at an unacceptable LOS but are masked
at the intersection level by the acceptable performance of higher-volume lane
groups. The analyst should always verHy that each lane group is providing
acceptable operation and consider reporting the LOS for the poorest.performing
¡ane group as a means of providing context to the interpretation of intersection
LOS.

Vo/ume-to-capacity Ratio
In general, a volume-to-capacity ratio greater than 1.0 is an indication of
actual or potential breakdown. In such cases, a muItiple'period analysis is
advised for this condition. This analysis would encompass all consecutive
periods in which a residual queue is presento
The critical intersection volume-to-capadty ratio is useful in cvaluating the
interscction from a capadty-only perspective. It is possible to havc a critical
intersection volume-to-capadty ratio of less than 1.0 and still have individual
movements oversaturated within the signal eycle.1f this situation occurs, then
the cycle time is generally not appropriately allocated amang the phases.
Reallocation of the cyclc time should be considercd, so that additional time is
given to the phases serving those lane groups with a volume-to-capacity ratio
greater than 1.0.
A critical intersection valume-to-capadty ratio greater than 1.0 indicates the
overall signal timing and geometric design provide inadequate capadty for the
given demand flows. Improvcments that might be considered indude the
following:
• Basic changes in intersection geometry (i.e., changc in thc numbcr or use
of lanes),

Chapter 19/5ignalized Intersections (ore Motorized Vehicle Methodology

Version 6.0 Page 19-5S
Highway Capacify Manual: A Guide tor Multimodal Mobility Analysis

• Increase in signal eyele length if it is determined to be too short, and

• Changes in signal phase sequence or timing.
Local guidelines should always be consulted before potential improvcments
are developcd.
Fully actuated control is intended to allocate cyele time dynamically to
movements on the basis of demand and, thercby, maintain efficient operation on
a eyele-by-cycle basis. Thc critical intersection volume-to-capacity ratio can
provide an indication of this effidency. In general. this ratio will vary betwcen
0.85 and 0.95 for most actuated intersections, with lower values in this range
more common for intersections having multiple detectors in the through traffic
lanes. A ratio less than 0.85 may indicate excessive green extension by random
arrivals, and the analyst may consider reducing passage time, minimum green,
or both. A ratio more than 0.95 may indicate frequent phase termination by max-
out and limited ability of the controller to reallocate cyele time dynamically on
the basis of detccted demando Increasing the maximum green may improve
operation in some instances; howcver, it may also degrade operation when phase
flow rates vary widely (because green extension is based on total flow rate
scrved by the phase, not flow rate per lane).
For semiactuated and coordinated.actuated control, the critical intersection
volume-to-capacity ratio can vary widely because of the nonactuated nature of
sorne phases. The duration of these phases may not be directly related to their
associated demand; instead, it may be dictated by coordination timing or the
demand for the other phases. A critical intersection volume-to-capacity ratio that
exceeds 0.95 has the same interpretation as offered previously for fully actuated
control.
The critical intersection volume-to-capacity ratio can be misleading when it
is used to cvaluate thc overall sufficiency of the intersection geometry, as is often
required in planning applications. The problem is that low flow rates dictate the
need for short cyeIe lengths to minimize delay. However, the equation for
calculating this ratio indicatcs thc desircd shorter cycle Icngth produces a higher
volume-to-capacity ratio. Therefore, a relatively large value for this ratio
(provided it is less than 1.0) is not a certain indication of poor operation. Rather,
it means e10ser attcntion must be paid to the adequacy of phase duration and
queue size, especially for the critical phases.

Vo/ume-to-capacity Ratio and De/ay Combinations

In sorne cases, delay is high even when the volume-to-capacity ratio is low.
In these situations, poor progression, a notably long cycle Icngth, or an inefficient
phase plan is generally the cause. When the intersection is part of a coordinated
system, the cyele length is determined by system considerations, and aIterations
at individual intersections may not be practica!.
It is possible for delay to be at acceptable levels even when the volume-to-
capacity ratio is high. This situation can occur when sorne combination of the
following conditions exists: the cyele length is relatively short, the analysis
period is short, the lane group capacity is high, and there is no initial queue. If a

Core Motorized Vehicle Methodology OJapter 19/5ignalized Intersections

Page 19.56 Version 6.0
 Highway Capacity Manual; A Guide lor Multimodal Mobility Analysis

residual queue is created in this scenario, then a multiple-period analysis is

necessary to gain a true picture of the delay.
When both delay levels and volume-to-capacity ratios are unacceptably high,
the situation is critica!. In such situations, the delay may increase rapidly with
small changes in demando The full range of potential geometric and signal design
changes should be considered in the search for improvemcnts.
In summary, unacceptable delay can exist when capacity is a problem as well
as when capacity is adequate. Further, acceptable delay levels do not
automatically ensure capacity is sufficient. Oclay and capacity are complex
variables that are infIuenced by a wide range of traffic, roadway, and
signalization conditions. The methodology presented here can be used to
estimate these performance measures, identify possible problems, and assist in
developing altemative improvements.

O1apter 19/5ignalized Intersections Core Motorized Vehicle Methodology

V&SiOn 6.0 Page 19-57
Highway Capacity Manual: A Guide for Multimodal MobUity Analysis

4. EXTENSIONS TO THE MOTORIZED VEHICLE

METHODOLOGY

CRITlCAL INTERSECTION VOLUME. TO.CAPACITY RATIO

Overview
A useful concept for analyzing signalized intersections is the critical
intersection volume-to-capacity ratio Xc'This ratio is computed by using
Equation 19-30 with Equation 19-31.

Equation 19-30 x, = (C~ J2:>,.,

iECi
with

Equation 19-31

where
X, critical intersection volume-to-capacity ratio,
e cyclc length (s),
Y,,; critical flow ratio for phase j = v/(N s,),

[u phase ilost time = lu + 121(s),

ci set of critical phascs on the critical path, and
L cycle lost time (s).
Thc summation term in each of thcse equations represents thc sum uf a
specific variable for the set of critical phases. A critical phase is one phase of a set
of phases that occurs in sequencc and whose combined flow ratio is the largest
for the signal cycle. The critical path and critical phases are identified by
mapping traffic movemcnts to a dual-ring phase diagram, as shown in Exhibit
19-2.

Equation 19-30 is based on the combined assumption that each critical phase
has the same volume-to-capacity ratio and that this ratio is equal to the critica!
intersection volume-to-capacity ratio. This assumption is valid when the effective
green duration for each critical phase iis proportional to y,jr.Qh¡)' When this
assumption ho!ds, the volume-to-capacity ratio for each noncritical phase is less
than or equal to the critical intersection volume-to-capacity ratio.

Identifying Critica! Lane Groups and Critica! Flow Ratios

Calculation of the critical intersection volumc-to-capacity ratio requircs
identification of the critical phases. This identification begins by mapping all
traffic movements to a dual-ring diagram.
Next, the lane group flow ratio is computed for each lane group served by
the phase. If a lane group is served only during one phase, then its flow ratio is
computcd as the lane group flow rate (per lane) divided by the lane group
saturation f10wrate [Le., v/(N sJ).1f a lane group is served during multiplc

Extensions to the Motorized Vehicle Methodology Chapter 19/5ignalized Intersections

Page 19-58 Vet:SiOn6.0
 Highway Capacity Manual: A Guide for Mu/timodal MobiJity Analysis

phascs (e.g., protected-permitted), then a flow ratio is computed for each phasc.
Spccifically, the demand flow rate and saturation flow rate that occur during a
given phase are used to compute the lane group flow ratio for that phase.
If the lane group is served in a permitted manner, then the saturation flow
rate Si used to determine the flow ratio is an average for the permitted green
periodo For left tums, it is computed with Equation 31.59 in Chapter 31 (with Eu
and Eu substituted for Eu", and ELI."" respectively) and the instructions that
follow this equation as they relate to shared or exclusive lane assignment. This
equation applies to lane groups served as permitted-only and to lane groups
servcd during the permitted phase of protected-permitted operation. For right
turns, Equation 31-61 is used (with ER substituted for fR.",)'
If the lane group is served by protected-permitted operation, then its volume
v, must be apportioned to the protected and permitted phases. To accomplish
this apportionment, it is appropriate to consider the phase that is displayed first
to be fuUy saturatcd by turning traffic and to apply any residual f]ow to the
phase that is displayed second. In this manner, the voluroe assigned to the first
phase is the smaller of the phasc capacity or the demand volume, and any
unassigned volume goes to the second phase.
Next, the phase flow ratio is determined from the flow ratio of eam lane
group served during the phase. The phase flow ratio represents the largest f10w
ratio of alllane groups servcd.
Next, the diagram is evaluated to identify the critical phases. The phases that
occur between one barrier pair are collectively evaluated to determine the critical
phases. This evaluation bcgins with the pair in Ring 1 and proceeds to the pair in
Ring 2. Each ring represents one possible critical path. The phase f10wratios are
added for cach phase pair in cach ringoThe larger of the two ring totals
represents the critical path, and the corresponding phascs represcnt the critical
phases for the barrier pair.
Finally, the process is repeated for the phases between the other barrier paie.
One critical flow rate is defined for each barrier pair by this process. These two
values are then added to obtain the sum of the critical flow ratios used in
Equation 19-30. The lost time associated with each of the critical phases is added
to yield the cycle lost time L.
The procedure for the basic intersection case is explained in the n('xt few
paragraphs by using an example intersection. A variation of this procedure that
applies when protected-permitted operation is used is described after the basic
case is described.

8asic case
For the basic case, consider an intersection with alead-lag phase sequence
on the major street and a permiUed-only sequence on the minar street, as shown
in Exhibit 19-22.The northbound right turn is provided an exclusive lane and a
grecn arrow indication that displays concurrently with the complementary left-
turn phase on the major street. Each of the left-turn movements on the major
strcet is served with a protected phase.

Chapter 19/5ignalized Intersections Extensions to the Motolized Vehide Methodology

Version 6.0 Page 19-59
Highwoy Copacity Manual: A Guide for Mulfimodol Mobility Analysis

Exhibit 19-22 Lane Groups

Oitical Path Determinaban Minar street
with Protected Left-Tum
Phases
4+7

.1
Majar Street
~.
¡-'

•., (0.30)
11(1
ECluivalent Dual-Rino Structure (now ratiosl
~1 (0.15) 4>~.25)
---.;¡ r
•., (O.2~ ••• ~.25)
(10.10)

~8
(O.3W
,,•
:(0.15)
B.mier
'""'"
Note: • Critical flow ratio.

Phases 4 and 8 represent the only phases bctwecn the barrier pair serving the
minor-street movements. Inspection of the fIow ratios providcd in the exhibit
indicates Phase 8 has two lane-graup fIow rates. The Iarger flow rate corresponds
to the shared left-tum and through movement. Thus, the phase flow ratio for
Phase 8 is 0.30. The phase flow ratio for Phase 4 is 0.25. Of the two phases, the
larger phase flow ratio is associatcd with Phase 8 (= 0.30), so it represents the
critical phase foc this barrier pairo
Phases 1, 2, 5, and 6 represent the phases between the other barrier paie.
They serve the major-street approaches. The phase flow ratio of Phase 1 is 0.15,
on the basis of the leH-tum lane group flow rate.
There are two possible critical paths through the major-street phase
sequence. One path is associated with Phases 1 and 2 (Le., Ring 1), and the other
path is associated with Phases 5 and 6 (Le., Ring 2). The total phase fIow ratio for
the Ring 1 path is 0.30 + 0.15 = 0045. The total phase flow ratio for the Ring 2 path
is 0.25 + 0.25 = 0.50. The Iatter total is larger and, hence, represents the criticaI
path. It identifies Phases 5 and 6 as the critical phases. Thus, the sum of critical
fIow ratios for the eycle is 0.80 (= 0.30 + 0.50).
One increment of phase lost time 11is associated with each phase on the
critical path. Thus, the eyde lost time L is computed as the sum of the lost time
for each of Phases 5, 6, and 8.

ExtenSlons to the Motorized Vehide MethodoIogy Chapter 19{5ignalized Intersectians

Page 19-60 VersiOn 6.0

L
 Highway Capacity Manual: A Guide tor Multimodal Mobility Analysis

Special case: Protecteci-Permitteci Left-Turn Operation

For the special case, consider an intersection with a lead-Iead phase
sequence on the majar street and a permitted-only sequence on the minar street,
as shown in Exhibit 19-23.The leit,tum movements on the major street operate in
the protected-permitted mode. Phases 4 and 8 represent the only phases between
one barricr paie. They serve the minor-street lane groups. By inspection of the
flow ratios provided in the exhibit, Phase 8 has the highest flow ratio ('"0.30) of
thc two phases and represents the critical phase for this barrier paie.

Lane Groups
Exhibit 19-23
Critical Path Determination
Minor Street with Protected-Permitted left-
4+7 Tum Operation

.1
Major Street

<1>, <1>2
111
Equivalent Dual-Ring Structtlre
(0.30) l
(f1ow ratios)
(0.15)
-_ ...•• CP~.25)

1(0.05) ~
<1>, (0.20) <1>. ~.25) <1>.
--_ ........ ,................. _-- (0.30)~
--! .;'(óji)
Barner
Note: • Oiticalllow ratio.

Phases 1, 2, 5, and 6 represent the phases between the other barrier pairo
They serve the major'street approaches. Each lefHum lane group is shown to be
served during two phases-once during the ¡efHum phase and once during the
phase serving the adjacent through movement. The fiow ratio for each of the four
left,turn service periods is shown in Exhibit 19-23.The following rules define the
possible critical paths through this phase sequence:
1. One path is associated with rhascs 1 and 2 in Ring 1 (0.35'" 0.05 + 0.30).
2. One path is associated with Phases 5 and 6 in Ring 2 (0.45'"0.20 + 0.25).
3. If a lead-lead or lag-lag phase sequcnce is used, then one path is
associated with (a) the lcft-tum phase with the larger flow ratio and (b)
the through phase that permissively serves the same left-tum lane
group. Sum thc protectcd and pcrmitted left-tum fIow ratios on this
path (0.35 '"0.20 + 0.15).
4. If aJead-lag phase sequence is used, then one path is associated with (a)
the leading left-tum phasc, (b) the lagging lefHurn phase, and (e) the

Olapter 19j5ignalized Intersections Extensions lo the Motorized Vehide Metnodology

Vers.0n 6.0 Page 19-61
Hjghway Capacity Manual: A Guide for Multimodal Mobility Analysis

lhrough phase (see discussion to follow). Sum the two

cOlltrol1illg
protected left-tum flow ratios and the one controlling permitted l£Oft-
tum flow ratio on this path.
lf alead-lag phase sequence is used, each of the through phases that
permissively serve a leH-tum lane group is considered in determining the
controlling through phase.lf both through phases have a permitted period, then
there are two through phases to considero The controlling through phase is that
phase with the larger permitted Jeft.tum flow ratio. For example, if Phase 1 were
shown to lag Phase 2 in ExhibilI9-23, then Phase 6 would be the controlling
through phase because the permitted left-tum flow ratio of 0.22 exceeds 0.15. The
critical path for this phase sequence would be 0.47 (= 0.20 + 0.22 + 0.05).
The first thrcc rules in the preceding list apply to the example intersection.
The calculations are shown for each path in parentheses in the previous Iist of
rules. lOe total f10w ratio for the path in Ring 2 is largest (= 0.45) and, hence,
represents th£ocritical path. It identifies Phases 5 and 6 as the critical phases.
lOus, the sum of critical flow ratios for the eycle is 0.75 (= 0.30 + 0.45).
lf Rule 3 in the preceding Iist applies. then the only lost time incurred is the
start-up lost time 11associated with the first critical phase and the c1earance lost
time 12associated with the second critical phase. lf Rule 1, 2, or 4 applies, then
one increment of phase lost time Iris associated with each critical phase. Rule 2
applies for the example, so the eycle lost time L is computed as the sum of the
lost time for each of Phases 5, 6, and 8.
Two flow ratios are associated with Phase 6 in this example. 60th flow ratios
are shown possibly to dictate the duration of Phase 6 (this condition does not
hold for Phase 2 becausc of the timing of the left-tum phases). This condition is
similar lo that for the northbound right-tum movement in Phase 1 of Exhibit 19-
22, and the treatment is the same. lOat is, both flow ratios are considered in
defining the phase flow ratio for Phase 6.

UNIFORM DELAY CAlCULATION USING QUEUE ACCUMULATION

POlYGON
Overview
This subsection describes a procedure for calculating uniform delay. lOis
incremental queue accumulation procedure (21, 22) is sufficiently general that it
can be applied to any Jane group, regardless of whether the lane group is shared
or exclusive or served with a protectcd, pcrmitted, or protected-permitted
operation.
lOe incremental queue accumulation procedure models arrivals and
departurcs as thcy occur during the average cycle. Specifically, it considers
arrival rates and departure rates as they may occur during one or more effective
green periods. lOe rates and resulting queue size can be shown in a queue
accumulation polygon, such as that shown previously in Exhibit 19-7. lOe
procedure decomposes the resulting polygon into an equivalent set of trapezoids
or triangles fOTthe purpose of delay estimation.

Extensions to the MotOri2ed Vehicle Methodology O1apter 19/5ignalized Intersections

Pase IS-62 Version 6.0
 Highway Capacity Manual: A Guide tor Multimodal Mobility Analysis

Polygon Construction
The key criterion for constructing a trapezoid or triangle is that the arrival
and departure rates must be effectively constant during the assodated time
periodo This process is illustrated in Exhibit 19-24for a lane group having two
different departure rates during the effective green periodo

Exhibit 19-24
Decomposition of Queue
Area of Trapezoid Represents Accumulation Polygon
One Delay Element

Q,

-- t",i

Time (s)

Construction of the queue accumulation pol)'gon requires that the arriva!

flow rate not exceed the phase capadty. If thc arrival flow rate exceeds capadty,
then it is set to egual the capadt)' for the purpose of constructing the polygon.
The queue can be assumed to equal zero at the end of the protected phase, and
the polygon construction process begins at this point in the cyclc. Once the
polygon is constructed, this assumption must be checked. If the ending queue is
not zero, then a second polygon is constructed with this ending queue as the
starting queue for the first interva1.
Construction of the queue accumulation polygon requires converting all flow
rate variables to common units of vehicles per second per Jane. This conversion is
implicit for a11flow rate variables shown in the exhibits that depict a queue
acrumulation polygon.
Polygon construction requires identifying points in the cycle al which one of
the following two conditions applies:
• The departure rate changes (e.g., due to the start or end of effective grcen,
a change in the saturation flow rate, depletion of the subject queue,
depletion of the opposing gueue, departure of sneakers), or
• The arrival rate changes (e.g., when a platoon arrival condition changes).
During the intervals of time behveen these points, the saturation f10w rate
and arrival flow rate are constant.
The determination of flow-rate change points may require an iterative
calculation process when the approach has shared lanes. For example, an
analysis of the opposing through movement must be completed to determine the
time this movement's queue clears and the subject left-tum lane group can begin
its service periodo This service period may, in tum, dictate when the permitted
left-tum movements on the opposing approach may depart.

Chapter 19¡5ignalized intersections Extensions to the Motorized Vehicle Methodology

Version 6.0 Page 19-63
Highway Capacity Manual: A Guide for Multimodal Mobilify Analysis

The procedure is based on defining arrival rate as having one of two fIow
states: an arrival rate during the green indication and an arrival rate during the
red indication. Further information about when each of these rates applies is
described in the discussion for platoon ratio in the Reguired Data and Sourccs
subsection. The proportion of vehides arriving during the green indication P is
used to compute the arrival fIow rate during each fIow state. Eguation 19.32 and
Eguation 19.33 can be used to compute thesc rates.
qP
Equation 19-32 qg = g/e
,nd

Equation 19.33

where
qg arrival f10wrate during the effective green time (veh/s),
q, arrival f10wrate during the effective red time (veh/s),
q ". arrival fIow rate = v/3,600 (veh/s),
P proportion of vehides arriving during the green indication (decimal),
,nd
g = effective green time (s).
A more detailed description of the procedure for constructing a queue
accumulation polygon for lane groups with various lane allocations and
operating modes is provided in Section 3 of Chapter 31.

Delay Calculation
The uniform delay is determined by summing the arca of thc trapezoids or
triangles that compase the polygon. The area of a given trapezoid or triangle is
determined by first knowing the queue at the start of the interval and then
adding the number of arrivals and subtracting the number of departures during
the specified time interva1. The result of this calculation yields the number of
vehicles in queue at the end of the interval. Eguation 19-34illustrates this
calculation foc interval i.

Equation 19.34 Qí = QI-1 - (3,:00 - ~) td.1 ~ 0.0

where
Qi queue size at thc end of interval i(veh),
N number of lanes in the lane group (In),
s = adjusted saturation f10wrate (veh/h/ln), and
td,i duration of time intcrval i during which the arrival fIow rate and
saturation flow rate are constant (s).

Extensions lo the Motorized Vehlcle Methodology Otapter 19/5ignalized Intersections

Page 19-64 Version 6.0
 Highway Capacity Manual: A Guide for Mulfimodal Mobilify Analysis

The uniform delay is calculated by using Eguaban 19-35with Eguation 19-36.

0.5 EeQí-I + Q¡) tu

dI = -- Equation 19-35
qC
with
Equation 19-36

when."
di uniform delay (s/veh),
tu duration of trapezoid or triangle in interval ¡(s),
wq queue change rate (Le., slope of the upper boundary of the trapezoid
or triangle) (veh/s), and
aH other variables are as previously defined.
The summation term in Equation 19-35 indudes aH intervals for which therc
is a nonzero queue. In general, tu wiII equal the duration of the corresponding
intervai. However, during sorne intervals the gueue wiII dissipate, and t,,; wiII
onl}' be as Jong as the time required for thc queue to dissipate (= Q¡_¡/wq).

INITIAL QUEUE DELAY CALCULATION

Overview
Initial queue delay accounts for the additional delay incurred due to an
initial gueue. This qucue is a result of unmet demand in the previous time
periodo It does IlOt inelude any vehides that may be in qucue duc to random,
cyelc-by-cyele f1uctuations in demand that occasionaIly cxceed capacity.
Exhibit 19-25 illustratcs the deJay due to an initial queue as a trapezoid shape
bounded by thick lines. Thc average del ay per vehide is represented by the
variable d). The initial queue size is shown as Qb vehicles. The duration of time
during thc analysis period for which the eHect of the initial qucue is still present
is reprcscnted by the variable t. This durabon is shown to equal the anal}'sis
pcriod T in Exhibit 19.25. However, it can be shortcr than thc analysis period
duration for sorne 10wer-voJume conditions.
Exhibit 19-25ilIustrates the case in which the demand f10wrate v exceeds the
capadty e during the analysis periodo In contrast, Exhibit 19-26 and Exhibit 19-27
illustrate altemative cases in which the demand f10wrate is less than the
capadty.
The remainder of this subsection describes the procedure for computing the
initiaJ queue deJa)' for a Jane group during a given analysis periodo

Chapter 19/5ignalized Intersections Extensions lo the Motorized Vehicle Methodology

Version 6.0 Page 19-65
Highway Capacity Manual: A Guide lor Multimodal MobiJjty Analysis

Exhibit 19-25
lnitial Queue Delay with
IllCreasing Queue Size
T" ,¡
,
••
J!
"•
>
-T/', '1
e
"
...-
, .-

•
~
,
!!
e,
u

...
,
. ..' Q,
o . •
'

o Time

Exhibit 19-26
lnmal Queue Delay with ,
Decreasing Queue Size
T '1
, '1
~
J! T/', I
"~ _LJ'-~
• ..'.'
ti
,
!!
,
e
u 1
..
'
,
...- Q,
o
o
• Time

Exhibit 19-27
Initial Queue Delay with
Queue Oearing T '1
~
"~
•
'/'4 td
-1 Lf~
.'

• .'.'
,
~
• .'
/-- .-
~
e,
u •
f
. .'
..- .' Q,
o .'
o Time

ExtensiollS to the Motorized Vehicle Methodology Olapter 19j5ignalized Intersections

Page 19-66 Version 6.0

J
 Highway Capacity Manual: A Guide for Mulfjmodal Mobility Analysis

Computational Steps
A. lnitial Queue Analysis
At the start of this step, the initial queue that was input for each movement
group nl.'eds to be converted to an initial queue for each lane group. When there
is a one-to-one correlation between the movement group and the lane group,
then the initial queue for the lane group equals the input initial gueue for the
movement group. When there is a shared lane on a multiple-Iane approach, then
the input initial queue needs to be distributed among the lane groups that serve
the movements sharing the lane. Specifically, the initial queue for each lane
group is estimated as being equal to the input initial queue rnultiplicd by the
number of lanes in the lane group and divided by the total number of shared and
through lanes.
The saturation fIow rate, phase duration, capacity, and uniforrn delay will
need to be recomputed for each lane group during this step. When these
variables are computed for a lane group with an initial queue, the arrival flow
rate for the lane group is inflated such that it equals the lane group capacity (Le.,
the actual input demand flow rate is not used). The remaining lane groups will
have their arrival flo\\' rate set to egual the smaller of the input demand flow rate
or the capacity.
The need to recompute these variables stems from the influence one lane
group often has on the operation of other lane groups. This influence is notably
adverse when one or more lane groups are operating in a saturated state for a
portion of the analysis periodo If the saturated lane group represents a conflicting
rnovement to a lane group that ineludes a permitted left-tum operation, then the
leh-turo lane group's operation will also be adversely affected for the same time
periodo Moreover, if the phase serving the ¡ane group is actuated, then its
capacity during the saturated state will be different from that of the subsequent
unsaturated state.
The uniform delay computed during this step is referred to as the saturated
uniform dc1ay. It is computed for each lane group by using the arrival flow rate,
capacity, and phase duration determined with the previous guidance.
The duration of unmet demand is cakulated in this step for each lane group.
Either Equation 19-37or Equation 19-38is used for this purposc.
lf v ~ c..,then
t=T Equation 19-37

Jfv<c" then
Equation 19-38

where
duration of unmet demand in the analysis period (h),
T analysis period duration (h),
Q. initial queue at the start of the analysis periad (veh),
v = demand flow rate (veh/h), and
c. saturated capacity (veh/h).

Chapter 19j5ignalized lntersectiOlls Extensions to the Motorized Vehide Methodology

Version 6.0 Page 19-67
 Highway Capadfy Manual: A Guide for Multimodal Mobility Analysis

For this calculation, the saturated capacity e, is equaI to that obtained from
the polygon constructed in this step and is reflective of the phase duration that is
associated with saturated operation (due to the initial queue).
Next, the average duration of unmet demand is calculated with Equation
19-39.
Equation 19.39

where
t. average duration of unmet demand in the analysis period (h),
ti duration of unmet demand for Iane group iin the anaJysis reriod (h),
and

Ng '" number of lane groups for which t exceeds 0.0 h.

The summation term in Equation 19.39 represents the sum of the t values for
only those lane groups that have a vaJue of t that exceeds 0.0 h. The average
duration t. is considered as a single representative value of t for alllane groups
that do not have an ¡nitial queue.

B. Compute Uniform De/ay

The uniform delay computed in 5tep 8 of the core motorized vehicle
methodology is adjusted in this step such that the adjusted uniform deJay reflects
the presence of the initial queue. Initially, the uniform deIay dI computed
previously is renamed as the baseline uniform deJay dl~ (Le., dlb = dI)' Next,
Equation 19-40 or Equation 19-41 is used to compute the uniform delay for each
lane group.
If lane group ¡has an initial queue, then
ti (T - ti)
Equation 19-40
du = ds•1 T + d1b•i T

If lane group idoes not have an initial queue, then

ta (T - ta)
Equation 19-41 dt•i = ds.i T + d1b.i T

where d, is the saturated uniform delay (s/veh), dl~ is the baseline uniform deJay
(s/veh), ti is the duration oi unmet demand for lane group i in the analysis period
(h), and other variables are as previously defined.

e Compute Average CiJpadty

Equation 19-42 and Equation 19-43 are used to compute the average capacity
for each lane group.
1£Iane group ¡has an initiaI queue, then
t, (T - t,)
Equation 19-42 CA'=C
,1 s. i-+C'---
T 1 T

Extensions lo the MotOlized Vehicle Methodology Chapter 19/5ignalized Intersections

Page 19-68 VefSK)n 6.0
 Highway Capadty Manual: A Guide for Multimodal MobWty Analysis

If lane group idoes not havc an initial queue, then

a t (T - ta)
CA.=C
,1
.-+c.----
S,IT ¡ T Equation 19-43

whcrc CA is thc avcrage capacity (veh/h).

D. Compute Initia/ Queue De/ay

Equation 19-44 through Equation 19-49 are used to compute the initial queue
delay for each lane group.
_ 3,600 ( Qb + Qe - Qeo Qi - Qio
d3 - -- tA ------ + ---- --- QE) Equation 19-44
vT 2 2 CA 2 CA

with
Equation 19-45

If v ~ CIV then
Qeo=T(V-CA) Equation 19.46

tA = T Equation 19-47

Jf v < cIV then

Qeo = 0.0 veh Equation 19-48

tA = Qb/(CA - v) s: T Equation 19-49

where
fA adjusted duration of unmet demand in the analysis period (h),
Q, queue at the end of the analysis period (veh),
Q", qucue at the end of the analysis period when v ~ CA and Qó = 0.0 (veh),
,nd
all other variables are as previously defined. The queue at the end of the analysis
period Q, is also referred to as the residual queue.
The last vehicle that arrives to an overflow queue during the analysis pefiod
will c1ear the intersection at the time obtained with Eguation 19-50.
te = tA + Qe/CA Equation 19-50

where t< is the gueuc-clearing time (h).

The gueue-clearing time is measured from thc start oí the analysis pefiod to
the time the last arriving vehicle c1ears the interscction.

Chapter 19¡5ignalized Intersections Extensions to the Motorized Vehicle Methodology

Version 6.0 Page 19-69
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

5. PEDESTRIAN METHODOLOGY

This section describes the methodology for evaluating the quality of service
providcd to pedestrians traveling through a signalized intersection.

SCOPE OF THE METHODOLOGY

The overall scope of the three methodologies was provided in Section 2. This
section identifies the additional conditions for which the pedestrian
methodology is applicable.
• Target travel modes. The pedestrian methodology addresses travel by
pedestrians walking across one or more legs of a signalized intersection. It
is not designed to evaluate the performance of other travel mcans (e.g.,
Segway, roUer skates).
• "Typical pedestrian" Jocus Jor pedestrian mefhodology. The pedestrian
methodology is not designed to reflect the perceptions of any particular
pedestrian subgroup, such as pcdestrians with disabilities. The
performance measures obtained from the methodology are not intended
to be indicators of a sidewalk's compliance with U.S. Access Board
guidelines related to Americans with Disabilities Act requirements. For
this reason, they should not be considered as a substitute for an Americans
with Disabilities Act compliance assessment of a pedestrian facility.

Spatial Limits
Intersection performance is separately evaluated for each crosswaIk and
intersection comer with this methodology. Unless otherwise stated, all variables
idellti/ied in this subsection are specific fo oue cr055walk aud one cornero A crosswalk is
assumed to exist across each intersection leg unless crossing is specifically
prohibited by local oedinance (and signed to this effect).

Peñormance Measures
Performance measures applicable to the pedestrian travel mode inelude
comer circulation area, crosswalk circulation area, pedestrian delay, and
pedestrian LOS score. Pedestrian deIay represents the average time a pedestrian
waits for a legal opportunity to cross an intersection legoThe LOS score is an
indication of the typical pedestrian's perception of the overall crossing
experience.
LOS is also considered a performance measure. 1t is useful for describing
intersection performance to eIected officials, policy makees, administrators, or
the publico LOS is based on the pedesteian LOS score.
The two circulation-area performance measures are based on the concept of
pedestrian space. One measure is used to evaluate the circulation area provided
to pedestrians while they wait at the comer. The other measure is used to
evaluate the area provided while the pedestrian is crossing in the crosswalk.

Pedestrian Methodology Olapter 19/5ignalized Intersections

Page 19-70 Versk¥7 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Circulatian area describes the space available ta the average pedestrian. A

larger area is more desirable from the pedestrian perspective. Exhibit 19-28can
be used ta evaluate intersection circulation area performance from the pedestrian
perspective.

Pedestrian Space (ftl/p) Description Exhibit 19-28

>60 Ability te move in desired path, no need te alter movements Qualitative Description of
>4<>-60 Dccasi041al need te adjust path te avoid conflicts Pedestrian Space
>24-40 Frequent need to adjust path to avoid oonflicts
>15-24 Speed and ability to pass slower pedestrians restricted
>8-15 5peed restricted, vef'( Iímited abílity to pass slower pedestrians

'" Speed severely restricted, frequent contact with other users

Limitations of the Methodology

This subsectian identifies the knawn limitations of the pedestrian
methodalogy. If one or more of these limitations is believed ta have an important
influence on the performance of a specific interse<:tian,then the analyst should
consider using alternative methods or tools for the evaluation.
The pedestrian methadolagy does not account for the effect of the following
conditions on the quality af service provided ta pedestrians:
• Grades in excess of 2%,
• Prescnce of railroad crossings,
• Unpaved sidewalk, and
• Free (Le., uncontrolled) channelized right turn with multiple ¡anes or
high-speed aperation.

Pedestrian Flow Conditions

Exhibit 19-29and Exhibit 19-30show the variables considered when one
comer and its two crosswalks are evaluated. Two flow conditions are illustrated.
Condition 1 corresponds to the minor-street crossing that occurs during the
major-street through phase. The pedestrians who desire to cross the major street
must wait at the comer. Condition 2 corresponds to the major-strect crossing that
occurs during the minor-street through phase. For this condinon, the pedestrians
wha desire to cross the minar street wait at the comer.

Chapter 19j5ignalízed intersectíons Pedestrian Methodology

Vet5kln 6.0 Page 19-71
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Exhibit 19-29
Condition 1: Minor'Street
Crossing

Major Street

Sidewalk @ HOld Area

w. (minor red)

,
___ J: Crosswalk @_
,
: Key Conditlon 1
,, V.,b"' sidewalk f10w

Minar Street
,, vd:>"' peds jolnlng Queue
, Veo "' outbound crossinll peds
: Veo O' inbound crosslng platoon
Crosswalk @ I w",b- width of sidewalks

Exhibit 19-30
~
Condition 2: Major.Street
Crossing
<, Sídewalk @
~_. .w.
~ Major Street

~
,"""""""""~:~---------
Sidewalk @ I
w.
i,
•
,,
Clossing Platooo

, ,
• @_
• ___ J: Crosswalk
•
•
• HOld Area Key CondjtjQo 2
• (major red)
• v••,/>_ sidewalk flow
•,
Minor Str~t , Veo "' peds jolnlnllllueue

,,, vd:>•• outbound

va O' Inbound
crossinll
crossinll
peds
platoon
, Crosswalk@ W.,b- wldth of sldewalks

REQUIRED DATA AND SDURCES

This subsection describes the input data needed for the pedestrian
methodology. These data are listed in Exhibit 19-31.The second column (labeled
Basis) of the exhibit indicates whether the input data are needed for each traffic
movement, each signal phase, each intersection approach, or the intersection as a
whole. The exhibit also Iists default values that can be used if local data are not
available (25, 30).

Pedestrian Methodology Qlapter 19/5ignatized Intersections

Page 19-72 V~6,O
 Highway Capacify Manual: A Guide for Multjmodal Mobilify Analysis

Potential Data Exhibit 19.31

Required Data and Units Basis Source(s) Suggested Default Value Required Input Data, Potential
Trame Characteristics Data Sources, and Default
Motorized vehde demand fIow rate
Values for Pedestrian Analysis
M Field data, past counlS Must be provided
(veh/h)
Right-tum-oo-red fIow rate (veh¡h) A F.eld data, past oounts Must be provided
Permitted 'eft-tum fIow rate
M Field data, past oounts See discussion
(veh/h)
M'dsegment 85th percentile speed
A Field data Speed limrt:
(mi/ti)
Pedestrian fIow rate (veh/h) M Reld data, past oounts Must be provióed
Geometric Design
Number 01 lanes (In) 1 Field data, aerial photo Must be provided
Numbe¡ of right-turn islands
(O, 1, 2)
1 Field data, aerial photo O
BusJness or offlCe land use: 9.0 ft
Total walkway w;dlh (It) A Field data, aerial photo
Reside!1tial or industrialiand use: 11.0 ft
Crosswalk width (ft)
Crosswalk Iength (It)
1
1
F.eld data, aerial photo
Fleld data, aerial photo
1,"
Must be provided
Trucks and buses in turn volume: 4S ft
Comer radius (ft) A Field data, aerial photo
No trueks Of buses in turn volume: 2S ft
Signal Control
Aetuated: 7 s
Walk (s) P Field data Pretimed: green interval minus
pedestrian clear
Pedestrian dear (s) P Field data 8ased on 3.5-ftjs walking speed
Rest in walk (yes or no) P Field data Not enabled
Cyde length (s) I Fleld data Same as motoriled vehide mode
Yellow change .;. red clearance (s)" P
Duration of phase serving
pedestrians (s)
p
Field data

Field data
"
Same as motorized vehlcle mode

Pedestrian signal head presence

P Field data Must be provided
(yes or no)

Other Data
Analys;s peroo duration (h)" Set by analyst 0.25 h
Notes: M= movement: one value for eilCh Ieft-tum, through, and right-tum movement.
A = approach: one value for the intersection approach.
L = leg: one vaiue for the intersecbon leg (approach plus departure sides).
P = phase: one value or conditioo for eilCh signal phase.
1 = intersection: one value for the intersection.
"5pecifoc values 01' yellow change and red clearance should be determined by kx:al guidelines or practíce,
"Anatysis peliod durabon is as defined for Exhibit 19-11.

The data elements Iisted in Exhibit 19-31do not indude variables that are
considered to represent calibration factors. Default valucs are provided for thcse
factors because they typically havc a Telatively narrow range of reasonable
values ur have a small impad on the accuracy of the performance estima tes. The
rccommended value for each calibration factor is identified at the relevant point
during presentation of the methodology.

Traffic Characteristics Data

This subsection describes the traffie characteristics data listed in Exhibit 19-
31. These data describe the traffic streams traveling through the interst'ction
during the study periodo Thc demand flow rate of motorized vehicles and RTOR
flow rate are defined in Section 3 for the motorized vehicle mode.

Permitted Left-Turn Flow Rate

The permitted left-turn fluw rate is defined as the count of vehicles that turo
left permissively, divided by thc analysis period duration. lt is expressed as an

Chapter 19/5ignalized Intersections Pedestrian Methcxlology

Version 6.0 Page 19-73
Highway Capacify Manual: A Guide for Multimodal Mobility Analysis

hourly flow rate but may represent an analysis period shorter than 1 h. A
permitted left-tum movement can occur with either the permitted or the
protected-permitted lefHum mode. For left-rurn movements served by the
permitted mode, the permitted left-tum flow rate is equal to the left-tum
demand flow rate.
Defuult Vulue. For left-tum movements served by the protected-permitted
mode, the permitted left-tum flow rate should be measured in the ficld occause
its value is influenced by many factors. However, a default flow rate can be used
if the analysis involves future conditions or if the permitted lefHum flow rate is
not known from field data.
The default permitted left-rurn flow rate for movements served by the
permitted mode is equal to the left-tum demand flow rate.
The default permitted left-tum flow rate for movements scrved by thc
protected-permittcd mode is equal to the left-tum arrival rate during the
permitted period. This arrival rate is estimated as the left-tum flow rate during
the effective red time [Le.,qT = (1 - P) q CIr].

Midsegment 85th Percentile Speed

The 85th percentile speed represents the speed of the vehide whosc specd is
exceeded by only 15%of the population of vehides. The speed of interest is that
of vehides traveling along the street approaching the subject intersection. lt is
measured at a location sufficientiy distant from the intersection that speed is not
influenced by intersection operation. This sp~d is likely to oc influenced by
traffic conditions, so it should reflect the conditions present during the analysis
periodo

Pedestrian Flow Rate

The pedestrian flow rate rcpresents the count of pedestrians traveling
through each comer of the intersection divided by the analysis period duration.
It is expressed as an hourly flow rate but may represent an analysis period
shorter than 1 h. This flow rate is provided for each of five movements at cach
intersection comer. These five movements (Le.,vri' v"" Vd;' Vd<>'and V",b) are shown
in Exhibit 19-32as they oceor at one intersection comer.

Geometric Design Data

This subsection describes the geometric design data listed in Exhibit 19-31.
These data describe the geometric elements that influeoce intersection
performance from a pedestrian perspective. The number-of-Ianes variable is
defined in $ection 3 for the motorized vehide mode.

Number of Rjght~ Turn Islands

The number of right-tum islands represents the count of channelizing
islands encountered by pedestrians while crossing one intersection legoThe
island should be delineated by a raised curb and of sufficient size to be
considered a refoge for pedestriaos. The number provided must have a value of
O, 1, oc 2.

Pedestrian Methodology O1apter 19/5ignalized Intersections

Page 19-74 v~6.0
 Highway Capacify Manual: A Guide for Multimodal Mobility Analysis

@ Exhibit 19-32
intersection Comer Geometry
anó Pedestrian Movements

Ared e O.21S~

Minar Street
W, v •• pedestrian flow

r Crosswdl1c.
;

0=
W=
L=
R ••
& lnbound flow
outbound flow
crosswalk width
crosswdl1c.length
corner radius

Total Walkway Width, Crosswalk Width and Length, and Comer Radius
The geometric dcsign data of total walkway width, crosswalk width and
length, and comer radius describe the pedestrian accommodations on each
comer of the intcrsection. These data are shown in Exhibit 19-32.The total
walkway width (Le., W. and W~)is measurcd from the outside edge of the road
pavement (or face of curb, if present) to the far edge of the sidewalk (as
sometimes delineated by a building face, fence, or landscaping).
The crosswalk width (Le.,W, and Wd) represcnts an effective width. Unless
there is a known width constraint, the crosswalk's effective width should be the
same as its physical width. A width constraint may be found when vehicles are
observed to encroach regularly into the crosswalk area or when an obstruction in
the median (e.g., a signal pole oc reduced~width cut in the median curb) narrows
the walking spacc.
Thc crosswalk length (Le.,Le and Ld) is measured from outside edge to
outside edge of road pavement (or curb to curb, if prescnt) along the marked
pedcstrian travel path.

Signal Control Data

This subsection describes thc data in Exhibit 19-31that are identified as
signal control. The walk, pedestrian clear, yeJlow change, and red clearance
seUings are defined in Section 3 for the motorized vehicle mode.

Rest in Walk
A phase with the rest.in-walk mode enabled will dwell in walk as long as
there are no conflicting calls. When a conflicting call is received, the pedestrian
clear intcrval will time to its setting value befare ending the phase. This mode

O1apter 19j5ignalized Intersections Pedestrian Methodology

Version 6.0 Page 19-75
Hjghway Capacity Manual: A Guide for Multimodal MobiJity Analysis

can be enabled foc any actuated phase. Signals that operate with coordinated-
actuated operation may be set to use a coordination mode that enables the rest-
in-walk mode. Typically, the rest-in-walk mode is not enabled. In this case, the
walk and pedestrian clear intcrvals time to their respective setting values, and
the pedestrian signal indication dwells in a steady DON'TWAlKindication until a
conflicting call is received.

Cycle Length
Cycle length is predetermined for pretimed or coordinated-actuated control.
Chapter 31 provides a procedure for estimating a reasonable cycle length for
these two types of control when cycle length is unknown. Default values for
cycle length are defined in Section 3 of the present chapter for the motorized
vehicle mode.
For semiactuated and fully actuated control, an average cycle length must be
provided as input to use the pedestrian oc bieycle methodologies. This length can
be estimated by using the motorized vehicle methodology.

Duration of Phase 5erving Pedestrians

The duration of each phase that serves a pedestrian movement is a required
input. This phase is typically the phase that serves the through movement that is
adjacent to the sidewalk and for which the pedestrian, bicycle, and through
vehicle travel paths are parallel. Foc example, Phases 2, 4, 6, and 8 are the phases
serving the pedestrian movements in Exhibit 19-2.

Pedestnan Signal Head Presence

The presence of a pedestrian signal head influences pedestrian crossing
behavior. lf a pedestrian signal head is provided, then pedestrians are assumed
to use the crosswalk during the WAlKand flashing DON'TWAlKindications. lf no
pedestrian signal heads are provided, then pedestrians will cross during the
green indication provided to vehicular traffic.

OVERVIEW OF THE METHODOLOGY

This subsection describes the methodology for evaluating the performance of
a signalized intersection in terms of its service to pedestrians. The methodology
is applied through a series of five steps that determine the pedestrian LOS for a
crosswalk and assodated comers. These steps are illustrated in Exhibit 19-33.

Pedestrian MethodoIogy Chapter 19/5ignalized IntersectiOns

Page 19-76 Version 6.0
 Highway Capacity Manual: A Guide tar Multimodal MobiUty Analysis

Exhibit 19-33
1 Pedestrian Methodology for
Step 1: Determine Street Corner Signalized intersections
Step 3: Determine Pedestrian Delay
CirOJlation Area

l I
Step 2: Determine Crosswalk Step 4: Determine Pedestrian LOS
Circulation Area Score for Intersection
I
I
Step 5: Determine LOS

The methodology is focuscd on the analysis of signalized intersection

performance. Chapter 18, Drban Street Segments, and Chapter 20, Two-Way
STOp.Controlled Intersections, describe methodologies for evaluating thc
performance of these system elernents with respect to the pedestrian mode.

COMPUTATIONAL STEPS

Step 1: Determine Street Corner Circulation Area

This step describes a procedure for evaluating the performance of one
intersection comer. It is repeated for each intersection comer of interest.
The analysis of drculation area at the street corners and in the crosswalks
compares available time and space wHh pedestrian demando The product of time
and space is the critical parameter. It combines the constraints of physical design
(which limits available space) and signal operation (which limits avaiJable time).
This parameter is referred to as time-space.

A. Compute Available Tíme-Space

The total time-space available for circulation and queuing in the interscction
comer equals the product of the net comer area and the cyde length C. Equation
19-51 is used to compute the time-space available at an intersection comer.
Exhibit 19-32identifies the variables used in the equation.
TScomer = C(Wa W¡, - 0.215 R2) Equation 19-51

where
TSrorrv<' = available comer time-space (ftLs),
e cyde length (s),
W. total walkway width of Sidewalk A (ft),
W. total walkway width of Sidewalk B (ft), and
R radius of comer curb (ft).
If the comer curb radius is larger than either W. oc W¡...then the variable R in
Equation 19-51should equal the smaller of W. or W.,

Chapter 19/5ignalized intersections Pedestrian MetIlodology

Version 6.0 Page 19-77
 Highway Capacify Manual: A Guide for Multimodal Mobility Analysis

8. Compute Holding-Area Waiting Time

£i""'",~ '.+l
. \ Crosswlllk D
The average pedestrian holding time represcnts the average time that
pedestrians wait to cross the street when departing from the subject comer. The
Mlrlor equation for computing this time is based on the assumption that pedestrian
street arrivals are uniforrnly distributed during the eycle.

r
Mlljor
"n"t Condit:ion 1: Minor-Street Crossing
Far Condition 1 (shown in Exhibit 19-29),Equation 19-52, with Equation 19-
53, is used to compute holding-area time for pedestrians waiting to cross the
majar street.
2
Equation 19-52 _ Ndo(C - 9Walk.mi)
Q tdo- 2C
with

Equation 19-53

where
Q,~ total time spent by pedestrians waiting to cross the major street during
one eycle (p-s),
number of pedestrians arriving at the comer during each cycle to cross
the majar street (p),
effective walk time for the phase serving the minor-street through
movement (s),
e cycle length (s), and
flow rate of pedestrians arriving at the comer to cross the major street
(p/h).
Research indica tes pedestrians typically continue to enter intersections with
pedestrian signal heads during the first few seconds of the pedestrian clear
interval (26, 32). This behavior effectively increases the effective walk time
available to pedestrians. A conservative estimate of this additional walk time is
4.0 s (26). A nonzero value for this additional time implies sorne pedestrians are
initiating their crossing during the flashing DON'TWALK indication.
The following guidance is provided to estimate the effective walk time on the
basis of these research findings (26, 32). If the phase providing service to the
pedestrians is either (a) actuated with a pedestrian signa! head and rest in walk is
llOt enabled or (b) pretimed with a pedestrian signal head, then Eguation 19-54is

used.
Equation 19-54 9Walk,mi = Walkmi + 4.0
If the phase providing service to the pedestrians is actuated with a
pedestrian signal head and rest in walk is enabled, then Equation 19-55is uscd.
Equation 19-55 gWaJk,mi = Dp•mi - Ymi - Re.mi - PCmi + 4.0
If otherwise (Le., there is no pedestrian signa! head), Equation 19-56is used.
Equation 19-56
9Walk,mi = Dp,mi - Ymi - Re.mí

Pedestrian Methodology Chapter 19/5ignalized Intersections

Page 19-78 VetWn6.0
 Highway Capac;ty Manual: A Guide for Multimodal Mobility Analysis

whcre
gW.U,mt effective walk time for the phase serving the minor-strt.'et through
movement (s),
Walkm,: pedestrian walk setting for the phase scrving the minor-strcet through
movement (s),
pem, pedestrian clear setting for thc phasc scrving the minor-strcet through
movement (s),
Dp,m¡ duration of the phasc serving the minor-street through mavement (s),
Ym; yellow change interval of thc phasc scrving the minor-strt.'et through
movement (s), and
R(,m, red clearance interval of the phase scrving the minor-street through
movement (s).
The effective walk time estimated with Equation 19-54 oc Equation 19-55can
vary widely among intersections (26, 32). At a given intersection, the additional
walk time can vary from 0.0 s to an amount equal to the pedestrian clear interva1.
The amount of additional walk time used by pedestrians depends on many
factors, including the extent of pedestrian delay, vchicular volume, leve! of
enforcement, and the presence of countdown pedcstrian signa! heads.
The effective walk time estimated with Equation 19-54or Equation 19-55is
considered to be directly applicable to design or planning analyses because it is
conservativc in the amount of additional walk time it includes. A larger value of
effcctive walk time may be applicable to an operational analysis if (a) field
observation or experience indica tes such a value would be consistent with actual
pedestrian use of the flashing DON'TWALK indication; (b) an accurate cstimate of
pedestrian delay or qucue size is desired; or (e) the predicted performance
estima tes are understood to reflect sorne ilIegal pedestrian behavior, possibly in
response to constrained spaces or inadequate signal timing.

Condition 2: Major-Street Crossing

Holding oreo
For Condition 2, as shown in Exhibit 19.30, the previous equations are
repeated to compute the holding-area time for pedestrians waiting to cross the
minor street QIro' For this application, the subscript leUers do are replaced with L
Minar
the letters co to denote the pcdestrians arriving at the comer to cross in
"'eet

r
Crosswalk C. Similarly, thc subscript letters mi are replaced with mj to denote
signal-timing variables associatcd with the phase serving the major-street Mojar
through movemcnt. "'~t

e Compute O'rculation nme-Space

The time-space availab!e for circulating pedestrians equals the total available
time--space minus the time-space occupied by the pedestrians waiting to cross.
The latter value equals the product of the total waiting time and the area used by
waiting pedestrians (: 5.0 ft2/p). Equation 19.57 is used to compute the time-
space available far circulating pedestrians.

O1apter 19j5ignalized Intersections Pedestrian Methodology

VefSiOO6.0 Page 19-79
 Highway Capacify Manual: A Gujde for Multimodal Mobmfy Analysis

Equation 19-57 TS, = TSoomer - [5.0 (Qtdo + Quo)]

where T5< is the time-space available for circulating pcdestrians (ftLs).

D. Compute Pedestrian Comer arro/atian Area

The space required for circulating pedestrians is computed by dividing the
time-space available for circulating pedestrians by the time pedestrians consume
walking through the comer area. The latter quantity equals the total circulation
volume muItiplied by the assumed average circulation time (= 4.0 s). Equation
19.58, with Equation 19-59, is used to compute comer circulation area.
TS,
Equation 19-58 Mcomer = 4 ON
. tot
with
Equation 19-59

where
comer circulation area per pedestrian (ft2/p),
total number of circulating pedestrians who arrive each cycle (p),
flow rate of pedestrians arriving at the comer after crossing the minor
street (plh),
flow rate of pedestrians arriving at the comer to cross the minor street
(plh),
Vd; flow rate of pedestrians arriving at the comer after crossing the majar
strl.'('t(p/h),
V •. h = flow rate of pedestrians traveling through the comer from Sidewalk A
to Sidewalk B, or vice versa (p/h), and
aHother variables are as previously defined.
The circulation area obtained from Equation 19.58 can be compared with the

, Vd+L
ranges provided in Exhibit 19-28 to make some judgments about the
~V.,
.
performance of the subject intersection comer.

Mlnor CrosswlllkO Step 2: Determine Crosswalk Circulation Area

mee!
This step describes a procedure for evaluating the performance of one
crosswalk. It is repeated for each crosswalk of interest.
The procedure that follows describes the evaluation of Crosswalk O in
Exhibit 19-30(Le.,a crosswalk across the major street). The procedure is repeated

Vro .Ú'OSSwalk e
L
Minor
to evaluate Crosswalk C in Exhibit 19-29. For the second application, the
subscript letters do and di are replaced with the letters co and ci, respectively, to
denote the pedestrians associated with Crosswalk C. Similarly, the subscript
v"
Major
m~t
r
street
leHer d is replaced with the letter e to denote the length and width oi Crosswalk
C. Also, the subscript letters mi are replaced with mj to denote signal-timing
variables associated with the phase serving the major-street through movement.

Pedestrian Metho::lology Chapter 19/5ignalized Intersections

Page 19-80 Version 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

A. Establish Walking Speed

The average pedestrian walking speed SI' is needed ta evaluate carner and The rerommended walking
speeds reflect average (50th
crosswalk performance. Research indica tes walking speed is influenced by percentile) walking speeds tor
!he purposes of calculating
pedestrian age and sidewalk grade (26). If 0% ta 20% al pedestrians traveling LOS. Traffic signa! tímfng tor
along the subject segment are elderIy (Le.,65 years of age or alder), an average pedestrians is typically based
011 a 15th percentífe wa!king
walking speed of 4.0 ftis is recommended for intersection evaluation. If more
>peed.
than 20% of aH pcdestrians are elderIy, an average walking speed af 3.3 ft/s is
recommended. In addition, an upgrade af 10% or greater reduces walking speed
by 0.3 ft/s.

B. Compute Available Time-Space

Equation 19-60is used ta compute the time-space available in the crosswalk.
TScw = Ld Wd 9Walk,mi Equation 19.60

where
available crosswalk time-space (ft2-s),
L, length of Crosswalk D (ft),
W, effective width af Crosswalk D (ft), and
effective walk time far the phase serving the minor-street through
movement (s).

e Compute Effective Available Time-Space

The available crosswalk time-space is adjusted in this step to account for the
effect turning vehicles have on pedestrians. This adjustment is based on the
assumed occupancy of a vehicle in the crosswalk. The vehicle occupancy is
computed as the product of vehicle swept-path, crosswalk width, and the time
the vehicle preempts this space. Equation 19-61through Equation 19-63are used
for this purpose.
Equation 19-61

with
TStv = 40 Ntv Wd Equation 19-62

vU.perm + vrt - vrtQr

Ntv = 3,600 e Equation 19-63

where
TS:'" effective available crosswalk time-space (ftLs),
w,

J=t
TS", time-space occupied by turning vehicles (f12-s),
N", number of turning vehicles during the walk and pedestrian c1ear
intervals (veh),
L
Vll,,........
Vrt

v,..".
permitted Idt-tum demand flo\\' rate (veh/h),
right-tum demand flo\\' rate (veh/h), and
right-tum-on-red flow rate (veh/h).
"-l
- -- "--
v" "nd

r
VIft>"

O1apter 19j5ignalized Intersections Pedestrian Methodology

Ve~6.0 Page 19-81
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

The constant "40" in Equation 19-62represents the product of the swept-path

for most vehicles (= 8 ft) and the time a tuming vehicle occupies the crosswalk (=
5 s). The left-tum and right-tum flow rates used in Equation 19-63 are those
associated with movements that receive a green indication concurrently with the
subject pedestrian crossing and tum across the subject crosswalk.

D. Compute Pedestrian 5ervice Time

Total servicc time is computed with either Equation 19-64or Equation 19-65,
depcnding on the crosswalk width, along with Equation 19-66.This time
represents the elapsed time starting with the first pedestrian's departure from the
comer to thc last pedcstrian's arrival at the far side of the crosswalk. In this
manner, it accounts for platoon size in the sen'ice time (33).
lf crosswalk width Wd is greater than 10 ft, then
L N
Equation 19.64 t PS,dO' =32+~+27 ped.do
S'W p ,

If crosswalk width Wd is less than or equal to 10 ft, thcn

L,
Equation 19.65 tps,do = 3.2 + S + 0.27Nped.do
p

with

Equation 19-66 e- 9Walk,mi

Nped,do = Ndo e
where
tt"d" service time for pedestrians who arrive at the comer to cross the major
street (s), and
Nptd,d. = number of pedestrians waiting al the comer to cross the majar street
(p).
Equation 19.66 estimates the numbcr of pedestrians who cross as a group
following the presentation of the WALKindication (or green indication, if
pedestrian signal heads are not provided). It is also used to compute N¡w.di for the
other travel direction in the same crosswalk (using N djJ as defined below).
Equation 19-64or Equation 19-65is used to compute the service time for
pedestrians who arrive at the subject comer having waited on the other comer
before crossing the major strcet t,...di (using Nptd,d¡)'

E. Compute Crosswa/k OCcupancy Time

The total crosswalk occupaney time is computed as a product of the
pedestrian service time and the number of pedestrians using the crosswalk
during ane signal cycle. Equation 19-67is used, with Equation 19-68and resuIts
from previous steps, for the computation.
Equation 19-67
Toce = tps.doNdo + tps,diNdi
with

Equation 19-68

Pedestrian Methodology O1apter 19/5lgnallzed Intersections

Page 19-82 Version 6.0
 Hjghway Capadty Manual: A Guide for Mu/timodal Mobility Analysjs

where
Toce crosswalk occupancy time (p-s), and
Nd, number of pedestrians arriving at the comer each cycle having crossed
the majur street (p).

F. Compute Pedestrian Cro5Swalk Circulation Area

The circulatiun space provided for each pedestrian is determined by dividing
the time-space available for crossing by the total occupancy time, as shown in
Equation 19-69.
TS;w
Mew=-- Equation 19-69
Toce
where M<"l<' is the crosswalk circulation area per pedestrian (fF/p).
The circulation arca obtained from Eguation 19-69 can be compared with the
ranges provided in Exhibit 19.28 to make sorne judgmcnts about the
performance of the subject-intersection crosswalk (for the specified direction of
travel). For a complete picture of the subject crosswalk's performance, the
procedure described in this step should be repeated for the other direction of
traveI aIong the crosswaIk (i.e., by using the other comer associated with the
crosswalk as thc point of rcference).

Step 3: Determine Pedestrian Delay

This step describes a procedure fur evaIuating the performance of a
crosswalk at the intersection. H is repeated for each crosswalk of interest.
The discussion that follows describes the evaIuation of CrosswaIk D shown
in Exhibit 19-30. Thc procedure is applied again to evaluate Crosswalk C shown
in Exhibit 19-29. For the second application, the subscript letters mi are rcplaced
with mj to denote signal-timing variables associated with thc phase serving the
major-street through movement.
The pedestrian deJay while waiting to cross the major street is computed
with Eguation 19-70.
d _ (C-gwalk.mi)2
p - 2 e Equation 19-70

where dp is pedestrian delay (s/p). The delay obtained from Eguatiun 19-70
applies egually to both directions of trave1 along the crosswalk.
Research indicates average pedestrian delay at signalized intersection
crossings is not constrained by capacity, even when pedestrian flow rates reach
5,000 p/h (26). For this reason, del ay due to oversaturated conditions is not
induded in the value obtained from Eguation 19-70.
If the subject crosswalk is dosed, then the pedestrian del ay dp is estimated as
the value obtained from Equation 19-70 for the subject crosswalk, plus two
increments of the delay from this eguation when applied to the perpendicular
cross\vaIk. This adjustment reflects the additional delay pedestrians incur when
crossing the other three Iegs of the intersection so they can continue walking in
the desired direction.

Chapter 19j5ignalized intersections Pedestrian Methodology

Version 6.0 Page 19-83
Highway Capacify Manual: A Guide tor Multimodal Mobilify Analysis

The pedestrian delay cornputed in this step can be used to make sorne
judgment about pedestrian compliance. In general, pedestrians become
impatient when they experience delays in excess of 30 slp, and there is a high
likelih<XJdof their not complying with the signal indication (34). In contrast,
pedestrians are very likely to comply with the signal indication if their expectcd
deJay is less than 10 s/p.

5tep 4: Determine Pedestrian lOS SCOretor Intersection

This step describes a procedure for evaluating the performance of one
crosswalk. It is repeated for each crosswalk of interest.
The procedure that follows describes the evaluation of Crosswalk D in
Exhibit 19.30. The procedure is repeated to evaluate Crosswalk C in Exhibit 19-29.
For the second application, the subseript letter d is replaced with the letter e to
denote the length and width of Crosswalk C. AIso, the subscript letters mj are
replaced with mi to denote variables associatcd with thc minor strcct.
The pedestrian LOS seore for the intersection Ip.inl is calculated by using
Equation ]9-71 through Equation ]9-76.
Equation 19-71 Ip,lnt = 0.5997 + Fw + Fu + Fs + Fdelay
with
Equation 19-72 Fw = 0.681(Na)O.514

Fv = 0.00569
Vrtor +4VIt,perm) - Nrtci,d
(
0.0027 n15,mj - 0.1946
)
Equation 19-73 (

Equation 19-74 Fs = 0.00013 n15,mj S85,mj

1
Equation 19-75 Fdelay = 0.0401ln(dp,a)
0.25 '\'
Equation 19-76 n15,mj = N L Vi
d iEmd
where
I,.in' pedestrian LOS seore for intersection,
F., cross-scction adjustment factor,
Fp motorized vehide volume adjustment factor,
Fs '" motorized vehide speed adjustment factor,

F d<Loy ¡x'destrian delay adjustment factor,

In(x) natural logarithm of x,
N, number of traffic lanes crossed when traversing Crosswalk D (In),
number of right-turn channelizing islands along Crosswalk D,
count of vehides traveling on the major street during a lS-min period
(vcMn),

lMovements
t
crossing
v, demand flow rate for movement i (veh/h),
set of all motorizcd vehide movements that cross Crosswalk D (St.-'e
CrosswZllk D figure in margin),

Pedestrian MethodoIogy O1apter 19{5ignalized Intersections

Page 19-84 V~6,O
 Highway Capacity Manual: A Guide for Mulfimodol Mobility Anolysis

585''''i 85th percentilc s¡x.'Cdat a midsegment lacation on thc majar street

(mi/h), and
dr,d = pedestrian dclay when traversing Crosswalk O (s/p).
The leH-turn now rate VIr,pt"" used in Eguation 19-73 is the £low rate
associated with the lefHurn movement that receives a green indication
concurrently with the subject pcdestrian crossing and turns across the subject
crosswalk. The RTOR now rate V,~"is the naw rate associated with the approach
being crosscd and that also turns across the subjcct crosswalk. lt is not the same VI,.peml --
v__ used in Eguatian 19-63.

The pcdestrian LOS score obtained from this equation applies equally to
both directions of travel along thc crosswalk.
l
Nrtri< the variable for number of right-turn channelizing islands, is an intcgcr
with a value of O, 1, or 2.

Step S: Determine lOS

This step describes a process for determining thc LOS of one crosswalk. It is
repeated for each crosswalk of interest.
The pedestrian LOS is determined by using the pedestrian LOS scorc from
Step 4. This performance measure is compared with the thresholds in Exhibit 19-
9 to determine the LOS for the subject crosswalk.

Chapter 19{5ignalized Intersections Pedestrian Methodology

VefSOn 6.0 Page 1<}-S5
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

6. BICYCLE METHODOLOGY

This section describes the methodology for evaluating the quality of service
provided to bicyelists traveling through a signalized intersection.

SCOPE OF THE METHODOLOGY

The overall scope of the three methodologies was provided in Section 2. This
section identifies thc additional conditions applicable to the bicyele methodology.
• Target trave! modes. The bicyele methodology addresses travel by bieyele
through a signalized intersection. lt is not designed to evaluate the
performance of othec travel means (e.g., motorized bicycle, rickshaw).
• Shared or exclu.sive ianes. The bicycle methodology can be used to evaluate
the service provided to bieyclists when sharing a lane with motorized
vehieles or when traveling in an exclusive bieycle lane.

Spatial Limits
Intersection performance is evaluated separately for each intersection
approach. Ullless otherwise stated, all variables identified ;11 this sllbsection are specific
to olle illtersectioll approach. The bicycle is assumed to travel in the street (possibly
in a bicycle Iane) and in the same direction as adjacent motorized vehicles.

Performance Measures
Performance measures applicable to the bicyele travel mode inelude bicyele
delay and bicyele LOS score. The LOS score is an indication of the typical
bicyelist's perception of the overall crossing experience.
LOS is also considered a performance measure. lt is useful for describing
intersection performance to elected officials, policy makers, administrators, or
the publiCoLOS is based on the bicyelist LOS score.

Limitations of the Methodology

This subsection identifies the known limitations of the bicyele methodology.
lf one or more of these limitations is believed to have an important influence on
the performance of a specific intersection, then the analyst should consider using
aItematÍve methods or tools for the evaluation.
The bicycle methodology does not account for the eHect of the following
conditions on the quality of service provided to bicyclists:
• Presence of grades in excess of 2% and
• Pcesence of railroad crossings.

REQUIRED DATA AND SOURCES

This subsection describes the input data necded foc the bicycle methodology.
These data are Usted in Exhibit 19-34. The second column (labeled Basis) of the
exhibit indicates whether the input data are needed foc each signal phasc, each
intersection approach, oc the intersection as a wholc. The exhibit also lists default
values that can be usecl if local data are not available (25, 30).

Bicycle Methodology Chapter 19/5ignalized Intersections

Page 19-B6 VeNion 6.0
 Highway Capacity Manual: A Gu;de tor Multimodal Mobility Analysis

Potential Data Exhiblt 19-34

Required Data and Units Basis Source(s) Suggested Default Value Required Input Data, Potential
TrafflC Characteristics Data Sources, and Default
Moterized ~ehicle(lemand f\ow rate (vehfh)
Values for Bicycle Analysis
A Rekl data, ¡¡ast rounts Must be provided
Bicyde f\ow rate (bicyl;les{h) A Fiekl data, ¡¡ast counts Must be pro~ided
Proportion of on-street parking occupied
A Field data 0.50 (it parking Iane present)
(decimal)
Geometric Design
Street widttl (ft) A Field data, <lefial photo Baseden a 12-ft lane width
Number ef lanes A Fiekl data, aerial piloto MI.lStbe provided
W>dthof OlrtSidettlrough Iane (ft) A Flekl data, aerial piloto 12ft
Width of bicycle lane {ft} A Fleld data, <lefial piloto S.Oft (it pro~ided)
W>dthof paved outside shoulder (ft) A Fleld data, aerial piloto Must be provided
Wldttl of striped parking lane (ft) A Field data, aerial piloto 8.0 ft (it present)
Sigf1iJ1 Control Data
Cycle Iength (s) I Field data Same as motoriled V€hide mode
Yellow change red dearance (s)' P Field data
"
.¡.

Duration of phase serving bicydes (s) P Field data Same as rnotOlized~ehicle mode
Other Data
IAnalys<Speriod duratioll (h)" 5et by analyst 0.25 h
Notes: A = approacll: ene ~alue for the intersection approad'l.
1 = inter'SeCtion:ene ~alue for!he inlersection.
P = phase: ene ~alue or coOOitionfor eilCh signal phase.
• Specific values of yellow change aOOred dearance should be determined by local guidelines er practice.
b Analysis penod duration i!; as deflned for Exhibit 19.11.

The data elements listed in Exhibit 19-34do not inelude variables that are
considered to represent calibration factars. Default values are provided for these
factors bccause they typically have a relatively narro\\' range of reasonable
values or they have a small impact on the arruracy of the performance estimates.
The recommended value for each calibration factor is identified at the relevant
point during presentation of the methodology.

Traffie Charaeteristics Data

This subsection describes the traffic characteristics data listed in Exhibit 19-
34. Thcse data describe the traffic streams traveling through the intcrsection
during the study periodo The demand f10wrate of motorized vehicles and bicyele
f10w rate are dcfined in Section 3 far the motorized vchicle mode.
The variable for the proportion of on-street parking ocrupied represents the
proportion of the intersection's right-side curb line that has parked vehieles
present during the analysis periodo It is based on a zone that extends from a
point 250 ft upstream of the intersection to the intersection and a sccond zone
that exlends from the intersection to a point 250 ft down.<;treamof the
intersection. If parking is not allowed in these two zones, then this proportion
equals 0.0.

Geometrie Design Data

This subsection describes the geomctric design data Iisted in Exhibit 19-34.
These data describe the geomctric elernents that influence intersection
performance from a bicyclist perspectiv€'oThe number-of.lanes variable is
defined in Section 3 for the motorizcd vehicle mode.

Chapter 19/5ignalized Intersections Bicycle Methodology

Ver-siOn6.0 Page 19-87
Highway Capacity Manual: A Guide tor Mu/timodal MobiJity Analysis

5lreet Width
The street width represents the width of the cross street as measured along
the outside through vehiele lane on the subject approach between the extended
curb line limits of the cross street. It is measured £or each intersection approach.

Widths of Outslde Through Lane, Bicycle Lane, Outside Shoulder,

and Parking Lane
The widths oí several individual elements oí the cross section are considered
input data. These elements inelude the outside lane that serves motorized
vehieles at the intersection, the bicyele lane adjacent to the outside lane (if used),
paved outside shoulder, and striped parking lane.
The outside lane width daes nat inelude the width of the gutter.1f curb and
gutter are present, then the width oí the gutter is included in the shoulder width
(i.e., shoulder width is measured to the curb face when a curb is present).

Signal Control Data

This subsection describes the data in Exhibit 19-34 that are identified as
signal control. The yellow change interval and red clearance interval settings are
defined in Section 3 for the motorized vehicle mode.

Cycle Length
Cyele length is predetermined for pretimed or coordinated.actuated control.
Chapter 31 provides a procedure for estimating a reasanable eyele length foc
these two types of control when cycle length is unknown. Default values for
cyele length are defined in Section 3 of the present chapter for the motorized
vehicle made.
For semiactuated and fully actuated control, an average cyele length must be
provided as input to use the pedestrian or bicycle mcthodologics. This length can
be estimated by using the motorized vehiele methodology.

Duration of Phase Serving Pedestrians and 8icycles

The duratian of each phase that serves a bicycle movement is a required
input. This phase is typically the phase that serves the through movement that is
adjacent to the sidewalk and for which the bicyele and through vehiele travel
paths are parallel. For example, Phases 2, 4, 6, and 8 are the phases serving the
bieycle movements in Exhibit 19-2.

OVERVIEW OF THE METHODOLOGY

This subsection describes the methodology for cvaluating the performance of
a signalized intersection in terms of its service to bieyclists. The methodolagy is
applied through a series oí three steps that determine the bieyele LOS for an
intersection approach. These steps are illustrated in Exhibit 19-35.
The mcthodology is focused on analyzing signalized intersection
performance from the bicyclist poiot of view. Chapter 18, Drban Street Segments,
describes a methodology íor cvaluating urban street performance with respect to
the bicycle mode.

Bicycle Methodology Chapter 19/5ignalized Intersections

Page 19-88 V~6.0
 Highway Capacify Manual: A Guide for Multimodal Mobility Analysis

Exhibit 19-35
Step 1: Determine Blcyde Delay Bicyde Methodology for
Signalized Intersections

I
Step 2: Determine Bicyde LOS Score
for Intersection

I
Step 3: Determine LOS

COMPUTATlONAL STEPS
5tep 1: Determine Bicycle Delay
This step describes a procedure for evaluating the performance of one
intersection approach. It is repeated for cach approach of interest.
Bicyele delay can be cakulated only (or intersection approaches that have an
on-slrecl bicyele lane oc a shoulder that can be uscd by bicyclists as a bicyele
lane. Bicyclists who share a lane with motorized vehicle traffic will incur the
same del ay as the motorized vchicles.

A. Compute Sicycle Lane Capacity

A wide range of capacities and saturalion flow rates have been reported by
many countries for bicyelc lanes at intersections. Research indicates the base
saturation flow rate may be as high as 2,600 bicyeles/h (35). However, few
intersections provide base conditions for bieyclisls, and current information is
insufficient to calibrale a serics of appropriale saturation £low adjustmenl factors.
Until such factors are developed, it is recommended that a saluration flow rate of
2,000 bieycles/h be used as an average value achievablc at most intersections.
A saturation How rate of 2,000 bicycles/h assumes right-tuming motor
vehicles yield the right-of-way to through bicyclists. Where aggressive right-
tuming traffic exists, 2,000 bicycles/h may not be achievable. Local observations
to determine a saturation £Iowrate are recommended in such cases.
The capacity of the bicyele lane al a signalized interscction may be computed
with Equation 19-77.

Equation 19-77

where
Co capacity of the bicycle lane (bicycles/h),
Sb saturation £Iowrate of the bicycle lane = 2,000 (bicyeles/h),
gb effective '('n
gfl.. time for the bicycle ¡ane (s), and
e cycle lcngth (s).
The effective grecn time for the bicydc lane can be assumed to equal that for
the adjacent motor vehiele traffic stream that is scrvcd concurrently with the
subject bicycle lane (i.e., go '" Dr -/1 -1z).

O1apter 19/5ignalized IntersectiOfls Bicycle Methodology

Version 6.0 Page 19-89
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

B. Compute Bicycle De/ay

Bicycle delay is computed with Equation 19-78.
0.5 C (1 - B,fC)'
Equation 19-78 db = ---------
1 - min (V;~c,
1.0)~

where db is bicyde delay (s/bicycle), vbicisbicycle flow rate (bicyeles/h), and aH

other variables are as previously defined.
This delay equation is based on the assumption there is no bieyde
incremental delay oc initial queue delay. Bicyclists will not normally toJerate an
oversaturated condition and will select other routes oc ignore traffic regulations
to avoid the associated delays.
At most signalized intersections, the only deJay to through bieycles is caused
by the signa!, because bieycles have the right-of-way over right-turning vehides
during the green indication. Bieycle delay could be longer than that obtained
fram Equation 19-78 when (a) bieycles are forced to weave with right-turning
traffic during the green indication or (b) drivers do not acknowledge the bicycle
right-of-way because of high flows of right-turning vehides.
The delay obtained from Equation 19-78 can be used to make sorne judgment
about intersection performance. Bieyclists tend to have about the same tolerance
for deJay as pedestrians. They tend to bccome impatient when they experience a
deJay in exeess of 30 s/bieycle. In contrast, they are very Iikely to comply with the
signal indication if their expected deJay is less than 10 s/bieyele.

Step 2: Determine Bicycle LOS Score for Intersection

This step describes a procedure for evaluating the performance of one
intersection approach. It is repeOlted for eOlchOlpproach of intcrest. The bieydc
LOS score can be cOllculatedfor Olnyintersedion Olpproach, rcgardless of whether
it has an on-strcet bicycle lanc.
The bicycle LOS score for the intersection Ib,i"1 is calcuJated by using Equation
19-79 through Equation 19-82.
Equation 19.79 Ib•int = 4.1324 + Fw + Fv
with
Equation 19-80 Fw = 0.0153 WCd - 0.2144 Wt

Equation 19-81
F = 0.0066 Vjt
v
+ Vth + Vrt
4Nth
Equation 19-82 Wt = W01 + Wbl + Wo~+ IpkWpk
where
bicyele LOS score for intcrscction;
cross-section adjustment factor;
F, motorized vehide volume adjustment factor;
curb-to-curb width of the cross street (ft);

Bícyde Methodology Chapter Ig¡Signalized intersections

Page 19-90 VersiOn 6.0
 Highway Copacify Manual: A Guide for Mulfimodol Mobilify Analysis

Wt = total width of the outside through lane, bicycle lane, and paved
shoulder (ft);
VII left-turn demand flow rate (veh/h);
Vth through demand flow rate (veh/h);
vrt right-turn demand flow rate (veh/h);
Nlh number of through lanes (shared or exclusive) (In);
WaI width of the outside through lane (ft);

WbI width of the bicycle lane (""0.0 if bicycle lanc nol provided) (ft);
Wr< width of striped parking lane (ft);

lp1; indicator variable for on.street parking occupancy (= O if PpI; > 0.0,
1 otherwise);
PpI: proportion of on-street parking occupied (decimal);
W"" width of paved outside shoulder (ft); and
W", adjusted width of pavcd outside shoulder (if curb is present, W", = Woo
- 1.5 ~ 0.0; otherwise, W", = W",) (ft).

Step 3: Determine LOS

This step describes a process for determining the LOS of one intersection
approach. It is repeated for each approach of interest.
The bicycle LOS is determined by using the bicycle LOS score from Step 2.
This performance measure is compared with the thresholds in Exhibit 19-9 to
determine the LOS for the subject approach.

Chapter 19/5ignalized Intersections Bicycle Methodology

Versien 6.0 Page 19-91
Highway Capacify Manual: A Guide for Multimodol Mobility Anolysis

7. APPLICATIONS

EXAMPLE PROBLEMS
Chapter 31, Signalized Intersections: Supplemental, describes the application
of each oi the three methodologies through the use oi example problems. There
is one example problem associated with each methodology. The examples
illustrate the operational analysis type.

GENERAUZEO OAILY SERVICE VOLUMES

Exhibit 19-36 shows an illustrative generalized service volume table for a
signalized interscction. This particular exhibit has been prepared for illustrative
purposes muy and should not be used for any specific planning or preliminary
enginccring application because the values in the table are highly dependent on
the assumed input variables. Care must be taken in constructing atable that the
analyst believes is representative of a typical signalized intersection within the
planning area. In the example table, the volumes represent the total approach
volume (sum of the leit-, through, and right-turn movements). This particular
table illustrates how hourly service volumes vary with the number of through
lanes on the approach and the through movement glC ratio.

Exhibit 19-36 Through NO.of

IIIustrative Gef1eralized Movement Through
ServiCe Volumes for Signalized g/CRatio unes LOS8 LOS e LOSD lOS E
Intersections (veh/h) 1 130 610 730 800
0.40 2 270 1,220 1,430 1,550
3 380 1,620 1,980 2,000
I 320 720 840 910
0.45 2 630 1,410 1,610 1,740
3 840 1,780 2,000 2,250
I 490 830 940 1,020
0.5 2 940 1,580 1,790 1,930
3 1,180 1,930 2,000 2,500
Notes: LOS E threshold is delined by control delay greater than 80 s/veh or voIume-tlKapacity
ralíO >1.0.
Assumed values for all entries:
Heavy vehicles: 0%
Peak hour factor: 0.92
Laroewidth: 12 ft
Grade: 0%
5eparate Ief'l:-tum larle: Ves
Separate right.turn larle: no
Pretimed control
Cyde Iength: 90 s
Lost time: 4 S/phase
Protecteclleft-turn phasing: ves
g/e ralíO for Ieft.turn movement: 0.10
Parking maneuve<s per hour: O
Buses stopping per hoor: O
Percenlage Ief'l:tvrns: 10%
Percentage right turns: 10%

The hourly service volumes could easily be converted to daily service

volumes with the application of appropriate K- and D-factors. Step-by-step
instructions are provided in Appendix B oi Chapter 6, HCM and Altcrnative
Analysis Tools, for users wishing to learo more about constructing one's own
service volume table.

Applications Chapter 19j5ignalized Intersections

Page 19-92 Vet5".OO 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

ANAL YSIS TYPE

The thrce methodologies described in this chapter can each be used in threc
types of analysis. These analysis types are described as operational, design, and
planning and preliminary enginecring. The charactcristics of each analysis type
are described in the subsequcnt paragraphs.

Operational Analysis
The objective of an operational analysis is to determine the LOS for current
or near-term conditions when details of traffic volumes, geometry, and traffic
control conditions are known. AlI the methodology steps are implemented and
aH calculation procedures are applied for the purpose of computing a wide range
of performance measures. The operational analysis type will providc the most
reHable resuIts because it uses no (or mini mal) dcfauIt values.

Design Analysis
The objective of a design analysis is to identify thc altematives that operate
at the target level of the specified performance measures (or provide a better
level of performance). The analyst may then recommend the "best" design
altemative after consideration of the fuH range of factors.
The design analysis type has two variations. 60th variations require
specifying the traffic conditions and target levels for a set of performance
measures. One variation requires the additional specification of the signalization
conditions. The methodology is then applied by using an iterative approach in
which altemative geometric conditions are separately evaluatcd.
The second variation of the design analysis requires the additional
specification of geometric conditions. The methodology is then applied by using
an iterativc approach in which altemative signaHzation conditions are evaluatcd.

Planning and Preliminary Engineering Analysis

The objective of a planning and preliminary engineering analysis can be (a)
to determine the LOS foc either a proposed intersection or an existing
intersection in a future year or (b) to size the overall geometrics of a proposed
intersection.
The level of prccision inherent in planning and preliminary engineering
analyses is typicaHy lower than for operational analyses bccause default vatues
are often substituted for field-measured values of many of the input variables.
Recommended default values for this purpose are described abovc in thc
sections associatcd with cach methodology.
The requirement for a complete description of the signal.timing plan can be a
burden for sorne planning analyses. The planning-Ievel analysis application
described in Chapter 31, Signalized lntersections: SupplementaL can be used to
estimate a reasonable timing plan, in conjunction with the defauIt values
provided.

Chapter 19/5ignalized Intersections Applications

Version 6.0 Page 19-93
Highway Capacity Manual: A Gujde for Multimodal Mobi/ity Analysjs

USE OF ALTERNATIVE TOOLS

General aftemative tooI General guidance for the use of altemative traffic analysis tools for capacity
guidaf1Ce is provided in
Ch<Jpters 6 and 7. and LOS analysis is provided in Chapter 6, HCM and A1temative Analysis Tools,
and Chapter 7, Interpreting HCM and Altemative Tool Results. This subscction
contains specific guidance for the application of alternative tools to the analysis
of signalized intersections. Additional information 00 this tapie is provided in
the Technieal Refereoce Library in Volume 4. The focus of this subsection is the
application of alternative tools to evaluate motorized vehicle operation.

Comparison of Motorized Vehic1e Methodology and Alternative Tools

Motorized Veh;c1eMethodology
The motorized vehicle mcthodology models the driver-vehicle-road-signal
system with reasonable accuraey for most applications. It accounts for the
operation of actuated phases, shared Janes, and permiUed tum movemeots. It
can account for the eHect of initial queues and signal progression on deJay.
The motorized vehicle methodology offers several advantages over
alternative analysis tools. One advantage is that it has an empirically calibrated
procedure for estimating saturation flow rateoAlternative tools often require
saturation flow rate as an input variable. A second advantage is that it produces
a direct estímate of capacity and volume-to-capacity ratio. These measures are
not directly available from simulation tools. A third advantage is that it produces
an expccted value for each of several performance measures in a single
application. Simu1ation tools require muItiple runs and manual calculations to
obtain an expected value for a given performance measure. A fourth advantage
is that its analytie procedures are described in the HCM so that analysts can
understand the driver-vehicle-road interactions and the means by which they
are modeled. Most praprietary alternative tools operate as a "black box,"
praviding Hule detail describing the intermediate calculations.

Altemative Tools
Deterministie tools and simulation tools are in common use as alternatives to
the motorized vehicle methodology offered in this chapter. Deterministic too1s
are often used for the analysis of signalized intersections. The main reasans for
their popularity are found in the user interface, optimization options, and output
presentation features. Sorne also offer additionaI performance measures such as
fuel consumption. air quality, and operating cos1.

Conceptual Differences
Conceptual differences in modeling approach may preclude the direct
comparison of performance measures from the motorized vehicle methodology
with those fram alternative tools. The treatment of random arrivals is a case in
poinL There is a common misconception among analysts that alternative tools
trcat random arrivals in a similar manner.
A simple case is used to demonstrate the different ways altemative tools
model random arrivals. Consider an isolated intersection with a two-phase
sequence. The subject intersection appraach serves only a through movement;

Applications Chapter 19/5ignalized Intersections

Page 19-94 Version 6.0
 Highway Capacity Manual: A Guide for Mulfimodal Mobilify Analysis

there are no turning movements from upstrcam intcrscctions or driveways. The

only parameter that is allowed to vary in this example is the cycle length (all
other variables are held constant).
The rcsults of this cxperiment are shown in Exhibit 19-37.The two solid lines
represent deJay estima tes obtained from the motorized vehicle methodology.
Uniform delay is shown to inerease linearly with cycle length. Incremental delay
is constant with respeet to cycle length because the voJume-to-eapacity ratio is
constant. As a result, control delay (the sum of the uniform deJay and
incremental deJay) is also shown to increase linearly with cycle length.

~ 60
70
("-í' + d,) r Exhibit 19.37
Effect of (yele Length on
oelay

-••
Con"" Qol.y
~
•
> 50 -
Simulated
Control DelllY
~•
---f
•••
40

e 30
o
_--~_I
--- I---~---

" 20
~
e
o U~fO!Tl!DelllY (di)
u 10
Incrementlll Delay (oi) = 21 s/veh
O
O 30 60 90 120 150 180

Cycle Length (s)

The dashed linc in Exhibit 19-37 represents the control delay estimate
obtained from a simulation-based analysis tool. The simulation-based tool shows
clase agreement with the motorized vehicle methodology for short eycles, but it
deviates for longer cycles. There are likely to be explainable rcasons for this
difference; howcver, thc point is that such differences are likely to exist among
tool5. The analyst should understand the underlying modeling assumptions and
limitations inherent in any tool (including the motorized vehicle methodology)
when it is used. Moreover, the analyst should fully understand the definition of
any performance measure used so as to interpret the results and observed trends
properly.

Alternative Tool Application Guidance

Development of HCM-Compatible Performance Measures
The motorized vehide methodology is uscd to predict control delay, which is
defined as the excess travel time caused by the aetion of the control device (in
this case, the signal). Simulation-bascd analysis tools often use a definition of
delay that differs from that used in the motorized vehide methodology,
especially for mavements that are oversaturatcd at sorne point during the
analysis. Therefore, sorne care must be taken in the determinatian of LOS when
simulation-based delay estima tes are used. Delay eomparison amang different
tools is discussed in mare detail in Chapter 7.

Chapter 19/5lgnalized Intersections Applications

Ve1SiOn 6.0 Page 19-95
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

An accurate estimate of control delay may be obtained from a simulation to01

by performing simulation runs with and without the control device(s) in place.
The segrnent deJay reported with no control is the delay due to geometrics and
intcraction between vehicles. The additional delay reported in the run with the
control in place is, by definition, the control delay.

Adjustment of Parameters
For applications in which either an alternative tool or the rnotorized vehicle
methodology can be used, sorne adjustrnent will generally be required for the
alternative tool if sorne consistency with the motorized vehicle methodology is
desired. For example, the parameters that determine the capacity of a signalized
approach (e.g., saturation flow rate and start-up last time) should be adjusted to
ensure the lane group (or approach) capacities from the alternative taol match
those estimated by the motorized vehicle methodology.

Sample Calculations IlIustrating Altemative Tool Applications

Chapter 31, Signalized Intersections: Supplemental, includes exarnples to
iIlustrate the use of simulation tools to address the stated Iimitatians of this
chapter. Specifically, these examples address the following conditions: left-turn
storage-bay overflow, RTOR operation, short through 1anes, and closely spaced
intersections.

Applications Chapter 19/5ignalized Intersections

Page 19-96 V£'fSiOn 6.0
 Highway Capacity Manual: A Guide for Multjmodal Mobility Analysis

8. REFERENCES

1. Signalized lllterseetioll Capacify Method. NCHRP Projcct 3-28(2). JHK & Sorne of t!lese references tAn
be found in Che Tedmk:iJf
Associates, Tucson, Ariz., Feb. 1983. Reference Library In Vofllme 4.

2. Signalized /nterseetiolJ Capacity Sfudy: Fina/ Report. NCHRP Project 3-28(2).

JHK & Associatcs, Tucson, Ariz., Dec. 1982.
3. Messer, C J., and D. 13.Fambro. Critical Lane Analysis for Intersection
Design. In Transporlation Researcl¡ Record 644, Transportation Research Board,
National Research Council, Washington, D.C, 1977, pp. 26-35.
4. Berry, D. S., and P. K. Gandhi. Headway Approach to Intersection Capacity.
In Higlmmy Researcll Record 453, Highway Research Board, National Research
Council, Washington, D.C, 1973, pp. 56-60.
5. Miller, A. Austra/ian Road Researeh Bu/letiIl3: The Capacity ofSigllalized
Interseetiolls in Australia. Australian Road Research Board, Kew, Victoria,
1968.
6. Webster, F., and B. Cobbe. Traffie Signa/s. Road Research Technical Paper 56.
Her Majesty's Stationery Office, London, 1966.
7. Petersen, B., and E. ¡mre. Swedish Capaeily Malllla/. National Swedish Road
Administration, Stockholm, Feb. 1977.
8. Catling, I. A Time-Dependent Approach to Junction Delays. Traffte
Engilleering ami Conlrol, Vol. 18, No. 11, 1977, pp. 520-526.
9. Akc;elik, R. Allstraliml Road Researeh Report 123: Traffie Signals: Capacify alld
Timing AlIalysis. Australian Road Research Board, Kew, Victoria, 1981.
10. Reilly, W., C Gardner, and J. Kell. A Teehnique for MeasllremCllt of De/ay at
/flterseetions. Report FHWA-RD.76.135/137. Federal Highway
Administration, Washington, D.C, 1976.
11. Courage, K. G., D. B. Fambro, R. Akc;elik, P.-S. Lin, M. Anwar, and F. Viloria.
NCHRP Web Documellt 10: Capacity Analysis ofTraffie-Aetuated lntersectiolls.
Transportation Research Board, National Research Council, Washington,
D.C, Dec. 1996.
12. Ake;elik, R. Estimation of Green Times and Cyele Time for Yehide-Actuated
Signals. In Trallsporfatioll Rescareh Record 1457, Transportation Research
Board, National Research Council, Washington, D.C, 1994, pp. 63-72.
13. Lin, F.-B. Estimation of Average Phase Durations for Full-Actuated Signals.
In TrallSportafioll Rescareh Record 881, Transportation Research Board,
National Hescarch Council, Washington, o.C, 1982, pp. 65-72.
14. Lin, F.-B. Predictive Models of Traffic-Actuated Cyde Splits. Transportafioll
Researeh, Yol. 16B, No. 5,1982, pp. 361-372.
15. Cowan, R. Useful Headway Models. Transportafioll Researeh, Vol. 9, No. 6,
1975. pp. 371-375.

Chapter 19j5ignalized intersections References

Version 6.0 Page 19-97
Highway Capacity Manual: A Gu;de tor Multimodal Mobility Analysis

16. Ak¡;elik, R., and E. Chung. Calibration of the Bunched Exponential

Distribution oí Arrival Headways. Road and Transport l~esearch, Vol. 3, No. 1,
1994, pp. 42-59.
17. Un, F.-B. Optimal Timing Settings and Detector Lengths of Presence Mode
Full.Actuated Control. In Transportation Research Record 1010, Transportation
Research Board, National Research Council, Washington, D.C., 1985,
pp. 37-45.
18. Milazzo, J. S., JI, N. M. RouphaiL J. E. Hummer, and D. P. Allen. Effect of
Pedestrians on Capacity oí Signalized lntersections. In Transportation Research
Record 1646, Transportation Research Board, National Research Council,
Washington, D.C., 1998, pp. 37-46.
19. AlIen, D. P., J. E. Hummer, N. M. Rouphail, and J. S. Milazzo n. Effect of
Bicyeles on Capacity of Signalized lntersections. In Transportation Research
Record 1646, Transportation Research Board, National Research Council,
Washington, D.C., 1998, pp. 87-95.
20. Rouphail, N., J. Hummer, J. Milazzo, and D. AlIen. Recommended Procedures
Jor Chapter 9, Signalized lntersections, of the Highway Capacity Manual. Report
FHWA-RD-98-106. Federal Highway Administration, Washington, D.C., 1999.
21. Strong, D. W., N. M. Rouphail, and K. G. Courage. New Calculation Method
for Existing and Extended Highway Capacity Manual Delay Estimation
Procedures. Presented al 85th Annual Meeting of the Transportation
Research Board, Washington, D.C., 2006.
22. Strong, D. W., and N. M. Rouphail. Incorporating Effects of Traffic Signal
Progression into Proposed Incremental Qucue Accumulation Method.
Presenled at 85th Annual Meeting of the Transportation Research Board,
Washington, D.C., 2006.
23. Prassas, E. S., and R. P. Roess. Left-Tum Adjustment foc Permitted Tums
from Shared Lane Groups: Another Look. In Transportation Research Record
1398, Transportation Rescarch Board, National Research Council,
Washington, D.C., 1993, pp. 75--81.
24. Bonncson, J. A, M. P. Pratt, imd M. A Vandehey. Predictillg the Performallce of
Automobile Traffic 011 Urban Streets: Final Report. National Cooperative
Highway Research Program Project 3-79. Texas Transportation Institute,
Texas A&M University, CoIlege Station, Jan. 2008.
25. Dowling, R. G., D. B. Reinke, A Flannery, P. Ryus, M. Vandehcy, T. A
Petritsch, B. W. Landis, N. M. Rouphail, and J. A. Bonneson. NCHRP Report
616: Multimodal Level of Seroice Analysis for Urball Sfreets. Transportation
Research Board of the National Academies, Washington, D.C., 2008.
26. Rouphail, N.,]. Hurnmer, J. Milano, and D. AlIen. Capacity Analysis of
Pedestriall and Bicycle Facilities: Recommended Procedures Jor the "Pedesfrians"
Chapter of fhe Highway Capacify Manlla/. Report FHWA-RD-98-107. Federal
Highway Administration, Washington, D.C., 1998.

References Chap1er19/5ignalized
Intersections
Page 19.98 Version 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

27. Rouphail, N., J. Hummer, J. Milazzo, and O. AlIen. Capacity AlIalysis of

Pedestriall alld Bicycle Facilities: Rccommended Procedurcs for thc "Bicycles"
Chapter of the Highway Capacify Manual. Report FHWA-RD-98-108. Federal
Highway Administration, Washington, D.C., 1998.
28. NatiOllal Transportation Commullications for ITS Protoco/: Object Definitiolls for
Actllated Traffic Sigila/ Control/er (ASO U"its-1202. National EJeetrical
Manufaeturers Association, RossJyn, Va., Jan. 2005.
29. Messer, C. J., and J. A. Bonneson. NCHRP Web Document 12: Capacify A/lalysis
of lnterchange Ramp Termillals. Transportation Research Board, NationaJ
Research Council, Washington, D.C., April1997.
30. Zegeer, J. D., M. A. Vandehey, M. B1ogg, K. Nguyen, and M. Ereti. NCHRP
Report 599: Default Va/ues Jor Highway Ca¡mcity and LeveI ofSeroice Ana/yses.
Transportation Research Board of the National Aeademies, Washington,
D.C., 2008.
31. Nevers, B., H. 5teyn, Y. Mereszczak, Z. Clark, N. RouphaiL J. Hummcr, B.
Schrocder, Z. Bugg, J. Bonncson, and D. Rhodes. NCHRP Reporf 707:
Guidelines 011 t1le Use of Auxiliary Througll talles at Sigllalized 1lltersections.
Transportation Research Board of the National Academies, Washington,
D.C., 2011.
32. Virkler, M. R. Pedestrian Compliance Effects on Signal DeJay. In
Transportation Research Record 1636, Transportation Research Board, National
Rcseareh Couneil, Washington, O.c., 1998, pp. 88--91.
33. Virkler, M. R. Scramble and Crosswalk Signal Timing. In Trmlsportatioll
Research Record 1636, Transportation Research Board, National Research
Couneil, Washington, D.C., 1998, pp. 83-87.
34. Dunn, R, and R Pretty. Mid-Block Pedestrian Crossings-An Examination of
Delay. Proc., 12th An/llw/ Australia" Road Research Board COllference, Habart,
Tasmania, Australia, Aug. 1984.
35. Opiela, K. S., S. Khasnabis, and T. K. Datta. Oetermination of the
Charactcristics of Bicyelc Traffie at Urban lnterseetions. In TransportafiO/I
Researcll Record 743, Transportation Research Board, National Research
CounciL Washington, D.C., 1980, pp. 30-38.

Chapter 19/5ignalized lntersections References

Version 6.0 Page 19-99
 Highway Capacity Manual; A Guide for Multimodal Mobility Analysis

CHAPTER 20
TWO-WAY STOP-CONTROLLED INTERSECTIONS

CONTENTS

1. INTR a D UCTI a N ......................................................................................•....•...•. 20-1

Chapter Organization . . 20-1

Related HCM Content . . 20-2

2. CONCEPTS .................................................•...........•...•........•..••..............................20-3
Intersection Analysis Boundaries and Travel Modes 20-3
Gap Acceptance Theory 20.4
LeveI-of-Servi ce Cri teria 20-6

3. MOTORIZED VEHICLE CORE METHODOLOGY 20-7

Scope of the Methodology 20-7
Required Input Data And Sources 20-10
ComputationaI Steps 20-11

4. EXTENSION TO THE MOTORIZED VEHICLE METHODOLOGY 20.33

lntroduction 20-33
Replacement Steps to Incorporate Pcdestrian Effects on Motorized
Vehicle Capacity 20-33

5. PEDESTRIAN MODE ..•..••...•.....................................................•..•...•...•....•.......

20.37
Scope of thc Methodology 20-37
Required Input Data And Sources 20-38
Computational Steps 20-38

6. BICYCLE MaDE 20-45

7. APPLICA TIONS .........•...•..••..••...•...................................................•...•...••..•........

20-46
Types of Analysis 20-46
Example Problems 20.46
Example Results 20-47

8. REFERENCES ......................•..••..•...•............................................•...••..••...............
20.49

Olapter 20fTwo-Way SToP-Controlled Intersections Contents

Version 6.0 Page 20-i
Hjghway Capacity Manual: A Guide for Multimodal MobiJity Analysis

LIST OF EXHIBITS

Exhibit 20-1 Yehicular and Pedestrian Movernents at a TWSC

lntersection 20-5
Exhibit 20-2 LOS Criteria: Motorized Yehide Mode 20-6
Exhibit 20-3 LOS Criteria: Pedestrian Mode 20-6
Exhibit 20-4 Lirnitations of the HCM TWSC Intersection Motorized
Yehide Proced ure 20-8
Exhibit 20-5 Required Input Data, Potential Data Sources, and Default
Yalues for TWSC Motorized Yehide Analysis 20-10
Exhibit 20-6 nvsc Intersection Methodology 2Q..11
Exhibit 20-7 Illustration of Conflicting Movernents Eor Major-Street Left-
Turn Moverncnts 20-13
Exhibit 20-8 Illustration of Conflicting Movernents Eor Minor-Street
Right- Tum Movements 20-14
Exhibit 20-9 Iliustration of Conflicting Movernents for Major-Street
U-Tum Movements 20-15
Exhibit 20-10 IIIustration of Conflicting Movernents Eor Minor-Street
Through Movements 20-16
Exhibit 20-11 lI1ustration of Conflicting Movements for Minor-Street
Left-Tum Movements 20-17
Exhibit 20-12 Base Critical Headways Ear TWSC Intersections 20-18
1
Exhibit 2Q..13Base Follow-Up Headways Eor lW"SC Intersections 2Q..19
Exhibit 2Q..14Proportion oE Analysis Period Blocked for Each Movement.. 20-20
Exhibit 20-15 Short Left-Tum Pocket on Major-Street Approach 20-23
Exhibit 20-16 Adjustment to Impedance Factors for Major-Street Left-
Tum Movement and Minor-Street Crossing Movement 20-26
Exhibit 20-17 Capacity of a Flared-Lane Approach 2Q..28
Exhibit 20-18 Relative Pedestrian-Yehide Hierarchy for Rank 2
Movements 2Q..34
Exhibit 2Q..19Relative Pedestrian-Vehide Hierarchy Eor Rank 3
Movernen ts 20-35
Exhibit 20-20 Relative Pedestrian-Yehide Hierarchy for Rank 4
Movernen ts 20-36
Exhibit 20-21 Limitations of the HCM TWSC Pedestrian Procedure 20-38
Exhibit 20.22 Required Input Data, Potential Data Sources, and DeEault
Values far lW"SC Pedestrian Analysis 20-38
Exhibit 20-23 TWSC Pedestrian Methodology 20-38
Exhibit 20-24 Effect of Pedestrian Crossing Treatments on Motorist Yield
Rates 20-42

Contents Chapter 20fTwo-Way 5T0P--C0ntrolled Intersectlons

Page 20-11 Version 6.0
 Highway Capadty Manual: A Guide for Multimodal Mobi/ity Analysis

1. INTRODUCTION

Two-way STOP-controlled(TWSC)intersections are common in the United VOlUME 3: INTERRUPTED FLOW

16. Urban 5treet Facilities
States. One typical configuration is a four-Ieg intersection in which one street- 17. Urban Street Reliability and
the major street-is uncontroll€'d, and the other street-the minor sfreet-is ATOM
18. Urban 5treet Segments
controlled by STOPsigns. The other typical configuration is a three-Ieg 19. ~9nalized Intersections
intersection in which the single minor-street approach (Le., the stem of the T 20. TWSC Intersections
21. AWSC Intersections
configuration) is controlled by a STOPsignoMinor-street approachcs can be public 22. Roundabouts
streets or private driveways. This chapter presents concepts and procedures for 23. Ramp Terminals and Altemative
Intersections
analyzing these types of intersections. Chapter 9 provides a glossary and list of 24. Off-Street Pedestrian and BiC)'cle
symbols, including those used for TWSC intersections. Fadlities

Capacity analysis of TWSC intersections requires a dear description and Th~1eg intersections in
which the stem of the T is
understanding of the interaction between travelers on thc minar (i.e., STOP- controlled by il STOP sign are
controlled) approach with travelers on the majar street. Both analytical and considered a standard type of
n¥.5C intersection.
regression models have been developed to describe this interaction. Procedures
described in this chapter rely primarily on field measurements of TWSC
performance in the United States (1) that have been applied to a gap acceptance
model developed and refined in Cermany (2).

CHAPTER ORGANIZAnON
This chapter is organized into the following sections:
• Section 1 (this section) introduces the chaptee.
• Section 2 describes the basic concepts of the TWSC procedure. Most
notably, the concept of gap acceptance-which is the basis of TWSC
intersection operations-is described. Performance measures and level-of-
service (LOS) criteria are also discussed.
• Section 3 provides the details of the TWSC interscction analysis proccdure
for the motorized vehide mode, including required input data and
detailed computational steps.
• Section 4 extends the motorized vehide mode procedure to account for
the eHects oi pcdcstrians 00 capadty.
• Section 5 presents a procedure for analyzing pedestrian operations at a
TWSC intersection, induding required data and computational steps.
• Section 6 qualitatively discusses bicydc operatioos at a TWSC interscction
and difects the reader to related research.
• Section 7 describes example problems included in Volume 4, suggests
applications far alternative tools, and provides guidance on interpreting
analysis results.

Chapter lOfTwo-Way SToP-Controlled Intersectioos Introduction

Version 6.0 Page 20-1
Highway Capacity Manual: A Guide tor Multimodal Mobility Analysis

RELATED HCM CONTENT

Other HCM content related to this chapter ineludes the following:
• Chapter 4, Traffic Opcrations and Capacity Concepts, introduces concepts
of traffic flow and capacity that apply to nvsc intersections, ineluding
peak hour factor, gap acceptance, and control delay.
• Chapter 5, Quality and Level.of.Service Concepts, provides an overview
of the LOS concept used throughout the HCM.
• Chapter 32, STOp.Controlled Intersections: Supplemental, provides
example problems demonstrating the TWSC methodoJogy.
• Case Study L U.S. 95 Corridor, and Case Study 5, Museum Road, in the
HCM Applications Cuide in Volume 4, demonstrate how this chapter's
methods can be applied to the evaluation of an actual nvsc intersection.
• Section M, STOP-Controlled Intersections, in the Plamring alrd Preliminary
Engineering Applicalions Cuide lo the HCM in Volume 4, provides guidance
on analyzing nvsc intersections in the context of a planning study.

Introduction Chapter lOfTwcrWay STOP.Qlntrolled Intersecbons

Page l()"l Vet3ion 6.0
 Highway Capacity Manual: A Guide far Multimadal Mability Analysis

2. CONCEPTS

nvsc intersections are unsignalized interseetions at which drivers on the

major street have priority over drivers on the minor-street approaehes. Minar-
street drivers must stop before entering the interscction. Left-tuming drivers
from the major street must yield to oncoming major-street through or right-
tuming traffie, but they are not required to stop in the absence of oncoming
traffie.
The methodologies presented rely on the required input data Usted in
$ection 3 to compute the polential capacity of eaeh minor movement, which is
ultimately adjusted, if appropriate, to compute a movemenl capacily for eaeh
movement. The movement capacity can be used to estimate the control delay by
movement, by approach, and for the intersection as a whole. Qucue Icngths can
also be estimated once movement eapadties are determined.
At TIVSC intersections, drivers on the SmP-controlled approaches are !he capacity of the controlled
legs is based primanly on three
required to sclcct gaps in the major-street f10win order to executc ceossing or factors: the dlstribu60n of gaps
turning maneuvers. In the presenee of a queue, each driver on the controlled In the majar stream, driver
judgment in setecting the I}iIPS,
approach must also spend time moving to the front-of-queuc position and ami the foIlow.up headways
prepare to evaluate gaps in the major-street flow. Thus, the eapacity of the required by each driver in a
q_.
controlled legs is based primarily on thece factors: the distribution of gaps in the
major-street traffic stream, driver judgment in selecting gaps through which to
execute the desired maneuvers, and the follow-up headways required by each
driver in a queue.
When traffic signals are
The basic eapacity model assumes gaps in the eonflicting movcments are
present on the majar street
randomly distributed. When traffic signals are present on the majar street upstream of the subject
upstream of the subjeet intcrsection, flows may not be random but willlikely intersedion, f!ows fT1<3y not be
random but wlfl fikely have
have sorne platoon structure. svme pfatoon stnJeture.

For the analysis of the motorized vehicle mode, the methodology addrcsscs
special cirrumstances that may exist at TWSC interseetioflS, including the
following:
• Two-stage gap acceptanee,
• Approaches with sharcd lanes,
• The presenee of upstream tralfie signals, and
• Flared approaches for minor-street right-turning vehicles.

INTERSECTION ANALYSIS BOUNOARIES ANO TRAVEL MOOES

The intersection boundaries for a TIVSC intcrscction analysis are assumed to
be those of an isolated intersection (Le., not affeeted by upstream oc downstream
intersections), with the exception of TWSC intersections that may be affccted by
vehicle platoons from upstream signals. This chapter presents mcthodologies to
assess TIVSCintersections for both pedestrians and motor vehicles. A discussion
of how the procedurcs for motor vehicles could potentially apply to an analysis
of bicycle movemcnts is also provided.

Chapter 20fTwrrWay STQl>oControlled Intersections Concepts

Version 6.0 Page 20-3
Hjghway Capacity Manual: A Guide for Multimodal Mobility Analysjs

GAP ACCEPTANCE THEORY

Gap acceptance models begin with the recognition that nvsc intcrscctions
give no positive indication or control to the driver on the minor street as to when
it is appropriate to leave the stop line and enter the major strcet. The driver must
determine when a gap on the major street is large enough to permit entry and
when to enter on the basis of the relative priority of the competing movements.
This dedsion-making process has becn formalized analytically into what is
commonly known as gap acceptance theory. Gap acceptance theory ineludes
three basic elements: the size and distribution (availability) of gaps on the major
street, the usefulness of these gaps to the minor-street drivers, and the relative
priority of the various movements at the interscction.

Availability of Gaps
The first element in gap acceptance thcory is the proportion of gaps of a
particular size on the major street offered to the driver entering from a minor
movement, as well as the pattern of vehiele arrival times. Thc distribution of
gaps between the vehieles in the different streams has a major effect on the
performance of the interscction.

Usefulness of Gaps
The second clement is the extent to which drivers find gaps of a particular
size useful when they attempt to enter the intersection. lt is general1y assumed in
gap acceptance theory that drivers are both consistent and homogeneous. This
assumption is not entirely correcto Studies have demonstratcd not only that
drivers have different gap acceptance thresholds but that the gap acceptance
threshold of an individual driver often changes over time (3). In this manual, the
critical headways and follow-up headways are considered representative of a
statistical average of the driver population in the United States.

Relative Priority of Various Movements at the Intersection

The third element in gap acceptance theory conceros the ranking of each
movement in a priority hierarchy. Typically, gap acceptance processes assume
drivers on the major street are unaffected by the minor movements. 1£this
assumption is not the case at a given intersection, the gap acceptance process has
to be modified.
In using thc nvsc intersection methodology, the priority of right-of-way
given to cach movement must be identified. Some movements have absolute
priority; other movements must yield to higher-order movements. Movements
can be categorized by right-of-way priority as follows:
Pedestrian movements • Movements of Rank 1 inelude through traffic on the major street, right-
crossing the /TIiljar street are
assumed ro be Rank 2 for Me tuming traffic from the major strcct, and pedestrian movements crossing
automobiJe analysis procedure. the minor street.
The effect of Rank 1vehides
yielding ro pedestrians ís • Movements of Rank 2 (subordinate to Rank 1) inelude left-turning and U-
included in the pedesrrian
analysis procedure. tuming traffie from the major street, right-tuming traffie onto the major
street, and pedestrian movements crossing thc major street (assumed for
this procedure).

Concepts Qlapter 20{Two-Way STOP-Controlledlntersections

Page 20-4 V~6.0
 Highway Capacity Manual; A Guide for Multimodal Mobility Analysis

• Movements of Rank 3 (subordinate to Ranks 1 and 2) inelude through

traffic on the minor street (in the case of a four-Ieg interscction) and left-
turning traffic from the minor street (in the case of a T-intersection).
• Movements of Rank 4 (subordinate to aH others) inelude left-turning !he minor-street leff-tum
movement is assigned Rank J
traffic from the minor street. Rank 4 movements occur only at four-Ieg priOnty at a T-intersedion and
interscctions. Rank 4 ptiOtity at a four-Ieg
Intersedion.
Exhibit 20-1 shows the assumed numbering of movements at both T- and
four-leg intersections.

Four-Leg Intersection T-Intersection Exhibit 20-1

Vehicular and Pedestrian

+L
Movements at a TWSC

~)jL__
Intersection

Salid arrows indicare vehicufar

~ 16 ~ "'-6 ~ ~ ---5
, , • S : : movements; dashed arrows
:13 : ,,-4 ,13 , ....-4 indicate pedestn'an

iu~
.-..
1-'"
'
:-
"
-:
'4"
~!.Y lU~ -
,
:,
14:
'~.,•
--;- -=. - 'U
movements.

I.----Trr~~.-----\y~
2 .: 14: 2 lo: ,
, , 3' ,

1 1
17891 1791

As an cxamplc of thc application of right-of-way priority, assume the

situation of a left-turning vehiele on the major street and a through vchielc from
thc minor strcct waHing to cross the major traffic stream. The first available gap
oi acccptable size would be takcn by thc left-turning vehiele. The minor-street
through vehiele must wait for the second available gap. In aggregate terms, a
large numbcr of such lcft-turning vchieles could use up so many of the available
gaps that minor-strcct through vehielcs would be severely impeded or unable to
make safe crossing movements.

Critical Headway and Follow-Up Headway

CriDca! headway defined.
Critica! headway te is defined as the minimum time interval in the major-street
traffic stream that allows intersection entry for one minor-street vehiele (4). Thus,
the driver's critica! headway is the minimum headway that would be acceptable.
A particular driver would rcjcct headways less than the critical headway and
would accept headways greater than or equal to the critical headway. Critical
headway can be estimated on thc basis of observations of the largest rejected and
smallest accepted headway for a given intersection.
The time between the departure of one vehiele from the minor street and the
departure of the next vehiele using the same major-strect hcadway, under a
condition of continuous queuing on the minor street, is caBed the follow-up Foflow-up headway define(/.
lJeadway tI' Thus, ti is the headway that defines the saturation flo\\' rate for the
approach if therc were no conflicting vehicles on movements of higher rank.

Chapter 20/Two-Way STOP-(ontrolled Intersections Concepts

Version 6.0 Page 20-5
Highway Capacity Manual: A Guide for Mu/tjmodol Mobility Analysis

LEVEL-OF-SERVICE CRITERlA
LOS is not defined for the
LOS for a nvsc intersection is determined by the computed or measured
major-street approaches or for
the overall intersa:tion as control delay. For motor vehicles, LOS is detcrmined for each minar-street
major-street /hrough vehides
movemcnt (or shared movement), as well as the majar-street left turns, by using
are assumed to experience no
delay. the criteria given in Exhibit 20-2. LOS is not defined for the intersection as a
whole or for major-street approaches far three primary reasons: (a) major-street
through vehicles are assumed to experience zero delay; (b) the dispropartianate
number of major-street through vehicles at a typical TWSC interscction skews
the weighted average of aH movements, resulting in a very low overall average
delay for all vehicles; and (e) the resulting low delay can mask LOS deficiencies
for minar movements. As Exhibit 20-2 notes, LOS F is assigned to a movement if
its volume-to-capacity ratio exceeds 1.0, regardless of the control delay.
The LOS crHeria for TWSC intersections differ somewhat from the criteria
used in Chapter 19 for signalized intersectians, primarily because user
perceptions differ amang transportation fadlity types. The expectation is that a
signalized intersection is designed to carry higher traffic volumes and will
present greater delay than an unsignalized intersection. Unsignalized
intersections are also assodated with more uncertainty for users, as delays are
less predictable than they are at signals.

Exhibit20-2
lOS Criteria: Motorized
Vehide Mode
~=~co:"~;t~ro
.~_I~",~e~I,:y~=~==;JL~O~Stb~y~Y~O~r~"~m~e~-t~.~-ca~P~'
Ha
fS1vehl v/c
A
v/c
:S1.0
F
~C~ity~.~.~ti~.====
> 1.0

>lo-lS B F
>15-25 e F
>25-35 o F
>35-50 E F
>50 F F
Note: lhe LOS criterí¡¡ ¡¡pply to eacIll¡me 00 ¡¡ given approach aOO10 eiid1 apprwch on the mlnor street. lOS is
not calaJlated for major-stTeet approadles ()( for the intersectiorl as a whole.

Pedestrian LOS at nvsc intersections is dcfined far pedestrians crossing a

traffic stream nat controlled by a STOPsign; it also applies to midblock pedestrian
crossings. LOS criteria for pedestrians are given in Exhibit 20-3.

Exhibit20-3
LOSOitetia: PedestrianMode
Control Delay
LOS (s{p) Cornrnents
A 0-5 Usuallyno conflictingtraffic
B 5-10 OCcasionaltysornedelaydue to conflictingtraffic
e 10-20 Delaynoticeableto pedestrians,but not inconveniencing
o 20-30 [)elaynoticeableand irritating, increasedlikelihoodof risk taking
E 30-45 Delayapproachestolerancelevel,risk-takingbehaviorIikely
F >45 DelayexceedstoIerancelevel, high likeJihoodof pedestrianrísktaking
Note: control delay may be interpreted as secoods per pedestrian group if groups o/ pedestrians were counted as
opposed to indiVidual pedestrians.

LOS F far pedestrians occurs when there are not enough gaps oí suitable size
to allow waiting pedestrians to cross through traffic on the major street safely.
This situation is typically evident from extremely long control delays. The
method is based on a constant critical headway.ln the Held, howevcr, LOS F may
also appear in the form of crossing pedestrians selecting smaller-than-usual gaps.
In such cases, safety could be a conccm that warrants further study.

Coocepts Chapter20fTwo-WaySTop-Controlled
Intersections
Page20-6 Vmkln 6.0
 Hjghway Capacity Manual: A Guide for Multimodal Mobility Analysis

3. MOTORIZED VEHICLE CORE METHODOLOGY

SCOPE OF THE METHODOLOGY

The version of the TWSC interscction analysis procedure presented in this
section is primarily based on studies conducted by National Cooperative
Highway Research Program Project 3-46 (1).

Spatial and Temporallimits

This methodology assumes the TWSC intersection under investigation is
isotated, with the exception of a 1VVSCintersection that may be affected by
vehide platoons from upstream signals. When interaction effects (e.g., queue
spillback, demand starvation) are likely between the subject nvsc intersection
and other intersections, the use of alternative to01s may result in a more accurate
analysis. Analysis boundaries may also indude different demand scenarios
related to the time of day or to different development scenarios that produce
various demand flow rates.
The recommended length oí the analysis period is the HCM standard of 15
min (ahhough longer periods can be examined).

Performance Measures
This method produces the following performance measures:
• Volume-to-capacity ratio,
• Control deJay,
• LOS based on control delay, and
• 95th percentile queue length.

limitations of the Methodology

The methodologies in this chapter apply to TWSC intersections with up to
threc through lanes (either shared or exclusive) on the major.street approaches
and up to three l,lOeson the minof-street approaches (with no more than one
exclusiVe lane fOfeach movement on the minor-street approach). Effects from
other interscctions are accounted fOfonly in situations in which a nvsc
intersection is located on an urban strl.'Ctscgment between coordinated
signalized intersections. In this situation, the intersection can be analyzed by
using the procedures in Chapter 18, Urban 5trL'£'tScgments. The methodologies
do not apply to nv5C intersections with more than four approaches or more
than one STOP-controlledapproach on each sirle of the major street.
The methodologies do not include a detailed method for estimating delay at Wlth dpPrOPrk1te c1langes in
the values of aiticQ/ headway
YIELD-controlledintersections; however, with appropriate changes in the values and foIlow-up headway, the
anafyst could apply the 7WSC
of key parameters (e.g., critical headway and follow-up headway), the analyst
method lO Ylao-controlled
could apply the nvsc method to YIELD.controlledintersections. Intersedions.
AH the rnethods are for steady statc conditions (Le., the demand and capadty
conditions are constant during the analysis period); the methods are not
designed to evaluate how fast or how often the facility transitions from one

Chapter 20{Two-Way STOP-Controlled Intersections Motoózed Vehide Core Methodology

Ver.sk1n 6.0 Page 20-7
Highway Capacity Manual: A Guide tor Multimodal Mobility Analysis

demand or capacity state to another. Analysts interested in that kind of

information should consider applying alternative tools, as discussed below.

Alternative Tool Considerations

Strengths of the HCM Procedure
This chapter offers a set of comprehensive procedures for analyzing the
performance of an intersection under two-way STOPcontrol. Simulation-based
tools offer a more detailed treatment of the arrival and departure of vchicles and
their interaction with the roadway and the control system, but for most purposes
the HCM procedure produces an acceptable approximation.
The HCM procedure offers the advantage of a deterministic evaluation of a
nvsc intcrscction, the results of which have been accepted by a broad
consensus of international experts. The HCM procedure also considers advanced
concepts such as two-stage gap acceptance and flared approaches based on
empirical evidence of their effects.

Umitations of the HCM Procedures That Might Be Addressed by Alternative Tools

The identified limitations for this chapter are shown in Exhibit 20-4, along
with the potential for improved treatment by alterna tive tools.

Exhibit 20.4 limitation Potential for Improved Treatment by Altemative Tools

limitations of tlle HCM TWSC Effects of upstream Simulation tools can inelude an unsignalized intersection explicitly
Intersection Motorized Vehicle intersections within a signalized arterial or network.
Procedure
YIELD-tontrolled intersection Treated explicitly by sorne tools. Can be approximated by varying
operations tlle gap acceptance parameters.
Non-steady-state conditions Most altemative tools provide for multiperiod variation of demand
for demand and capacity and, in sorne cases, capacity.
Atypical intersection
configurations, such as more Sorne alternative tools can be custornized te model the unique
than four Iegs or STOPcontrol configuration of these types of intersections.
on all but one leg

Most analyses for isolated unsignalized intersections are intended to

determine whether TWSC is a viable control alternative. Analyses of this typc are
handled adequately by the procedures describcd in this chapter. The most
!he most rommon applicatiOn common application of alternative tools for nvsc analysis involves an
of altemative tooIs for TWSC
anafysis involv'es an unsignalized intersection located between coordinatcd signalizcd interscctions.
unsignalized intersedion Most intersections between coordinated signalized intcrsections operate under
Iocated t>etween a>ordlfliJted
signafized intersections. TWSC. These intersections tend to be ignored in the analysis of the system
because their eHect on the system operation is mini mal. Occasionally, it is
necessary to examine a nvsc interscction as a part of thc arterial system.
Although the procedures in this chapter provide a method for approximating the
operation of a TWSC intersection with an upstream signa!. the operation of such
an intersection is arguably best handled by including it in a complete simulation
of the full arterial system. For example, queuc backup from a downstream signal
that blocks entry from the cross street for a portion of the cycle is not treated
explicitly by the procedures contained in this chapter.

Motorized Vehiele Cofe Methodology Chapter 20/Tw()-Way STOP-Controlled lntersections

Page 20-8 Version 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Another potential application for alternative too1s is modeling intcrscctions

with more than four legs or with control configurations other than the typical
priority control, such as 5TOl'control on aHbut one leg. The operanon of these
types of intersections has not bccn adequately researched, and no analytical
method has been developed to model their operation. It may be possible to use
an alternative tool to model these configurations provided the priorities betwcen
movemcnts can be customized to match field operations.

Development of HCM-Compatible Performance Measures Using Alternative Tools

Control delay, the performance measure that determines LOS for nvsc
intersections, is defined as that portion of the dela}' caused by a control device-
in this case, a 5TOl'signo Most simulation tools do not produce explicit estima tes
of control delay.
The best way to determine control delay at a STOI'sign from simulation is to
perform simulation runs with and \••.. ithout the control device in place. The
segment delays reported with no control rcpresent the delays due to geometrics
and interaction bctween vehic1cs.The additional delay reported in the TUnwith
the control devicc in place is, by dcfinition, the control delay.
Chapter 7, Interpreting HCM and Altcmative Tool Rcsults, discusses
performance measures from various tools in more detail, and Chapter 36,
Concepts: Supplemcntal, provides recommendations on how individual vchicle
trajectories should be interpreted to produce specific performance measures. Of
particular interest to TWSC opcration is the definition of a queued state and the
development of queue delay from that definition. For altemative tools that Delay and LOS should be
conform to thc queue delay definitions and computations presented in this estimated only by using
aftemative tooIs that confonn
manual, qucuc dcla}' will providc the bcst estima te of control delay for TIVSC to the defimtions ami
intersections. Delay and LOS should not be estimated by using altemativc tools computations o, queue delay
presented 117 thlS manual.
that do not conform to thesc dcfinitions and computations.

Conceptual Differences Between the HCM and Simulation Modeling That Preclude
Direct Comparison of Results
Deterministic tools and simulation tools both model TI\lSC opcrations as a
gap acceptance process that follows the rules of the road to determine the right-
of-way hicrarchy. To this extent, both types of tools use the same conceptual
framework. Dctcrministic tools such as the HCM base their estimates oí capacity
and delay on expectcd values computed from analytical formulations that have
been mathematically derived. Simulation tools, in contrast, take a more
microscopic view, treating cach vehicle as an independent object that is subject to
the rules of the road as well as interaction with other vehicles. Differences in the
treatment of randomness also exist, as explained in the guidance provided in
Chapter 19, Signalized Intersections.
When the opposing movcment volumes are very high, there is minimal
opportunity for thc SToP-controlled movements to accept gaps, and these
movements often have Iittle or no capacity. Simulation tends to produce slightly
higher capacitics under these conditions bccause of a tool-spccific overriding
logic that limits the amount of time any driver is willing to wait far a gap.

Chapter 20/Two-Way STOP-Controlled Intersections Motorized Vehide Core Methodology

Version 6.0 Page 20-9
Highway Capacjty Manual: A Guide for Multimodal Mobilify Analysis

In general, the simulaban results far a specific lVVSCintersection problem

should be close to the results obtained from the procedures in this chapter. Some
differences may, however, be expected among the analysis tools.

Adjustment of Simulation Parameters to the HCM Parameters

Critical headways and follow-up headways are common to both
deterministic and simulation models. It is therefore desirable that similar values
be used for these parameters.

Sample Calculations Illustrating Altemative Tool Applications

An example of the most common application for lVVSCsimulation,
unsignalized intersections within a signalized arterial system, is presented in
Chapter 29, Urban Street Facilities: Supplemental. An additional example
involving blockage of a cross-street approach with STOPcontrol by a queue from
a nearby diamond interchange is presented in Chapter 34, Interchange Ramp
Terminals: Supplemental.

REQUIRED INPUT DATA ANO SOURCES

Exhibit 20-5lists the information necessary to apply the motorized vehicle
methodology and suggests potential sources for obtaining these data. It also
suggests default values for use when intersection-specific information is not
available.

Exhibit 20-5 Suggested

Required Input Data, Potential Required Data and Units Potential Data Source(s) Default Value
Data Sources, and Default GeometTic Data
Values for TWSC Motorized
Number and configuratkm of lanes of each
Vehide Analysis Design plans, road inventory Must be provided
approach
Approach grades Design plans, road inventory 0%
Special geometrk: factors such as
• Unique channelization aspects
• Existence of a two-way left.tum lane or
raised or striped median storage (or both) Design plans, road inventory Must be provlded
• Existence of f1ared approaches on the
minor street
• Existence ef upstream signals
DemandData
Hourly tuming-movement demand volume
(veh/h) ANO peak hour factor
OR Reld data, modeling Must be provided
Hourly tuming-movement demand f10w rate
(veh/h)
Analysis period length (min) Set by analyst 15 min (0.25 h)
Peak hour factor (decimal) Fleld data 0.92
Heavy.vehide percentage (%) Fleld data 3%
saturation f10w rate fer major-sl:reet t:hrough
movement (fer analysis of shared or short Reld data 1,sao veh/h
major-street, left-tum lanes)
saturation f10w rate for maJor-street, right-tum
movement (for analysis of shared or short Field data 1,500 veh/h
major.street, left.turn lanes)

Motorized Vehicle Core Methodology Chapter 20/Two-Way SroP-Controlled Intersections

Page 20-10 Verskrn 6.0
 Highway Capacity Manual: A Gujde far Mulfimodal Mobj/ity Analysis

A comprehensive presentation of potential default values for interrupted

flow facilities is provided elsewhere (5), with specific recommendations
surnmarized in its Chapter 3, Recommended Default Values. These defaults
cover the key characteristics of peak hour factor and percentage oí heavy
vehicles. Recornmendations are based on geographic region, population, and
time of day. AHgeneral default values for interrupted.flow facilities may be
applied to the analysis of TWSC intersections in the absence of Held data or
projected conditions.

COMPUTATIONAL STEPS
The TWSC intersection methodology for the motorized vehicle mode is
applied through a series oí steps that require input data related to movernent
flow information and geometric conditions, prioritization of movements,
computation of potential capacities, incorporation of adjustments to compute
movement capacitics, and estimation of control delays and queue lengths. These
steps are ilIustrated in Exhibit 20-6.

Exhibit 20-6
I Step 1: Determine and label Movement Priorities
I TWSC Intersection
Methodology
••
I Step 2: Convert Movement Demand Volumes to Aow Rates
I
••
I Step 3: Determine Conflicting Flow Rates I
••
I Step 4: Determine Critica! Headways and Follow.Up Headways I
~
No
Coordinated Upstream
Signals Present?

y" ~
5tep 5b: Compute Potential
Step 5a: Compute Chapter 18 inputs
Cdpadties Adjusting fof <-
Potential Cdpacities (P/J,x)
Effects of Upstream Signals

¡ ¡
Step 6: Compute Rank 1 Movement I ¡
Cdpacities I Step 10: Final Cdpacity AdJustments I
¡
Step 7: Compute Rank 2 Movement 1 1 Step 11: Compute Movement Control
capacities Delay

~ ~
Step 8: Compute Rank 3 Movement Step 12: Compute Approach and
Cdpacities Intersection Control Delay

~ ~
Step 9: Compute Rank 4 Movement Step 13: Compute 95th Percentile
Cdpacities Queue Lengths

Chapter 20fTwo-Way STOP-eontrolled Intersections Motorized Vehicle Cofe Methodology

Version 6.0 Page 20.11
Highway Capacity Manual: A Guide lar Multimadal Mability Analysis

Step 1: Determine and Label Movement Priorities

The priority for each movement at a TWSC intersection must be identified to
designa te the appropriate rank of each movement for tuture steps in the analysis
process. This step's process also identifies the sequence in which the analyst will
complete the capacity computations. Because the methodology is based on
prioritized use of gaps by vehicles at a TWSC intersection, the subscquent
computations must be made in a precise order. The computational sequence is
the same as the priority of gap use, and movements are considered in the
following order:
1. Left turns from the major street,
2. Right turns from the minor strcct,
3. U-turns from the majar street,
4. Through movements from the minar street, and
5. Left turos from the minar strect.

Step 2: Convert Movement Demand Volumes to Flow Rates

For analysis of existing conditions wheo the peak 1S-min period can be
mcasurcd in the Held, the volumes for the peak lS-min period are converted to a
peak 1S-mio demand £low rate by multiplying the peak 1S-min volumes by four.
For analysis of projected conditions or when lS-min data are not available,
hourly demand volumes for each movcment are converted to peak lS-min
demand tlow rates in vehicles per hour, as shown in Equation 20-1, through use
of the peak hour factor for the intersection.
Vi
Equation 20-1 v.---
¡ - PHF

where
Vi demand flow rate for movement j (veh/h),
VI demand volume for movement j (veh/h), and
If PHF is used, a single PHF = peak hour factor for the intersection.
intersectionwide PHF should be
used rather tllan movement- If peak hour factors are used, a single peak hour factor for the entire
speaflC or approach-speaflC
PHFs. If individuaf approaches
intersection is generally preferred to decrease the likelihood of creating demand
or movements peak at scenarios with contlicting volumes that are disproportionate to the actual
different times, a series of 15-
min analysis pericds that
volumes during the 1S-min analysis period.lf peak hour factors for each
encompasses the peaking individual approach or movement are used, they are likely to gcnerate demand
shoufd be considered.
volumes from one IS-min period that are in apparent conflict with demand
volumes from another 1S-min period, but in reality these peak volumes do not
T11euse of a peak 15-min occur at the same time. Furthermore, to determine individual approach oc
traffic cvunt muftipfied by tour
is preferred for existing movemeot peak hour factors, actual lS-mio count data are likely available,
condltions when trafflC cvunts permitting the determination of actual lS-min demand and avoiding the nccd to
are avaiJable. T11euse of a 1-h
dem3nd voIume divided by a use a peak hour factor. lf individual approaches or movements are known to
peak hour factor is preferred have substantially different peaking characteristics or peak during different 15-
w;ff¡ projected voIumes or with
projeded voIumes that have mio periods within thc hour, a series of 15-min analysis periods that
been added to current encompasses the peaking should be considered instead of a single analysis
voIumes.
pcriod using a single peak hour factor for thc intersection.

Motorized Vehicle Core Methodology Chaptef 20{Two-Way STOP-<:Ontrolled lntersections

Page 20-12 Version 6.0
 Highwoy Copocity Manual: A Gu;de lor Multimodal MobWty Analysis

5tep 3: Determine Conflicting Flow Rates

Each movement at a nvsc intersection faces a different set of conflicts that
is directly related to the nature of thc subject movement. The following
subsections describe thc set of conflicts faong each minor movement (Rank 2
through Rank 4) at a TWSC intersection, The exhibits and equations illustrate the
computation of the parameter VCy the conflicting flow rate for movement x-that
is, the total flow rate (in vehicles per hour) that conflicts with movement x.
Pedestrians may also conflict with vehicular movements. Pedestrian flow
rates (defined as v.' with x noting the leg of the interscction being crossed)
should be included as part of the conflicting flow rates. Pedestrian f10wsare
included because they define the beginning or ending of a gap that may be uscd
by a minor-street movemcnt. Although this method recognizes sorne
peculiarities associated with pedestrian movements, it takes a uniform approach
to vehicular and pedestrian movements.

Major-Street Left. Turn Movements: Rank 2, Movements 1 and 4

Exhibit 20.7 illustrates the conflicting movemcnts and Equation 20-2 and
Equation 20-3 compute the conflicting £low rates encountered by major-street
ldt-tuming drivers. Thc ldt-tum movernent from the major street conflicts with
the total opposing through and right-tum flow, because the left-turning vchicles
must cross the opposing through movement and be in conflict with the right-
turning vehicles. Thc method does not differcntiate hetween crossing and
merging conflicts. Left-tuming vchicles from the majar street and the opposing
righHurning vehicles from the major street are considered to merge, regardless
of the number of lanes provided in the exit roadway.

Exhibit 20-7
.-_._----~
" , , , ••..
---,.,0 IIIustration of Conflicting
, , Movements for Major-5treet
left-Tum Movements
,,-_.
CD
.--------~
"
V"l = + V6 + v16
Vs Equation 20-2

V',4 = V2 + V3 + V1S Equation 20-]

If the major-street right tum is separated by a triangular istand and has to

comply with a YIELDor STOP sign, then the v6 and v3 terms in Equation 20-2 and
Equation 20-3, respectively, may bc assumed to be zero.

Minor-Street Right.Turn Movements: Rank 2, Movements 9 and 12

Exhibit 20-8 iIIustrates the conflicting movements encountered by minor-
street right-turning drivers. The right-turn movement from the minor street is
assumed to be in conflict with on1)' a portian oí the major-street through
movement when more than one major-street Iane is present. Also, one-half of
cach right-turn movemcnt from the majar street is considered to conflict with the
minor-street right-turn movement, as sorne of these tums tend to inhibit the
subjeet movement. Beeause righHurning \'Chicles from the minor street
commonly merge into gaps in the right-hand lane of the stream into which they

Chapter 20fTwo-Way STOP-Controlled Intersections Motorized Vehicle Core Methodology

Ver>ion 6.0 Page 20-13
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

tum, they typically do not require a gap across alllanes of the conflicting stream
(trns situation may not be true for sorne trucks and vans with long wheelbases
that encroach on more than one lane in making their turos). Furthermore, a gap
in the overall major-street traffic couId be used simultaneously by another
vehide, such as a majar-street left-turning vehide. Exhibit 2G-8does not indude
vehides making major-strcct U-turos as conflicting vehides. Although these
conflicts may be observed in practice, they are not assumed to be conflicts in this
methodology .
Exhibit 20-8
•,,,
Illustration of Conflicting
Movements for Minor-Street
Right-Turn Movements 2
,
,
,,:14
J~,
~--------~
,
3
15 T
•,,
~--------~ 13)
,, • s

<i T
,,

Equation 20-4 through Equation 20-9 compute the conflicting flow rates for
minor-street right-tum movements entering a majar street. If the major-street
right tum has its own ¡ane, the corresponding v) or V~term in these equations
may be assumed to be zero. Users may suppIy different lane distributions for the
V2 and Vs terms in the equations for four- and six-Iane major streets when

supported by field data.

Equation 20-4 and Equation 20-5 compute the conflicting flow rates for
minor-street right-tum movements entering two-Iane major streets:
Equation 20-4 VC.9 = V2 + 0.5V3 + V14 + VIS
Equation 20-5 Vc,12 = Vs + 0.5v6 + V13 + V16
Equation 20-6 and Equation 20-7 are used for four-lane majar streets:
Equation 20-6 Vc.9 = + O.sV3 + v14 + VIS
0,SV2
Equation 20-7 VC,12 = O.SVs + O,SV6 + V13 + V16

Equanon 20-8 and Equation 20-9 are used for six-Iane major streets:
Equation 20-8 VC,9 = O,SV2 + O.sV3 + V14 + VIS
Equation 20-9 Vc.12 = O.SVs + O,SV6 + V13 + V16
Major-Street U-Turn Movements: Rank 2, Movements lU and 4U
Exhibit 20-9 illustrates the conflicting movements encountered by majar-
street U-tuming drivers. The U-tum movement from the major street conflicts
with the total opposing through and right-tum flow, similar to the major-street
left-tum movement. Research has found that the presence of minor-street right-
tuming vehicles significantly aHects the capacity of major-street U-toms (6). The
methodology accounts for this eHect in the impedance calculation rather than in
the calculation of conflicting flow. lf a diHerent priority order is desired (e.g.,
minor-street right turos yield to major-street U-toms), the analyst should adjust
the computation procedure accordingly to replicate observed conditions.

Motorized Vehicle Core Methodology Olapter 20/Two-Way SrOP-Controlled Intersections

Page 20-14 Version 6.0
 Highway Capacify Manual: A Guide for Mulfimodal Mobilify Analysis

Exhibit 20-9
IllustratiOn
of Conflicting
, MovementsforMajor'Street
U-TumMovements

Equation 20-10 through Eguation 20-13 compute the con£licting £low rates for
major-street U-rurns. No field data are availabIe for U-tums on major streets wüh
fewer than four Ianes. If a major-street right tum has its own Iane, the
corresponding v30r VI>term in these eguations shouId be assumed to be zero.
Equation 20-10 and Equation 20-11 compute the conflicting £low rates for
major-street U-turns when the majar street has four Ianes:
Ve.1U = + V6
Vs Equiltion 20-10

ve•4u = V2 + V3 Equation 20-11

Eguation 20-12 and Eguation 20-13 compute the eonflicting flow rates for
major-street U-tums on six-Iane major streets:
ve.tU = 0.73vs + 0.73V6 Equation 20-12

Ve.4U = 0.73v2 + 0.73v3 Equation 20-13

Minor-Street Pedestrian Movements: Rank 2, Movements 13 and 14

Minar-street pedestrian movements (those pedestrians crossing the major
street) are in direct conflict with all vehicular movements on the major street
except the right-tum and left-tum movements on the major street approaching
from the far side of the interseetion. The volume of minor-street pedestrians is an
input parametcr in the computation of the eonflicting flow rates for al1 Rank 3
and Rank 4 movements.

Minor-Street Through Movements: Rank 3, Movements 8 and 11

Minor-street through movements have a direet crossing or merging confliet
with a1l movements on the major street except the right tum into the subject
approach. Similar to thc minar-street right-tum movement, one-half of each
right-tum movement from the major street is considcred to conflict with the
minor-street through movement. In addition, fieId researeh (1) has shown that
the cHect of left-turning vehic1cs is approximately twice their actual number.
Drivers executing minar-street through movements may complete their
maneuver in one or two stages. One-stage gap aeeeptanee assumes no median
refuge area is available for minor-street drivers to store in and that the minor-
street drivers will evaluate gaps in both major-strcct direetions simultaneously.
Conversely, thc two-stage gap aeeeptanee seenario assumes a median refuge arca
is available for minor-street drivers. During Stage 1, minor-street drivers evaluate
major-street gaps in the nearside traffie stream (eonflieting traffie from the left);
during Stage 11,minor-street drivers evaluate major-street gaps in the farside
traffie stream (eonflicting traffie from the right). For one-stage erossings, the
eonflieting f10ws for Stage 1and Stage JI are eombined; for two-stage erossings,
the eonflieting flows are considered separately.

Olapter10/Two-Way
5TOP-Controlled
Intersections Motorized
VehicleCoreMethodoIogy
Version 6.0 Page 20-15
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Exhibit 20-10 illustrates the conflicting movements encountered by minor-

street through-movement drivers.

Exhibit 20-10
5tage 1
IIIustration of Conflicbng
Movements for MinQf-$treet lO
+¡
~
j@
Through Movements 1
2 • .--------~
3
"<C--------~
15
16 , 6
• 5
4

@J " C; 4U

5tage II
•....16_---~ +¡ j@
, 6
lO
2
1 ~
•
• 5 3

@1 " C;
4
4U "<C---!? •• ~
Equation 20-14 and Equation 20-15 compute the conflicting flows
encountered by minor-street through-movement drivers during Stage I.1f there
is a right-tum lane on the majar street, the corresponding V3 or V6 term in these
equations may be assumed to be zero.

•
Equation 20-14 Ve,I.8 = 2(v¡ + V1U) + V2 + 0.5V3 + VIS
Equation 20-15 Ve.l.l1 = 2(V4 + V4U) + Vs + 0.5v6 + VI6
Equation 20-16 and Equation 20-17 compute the conflicting flows
encountered by minor-street through-movement drivers during Stage II.1f the
majoNtreet right tum is separated by a triangular island and has to comply with
a YIELDor STOPsign, the corresponding v3 or v6 tenn in these equations may be
assumed to be zero.
Equation 211-16 Vc,1I.8 = 2(V4 + v4u) + Vs + V6 + VI6
Equation 20-17 Ve.II.11 = 2(Vl + V1U) + V2 + V3 + VIS

Minor-Street Left. Tum Movements: Rank 4~Movements 7 and 10

The left-turn movement from the minor strcet is the most difficult maneuver
to execute at a nvsc intersection, and it faces the most complex set of conflicting
movements, which indude al1 major-street movements in addition to the
opposing right-tum and through movements on the minor street. Only one-half
the opposing righHum and through-movement flow rate is induded as
conflicting flow rate beca use both movements are STOP-controlled, which
diminishes their effect 00 left turns. The additional capacity impedaoce effects of
the opposing right-turn and through-movement flow rates are taken ioto account
elsewhere in the procedure.
Similar to minor-street through movemeots, minor-street left-turn

•
movemeots may be completed in one or two stages. Exhibit 20-11 illustrates the
conflicting movements encountered by mioor-street left-turning drivers.

Motorized Vehicle Core Methodology Chapter 20/Two-Way STOP-Controlled Intersections

Page 20-16 VerskJn 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Exhibit 20-11
5tage i
IIIustration of Conflicting
,
1U
•• Movements for Minor-Street
Left- Tum Movements
2
3 <t-------- •.
15
..•-------- .. 16

•
6
S
• 4
4U

Sto" ~ ') 1" ,

1U •,,,
, 2 :14
3 ,
,,
: ~ 6
1] : ~ S T

¡ d' 4
4
4U

Equation 20-18 through Equation 20-23 compute the conflicting flow rates for
minor-strect left.turn movements entering a majar street during 5tage I. If a
right-turn lane exists on the majar street, the corresponding v3 ar Vó term in these
equations may be assumed to be zero.
During 5tage L Equation 20-18 and Equation 20-19 compute the conflicting
flow rates for minor-street left-turn movements entering two-lane major streets:
ve,t.7 = 2V1+ V2 + O.sv] + VIS Equation 20-18

Ve,I,tO = 2V4 + Vs + 0.SV6 + VI6 Equation 20-19

Equation 20-20 and Equation 20-21are used for four-Iane majar streets:
Ve,!'7 = 2(vI + vw) + V2 + 0.SV3 + vts Equation 20-20

ve,uo = 2(V4 + V4U) + Vs + O,SV6 + VI6 Equation 20-21

Equation 20-22 and Equation 20-23 are used for six.lane majar streets:
vel7 = + ViU) + V2 + O.Su] + VIS
2(VI Equation 20-22

Ve,UO = 2(V4 + V4U) + Vs + 0.sV6 + VI6 Equation 20-23

5imilarly, Equation 20-24 though Equation 20-29 compute the conflicting

flow rates for minor-street left.turn movements entering a majar street during
5tage II. lf thc minor-street right turn is scparatcd by a triangular island and has
to comply with a YIElD or STOPsign, the corresponding v9 or Vl2 term in these
equations may be assumed to be zcro.
During 5tage ll, Equation 20-24 and Equation 20-25 compute thc conflicting
flow rates for minor-street ldt-turn movements entering two-lanc majar streets:
Ve,1I,7 = 2V4 + Vs + 0.5v6 + O.SV12 + O.svn + vn Equation 20-24

ve.ll.10 = 2Vl + V2 + O.Sv] + 0.sV9 + O.5vs + vt4 Equation 20-25

Equation 20-26 and Equation 20-27 are used for four-Iane majar strects:
Vell7 = + V4U) + O.svs + O.SVll + VI3
2(V4 Equation 20-26

ve,U,tO = 2(vI + vw) + 0,SV2 + O.Svs + vt4 Equation 20-27

Olapter 20/Two-Way STop-Controlled intersectíons MotOfized Vehíde Cofe Methodology

Vef5km 6.0 Page 20-17
Highway Capacity Manual: A Gu;de for Mu/timoda/ Mob;/ity Analysis

Equation 20.28

Equation 20-29
Equation 20-28 and Equation 20-29 are used for six-lanc major streets:
Ve.U,7

ve,n,to
= 2(v4+ V4U) + O.4vs + 0.5vn + V13
= 2(vt + V1U) + O.4vz + O.5va + va

Step 4: Determine Critical Headways and Follow-Up Headways

•
The critical headways t,~and follow-up headways tJ~ must be determined for
the major-street left turns (v,,! and V',4)' the minor-street right tums (V,.9 and Vd2)'
the major-street U-turns (V,.1U and V',4U)' the minor-street through movements (v,)!
and Vc,u), and the minor-street left turns (v,.? and V,.IO) as they occur at a nvsc
intersection.
To compute the critical headways for each movement, the analyst begins with
the base critical headway given in Exhibit 20-12 and makes movement-specific
adjustments relating to the percentage of heavy vehicles, the grade encountered,
and a three-leg versus four-leg intersection as shown in Equation 20-30.
Equation 20-30 te.x = te.base + te,HVPHV + te.GG - t3,LT

where
t,~ critical headway for movemcnt x (s),
t,."I>.>'" base critical headway from Exhibit 20-12 (s),
t,.HV = adjustment factor for heavy vehicles (1.0 for major streets with one
lane in each direction; 2.0 for major streets with two or three lanes in

•
each direction) (s),
PHV proportion oí heavy vehicles for movement (expressed as a decimal;
e.g., PHV = 0.02 for 2% heavy vehicles),
t,.G adjustment factor for grade for given movement (0.1 for Movements 9
and 12; 0.2 for Movements 7, 8, 10, and 11) (s),
G percentage grade (expressed as an integer; e.g., G = -2 for a 2%
downhill grade), and
túr iS applicable ro Movements tur adjustment factor for intersection geometry (0.7 for minor-street left-
7, 8, 10, and 11.
tum movement at three-leg interscctions; 0.0 otherwise) (s).

Exhibit 20-12 Vehicle Base Critical Headway. t:••••(s)

Base Critical Headways fer Movement Two Lanes Four unes Six Lanes
TWSC intersections left tum from majar
4.1 4.1 5.3
,;treet
U-tum from major
NA 5.6
,;treet
Right tum from minar
6.2 6.' 7.1
,;treet
1 stage: 6.5 1 stage: 65 1 stage: 65b
Throogh traffic on
2 stage, Stage I: 5.5 2 stage, Stage 1: 5.5 2 stage, Stage 1: S.Sb
minor street 2 sta Sta e II: S.Sb
2staeS eII:S.S 2 sta e S II: 55
1 stage: 7.1 1 stage: 7.5 1 stage: 6.4
left tum from minar
2 stage, Stage 1: 6.1 2 stage, Stage 1: 6.5 2 stage, Stage 1: 7.3
_t 2 sta e S II:6.1 2 sta e Sta e II: 65 2 sta e Sta e II: 6.7
Notes: NA '" not available.

•
I Narrow U-tI.Irns have a median nose widtlI <21 ft; wide U-turns have a median nose width <!:21 ft.
b lh;e cautiorl; values estimated.

Motorized Vehide Core Metliodology Olapter 20fTwo-Way STOP-Controlled Inter.;ections

Page 20-18 Version 6.0
 Highway Capacity Manual: A Guide for Multimodal MobUity Analysis

The critical headway data for four- and six-Iane sites aceount Eorthe actual
lane distribution of traffic flows measured at eaeh site. For six-Iane sites, minor-
street leEtturos were commonly observed beginning their movement while
apparentIy eonflicting vehieles in the farside major.street through stream passed.
The values for critica! headway for minor-street through movements at six.lane
streets are estimated, as the movement is not frequently observed in the field.
Similar to the computation of critical headways, the analyst begins the
computation of follow-up headways with the base follow.up headways given in
Exhibit 20-13. The analyst then makes movement-specific adjustments to the base
follow.up headways with information gathered on heavy vehicles and the
geometrics of the major street per the adjustment factors given in Equation 20-31.
tf.x = tl.base + tl.HVPHV Equation 20-31

where
t¡... follow-up headway for movement x (s),
t¡.N"" base follow-up headway from Exhibit 20-13 (s),
f¡;HV adjustment factor for heavy vehicles (0.9 Eor major streets with one
lane in each direction; 1.0 for major streets with two or three lanes in
each direction), and
PHI' proportion oE heavy vehicles for movement (expressed as a decimal;
e.g., PHV = 0.02 Eor 2% heavy vchicles).

Base Follow-UD Headway. tn._ ts) Exhibit 20-13

Vehicle Movement Two Lanes Four Lanes Six Lanes Base Follow-Up Headways for
left tum from major street 2.2 2.2 3.1 TWSC Intersections
U-turn from major street 2.5 (wide).
NA 2.3
3.1 (narrow)'
Right tum from minor street 3.3 3.3 3.9
Through traffie on minor street 4 .• 4.• 4.•
left tum from minor street 3.5 3.5 3.8
Notes: NA = not available .
• N"rrow U'wrns have a mediar¡ nose widtt1 <21 ft; wide U-tums have a median nose width 221 rt.

Valucs from Exhibit 20-12 and Exhibit 20-13 arc based on studies throughout
the United States and are representa tive oE a broad range of conditions. If smaller
values for Irand t¡ are observed, capadt)' will be increased. If larger values for le
and lIare used, capacity wiII be decreased.

5tep S: Compute Potential Capacities

Step Sa: Potential Capacity W,thout Upstream Signal Effects
The potential capadt)' cp•r of a movement is computed according to the gap
acceptance rnodel provided in Equation 20-32 (7).
e-ve ..•.
te..•.
/J,600
e p.x -- v e.x 1 _ t,....
e-lle ..•. /J.600
Equation 20-32

where
cp,r potential capadty of movcment x (veh/h),
ve... conflicting flow rate for movement x (veh/h),

Chapter 20/Two-Way STOP-Controlled Intersections Motorized Vehicle Core Methoclology

VersiOn 6.0 Page 20-19
Highway Capacify Manual: A Gu;de for Multimodal Mobility Analysjs

te..
ti"
critical headway for minor movement
follow-up headway for minor movement
X

For two-stage Rank 3 or Rank 4 movements, the potential capadty is

(s), and
X (s).

computed three times: cp .• assuming one-stage operation. cp,l> for 5tage I, and cr.lLx
for 5tage JI. The conflicting £low definitions for each calculation are as provided
•
in 5tep 4.

Step Sb: Potentiaf Capacity with Upstream Signa! Effects

To evaluate the impact of coordinated upstream signals, the urban street
segments methodology (Chapter 17) is used to estimate the propartion of time
that each Rank 2 or lower movement will be effectively blocked by a platoon. The
proportion of time blocked is denoted by P".v where X is the movement using the
movement conventions provided in Exhibit 20-1.
With these values, the propartion of the analysis period that is blocked far
each minor movement can be computed by using Exhibit 20-14.

Exhibit 20-14 Prooortion Blocked Jor Movement, Do.

Proportion of Analysís Period Two-stage Movements
Blocked fer Each Movement Movement s x One-Sta e Movements Sta el Sta eII
1,1U PO' NA NA
4,4U P" NA NA
7 P" PM PD,I
8 fu PM Pu

•
9 NA NA
P"
10 PD,IO PD,1 PD,4
11 PD,U Pu PM
12 ~.1] NA NA
Note: NA'" not applicab~.

The £low for the unblocked period (no platoons) is determined in this step.
This £low becomes the con£licting £low for the subject movement and is used to
compute the capadty for this movement. The minimurn platooned £low rate ve.,,,,n
is approximately 1,0OON, where N is the number of through Janes per direction
on the major street (8).
The conflicting £low for movement x during the unblocked period is given by
Equation 20-33.
_ {VC,X - l~Vc,minPb.x
ir Vc.x > l.SVc.minPb,X
vc.u,X - 1 Pb,X

O otherwise
where
Ve...... conflicting flow for movement X during the unblocked period (veh/h);
Ve... total conflicting £low for movement x as determined from 5tep 3
(veh/h);
ve•mi" minimum platooned f10w rate (veh/h), assumed to be I,OOON,where N
is the number of through lanes per direction on the major street; and

Motorized Vehíde Core Methodology

Page 20-20
proportion of time thc subjcct movement x is blocked by the major-
street platoon, which is detcrmined from Exhibit 20-14.

O1apter 20{Two-Way $TOP-Controlled Intersections

Version 6.0
•
 Highway Capacity Manual: A Guide for Mulfjmoda/ Mobility Analysis

The potential capacity of the subject movement x, accounting for the effect of
platooning, is given by Equation 20-34 and Equation 20-35.
Equation 20-34
cp,x = (1- Pb.x)Cr,x
e -vc .••..•.
tc..•.
/3.600
er,x =v c.tl,X -------
1 _ e-vc tt ..•.
.••..•. /3.600
Equation 20-35

where
Cr ,% potential capacity of movement x (veh/h),
Pb,% proportion of time that movement x is blocked by a platoon, and
c',% capacity of movement x assuming random flow during the unblocked
periodo
These equations use the same critical headway and follow-up headway
inputs as a normal calculation, but the)' use only the conflicting f10wduring the
unblocked periodo

5teps 6-9: Compute Movement Capacities

For darity, these steps assume pedestrian impcdance effects can be neglected,
and in man)' cases this assumption is reasonable. However, pedestrians can be
accounted for in the analysis of thc motorized vehide mode by replacing these
steps with those provided in Section 4, Extension to thc Motorized Vehide
Mcthodology, which incorporate the effects of pedestrian impedance.

5tep 6: Compute Rank 1 Movement Capacities

Rank 1 major-strect movcments are assumed to be unimpeded by any
movements of lower rank. This rank also implies that major-street movemeots of
Rank 1 are oot expected to incur delay or slowing as they travel through the
nvsc intersection. Empirical observations have shown that such delays do
occasionally occur, and they are accounted for by using adjustments provided
later in this procedure.

5tep 7: Compute Rank 2 Movement Capacities

Movemcnts of Rank 2 (left turns and U-turns from the majar street and right
turos from thc minar street) must yield to conflicting major-street through and
right-turning vehicular movements of Rank 1. Minor-street right turos are
assumed to yield to major-street U-turos, although somctimes the reverse occurs.

Step 7a: Movement Capacity for Major-Street Left-Turn Movements

The movement capacity cm,¡ for Rank 2 major-street left-turo movements
(1 and 4) is equal to its potential capacity 'r,i' as shown in Equation 20-36.
Equation 20-36

Step lb: Movement Capacity for Minor-Street Right-Turn Movements

The movement capacity cm,¡ for Rank 2 minor-street right-turn movemcnts
(9 and 12) is equal to its potential capacity 'r,'" as shown in Equation 20-37.
Equation 20-37
Cm,j = Cp.j

Chapter 20/Two-Way STOJ>-Controlledintersections Motorized Vehicle Core Methodology

VersiOn 6.0 Page 20-21
Highway Capacity Manual: A Guide for MultimodaJ MobiJity Analys;s

Step 7c: Movement Capacity for Major-Street U-Turn Movements

The movement capacity Cm,; far Rank 2 major-street U-tum movements (lU
and 4U) is faund by first computing a eapadty adjustment factor that accounts
for the impeding effeets of higher-ranked movements. Field observations are
mixed in terms of the degree to which major-street U-tum movements yield to
minor-street right-tum movements and vice versa (5). It is assumed that the
presence of minor-street right-tuming vehides will impede U-tuming vehides
from accepting gaps in the major-street traffic stream; therefore, the eapacity of
the U-tum movement is affected by the probability that the minor-strcet right-
tuming traffie will operate in a queue-free sta te. The capacity adjustment faetors
are denoted by hu and ffU for the majar-street U-tum movements IU and 4U,
respectively, and are given by Equation 20-38 and Equation 20-39, respectively.
Equation 20-38 ftU = PO.12 = 1---
v"
Cm,12

Equation 20-39 f4U = PO,9

= 1--
",
Cm.9

whcrc
fW,f4U "" capacity adjustment factor for Rank 2 major-street U-turo movements
1 and 4, rcspeetively;
Po,; probability that conflicting Rank 2 minor-street right-tum movement j
will operate in a queue-free sta te;
vj
c••.¡
flow rate of movement j;
capacity of movement j; and
j '" 9 and 12 (minor-street right-tum movements of Rank 2).
The movement eapacity for major-street U-tum movements is then
computed with Eguation 20-40.
•
Equation 20-40
Cm,jU = Cp,jU X Jiu
In afmost all cases, fTliJjor- where
street Ieft-tuming vehides
share a lane with U-tuming cm•jU movement eapadty for Movements 1U and 4U,
vehicles. If Rank 1 fTliJjor-sireet
U-tum movements are present cp•jU potential capacity far Movements lU and 4U (from 5tep 5), and
ro a signifiG3nt degree, then
Equation 1{}-4} shoukJ be /ISl!d /;u capacity adjustment factor far Movements 1U and 4U.
to compute Me shared-lane
capaciry. Beeause left-tum and U-tum movements are typically made from the same
lane, their shared.lane capacity is eomputed with Eguation 20-41.
Ly Vy
Equation 20-41
CSH = v
Ly~
m,y

where
CSH capacity of the shared lane (veM1),
vy fIow rate of the y movement in the subjeet shared lane (veh/h), and
C""y -= movement capacity of the y movement in the subject shared ¡ane
(veM1).

MotorizedVehicleCoreMethodology Olapter20{Tw(}-Way$TOP-Controlled
Intersections
Page20-22 Versfon6.0
 Highway Capacity Manual: A Guide for Multimodol Mobility Anolysis

Step 7d: Effect of Major.Street Shared Through and Left-Turn Lane

Thc probability that the major-street left-turning traffie will opera te in a
queue-free state is expressed by Equation 20-42.
v.1
PO,j = 1 __ Equation 20-42
Cm,j

where j = 1 and 4 (major.street lett-tum and U-tum movements of Rank 2, using

shared vo1ume and eapaeity as appropriate). Use Equation 20-42 to
compute the pro/Jobility of a
If, howcvcr, a sharcd IcfHum lanc or a short lefHum poeket is present on a queue-free stiJte for Rank 2
mow;menls.
major.street approach (as in Exhibit 20-15), thc analyst aecounts for this
Ir major-street through and
occurrenee by eomputing thc probability that there will be no queue in thc left-mm movements are
major.street shared lane, P~,i' aeeording to Equation 20-43.This probability is shared, use Equation 20-43.

then uscd by the analyst in lieu of Po.; from Equation 20-42.

Exhibit 20.15
Short Left- Tum Pocket on
Major-Street Approach

The methodology implicitly assumes an exclusive lane is provided to allleft-

tuming traffic from the major street. If a lett-tum lane is not provided oc the left-
tum pocket is not long enough to aecommodate all queuing left-tum and U-tum
vehicles, major-street through (and possibly right-tuming) traffic could be
delayed by left-tuming vehicles waiting for an acceptable gap in opposing major-
street through traffic. To aecount for this occurrenee, the factors P~,1 and PO,4 may
be computed aeeording to Equation 20-43 and Equation 20-44 as an indication of
the probability there wil! be no queue in the respective major-street shared or
short lancs (9).

Equation 20-43

Equation 20-44

where
probability of queue-free state for movement j assuming ao exclusive
left-tum lane on the major street (per Equation 20-42);

PO,j probabiJity of queue-free state for movement j assuming a shared 1eft-

tum lane on the major strcct;
Whenj = 1, i1 = 2 and i2 = 3;
j 1 and 4 (major-street left-tuming vehicular movements); wheflj = =
4, i1 5 and i2 = 6.
;1 2 and 5 (major-street through vehicular movcments);
a 3 and 6 (major-street right-tuming vehicular movements);

Chapter 20/Two--Wa'l STOl'-controlled Intersections Motorized Vehicle Core Methodolog'l

V~6.0 Page 20-23
Highway Capocity Manual: A Gu;de for Multimodol Mobility Anolysis

X',I.! combined degree of saturation for the major-street through and right-
tum movements;
Sil saturation flow rate for the major-street through movements (default
assumed to be 1,800veh/h; however, this parameter can be measured
in the field);
S,l saturation flow rate for the major-street right-tum movements (default
assumed to be 1,500veh/h; however, this parameter can be measured
in the field);
Vil major-street through-movement flow rate (veh/h);
V,l major-strt..
'Ct right-tum flow rate (veh/h) (Oif an exclusive right-turn
lane is provided); and
/Ir. = number of vehicles that can be stored in the left-tum pocket (see
Exhibit 20-15).
For the spedal situation of shared ¡anes (/Ir. = O),Equation 20-43 becomes
Equation 20-45 as follows:
l-p .
Equation 20-45 • -1 O,}
PO,i - - 1- X.
1,1+2

where all terms are as previously defined.

By using P~l and PO.t in Iicu of PQl and Po.t (as computed by Equation 20-42),

•
the potential for queucs on a major street with shared or short ¡eft-tum lanes may
be taken into aecount.

Step 8: Compute Rank 3 Movement Capacities

Rank 3 minor-street traffic movements (minor-street through movements at
four-leg intersections and minor-street ¡eft tums at three-leg intersections) must
yield to conflicting Rank 1 and Rank 2 movements. Not aH gaps of aceeptable
length that pass through the interseetion will normally be available for use by
Rank 3 movements, because sorne of these gaps are likely to be used by Rank 2
movements.
If the Rank 3 movement is a two-stage movement, the movement eapacity
for the one-stage movement is eomputed as an input to the two-stage ealrolation.

Step 8a: Rank 3 capacity for One-Stage Movements

For Rank 3 movements, thc magnitude of vehicle impcdancc dcpcnds on the
probability that major-strcct left-tuming vehicles will be waiting for an
aceeptable gap at the same time as vehicles of Rank 3. A higher probability that
this situation will oeror means greater capacity-reducing effects of the major-
street ¡eft-turning traffie on all Rank 3 rnovements.
The movement capacity c",;: for aHRank 3 movements is found by first
computing a capacity adjustment factor that aceounts for the impeding effects of
higher-ranked movements. Thc eapacity adjustment factor is denoted by f< for all
movements k and for aH Rank 3 movements and is given by Equation 20-46.
Equation 20-46
!k = npO.i
j

Motortzed Vehicle Core MethOOology Chapter 20/Two-Way STOP-Controlled Intersections

Page 20-24 VetSiOn 6.0
 Highway Capadty Manual: A Gu;de for Mu/fimodal MobUity Analysis

where n indicates the product of a series of terms, and

PO.¡ probability that conflicting Rank 2 movement j will opcrate in a queue-
free state, and
k = Rank 3 movements.
The movement capadty cm): for Rank 3 minor-street movements is computed
with Equation 20-47.
Equation 20-47
Cm,k = Cp.k X fk

where Cp.k is the potential capadty of Rank 3 minor-street movements, andA is the
capadt)' adjustment factor that accounts for the impcding effects of higher-
ranked movements computed according to Equation 20-46.

Step 8b: Rank 3 C8pacity for Two-Stage Movements

If the Rank 3 movement is a two-stage movcment, the pracedure for
computing the total movement capadt)' for the subjcct movement considering
the two-stage gap acceptance process is as follows. An adjustment factor a and an
intermediate variable y are computed with Equation 20-48 and Equation 20-49,
respectively.
a = 1 - 0.32e-1.3.¡n,;; for nm >O fquation 20-48

y~
Cl - cmx.
Equation 20-49
c[( - VL - cm,x

where
"m number of vehicles that can be stored in the median;
CI movement capadt)' for the Stage 1praccss (vch/h);
cn movement capadty for the Stage 11process (veh/h);
t'L major Idt-1om or U-tum £low rate, either VI + Vw or v4 + v4U (veh/h); and
cm~ capadt)' of subject movement, considering the total conflicting £low
rate for both stages of a two-stage gap acceptance process (from Step 8a).
The total capadty cT for the subject movement, considering the two-stage gap
acceptance process, is computed by using Equation 20-50 and Equation 20-51 and
incorporating the adjustment factors derived fram Equation 20-48 and Equation
20-49.
Fory,¡,l:
a
cr = y nm +1 _
1 [y{ynm - l)(c[( - vd + {y - l)cm.xl Equation 20-50

Fory=l:
Equation 20-51

Step 9: Compute Rank 4 Movement capacities

Rank 4 movements occur only at four-Ieg intersections. Rank 4 movements
(Le., only the minor-street left tums at a four-Ieg intersection) can be impeded by
all higher-ranked movements (Ranks 1, 2, and 3).

Chapter 20fTwo-Way $TOP-Controlled intersectiOfls Motorized Vehicle Core Methodology

Version 6.0 Page 20-25
Highway Capacity Manual: A Guide far Mulfimodal MobiJjty Analysis

Step 9a: Rank 4 G3paCity for One-Stage Movements

The probability that higher-ranked traffie movements will operate in a
queue.free state is central to determining their overall impeding effects 00 the
minor-street left-tum movement. However, not aH these probabilities are
independent of eaeh other. Speciíically, queuing in the major-street left-turning
movement affeets the probability oí a queue-free state in the minor-street
crossing movement. Applying the simple product of these two probabilities will
likely overestimate the impeding effects on the minor-street left-tuming traffic.
Exhibit 20.16 can be used to adjust for the overestimate eaused by the
statistical dependence between queues in streams of Ranks 2 and 3. The
mathematieal representation of this curve is determined with Equation 20-52.

Equation 20-52 p' = 0.6Sp" - ~ + 0.6 C;;p"

p"+3 vf'
where
p' adjustment to the major-street left, minor.street through impedance
factor;
p" (POoi)(POo~);
POol probability of a queue-free state for the eonflicting major-street left-
tuming traffie; and
Po.~ = probability of a queue-free state for the eonflieting minor.street

•
crossing traffic.
When determining p' foc Rank 4, Movement 7, in Equation 20-52,
p" = (Po.!J(Po.t)(PO.l1)' Likewise, when determining p' for Rank 4, Movement 10,
p" = (PO.l)(POA)(PO.s)'

Exhibit 20.16 LO
Adjustment to Impedance
Factors for Major-5treet Left- O., -
1
V
Tum Movement and Minor- : ,
5treet Crossing Movement O.,
I
; /
'/
0.7
, 1/
O.

•• 0.5
V ,

y ,
, :
O.•
!
0.3
/
02
V
O.,
1/
7 I
l.. -
1
0.0
0.0 O, 0.2 0.3 os o.• 0.7 O' o., LO
p'

MotOrized Vehicle Core Methodology Chapter 20(Two-Way STOP-Controlled intersections

Page 20-26 Version6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

The movement capacity C •• ,I for all Rank 4 movements is found by first

computing a capacity adjustment factor that accounts for the impeding effects of
higher-ranked movements. The capacity adjustment factor Ir'! for the Rank 4
minor-street left-turn movement can be computed with Equation 20-53.
{P,I = p' X PO,j Equation 20-53

where
minor-street kft-turn movement of Rank 4 (Movements 7 and 10 in
Exhibit 20-1), and
j = conflicting Rank 2 minor-street right-turn movement (Movements 9
and 12 in Exhibit 20-l).
Finally, the movement capacity for the minor-street lcft-turn movcmcnts of
Rank 4 is determined with Equation 20-54, where Ir,1 is the capacity adjustmcnt
factor that accounts for the impcding cffccts of higher-ranked movements,
cm,/ = Cp,l X {p.l

Step 9b: Rank 4 Capacity for Two-Stage Movements

The procedure for computing the total movement capacity for the subject
movement considering the two-stage gap acceptance process is as follows. An
adjustment factor a and an intermediate variable y are computed with Equation
20-55 and Equation 20-56, respectively,
a = 1 - 0.32e-l.3.[r1,;;" for nm >O Equation 20-55

Cl - cm,x
y=---- Equation 20-56
CII - VL - cm.x
where
11.. number of storage spaces in the median;
el movemcnt capacity for the 5tage l process (veh/h);
Cl! movement capacily for the 5tage JI process (veh/h);
vL majar left-turn or U-turn flow rate, either VI + Vw or v4 + V4U (veh/h); and
c••,x capacity of subject movement, inc1uding the total conflicting f10wrate
for both stages of a two-stage gap acceptance process (from 5tep 9a).
The total capacity Cr for the subject movement considering the two-stage gap
acceptance process is computed by using Equation 20-57 and Equation 20-58 and
incorparating the adjustment factors computed in Equation 20-55 and Equation
20.56.
Fory*l:
a Equation 20-57
cr = y"m+l _ 1{y(y"m - 1)(c[[ - VL) + (y - 1)cm,xl
Fory=l:
Equation 20-58

Olapter 20fTwo-Way STOP-Controlled Intersections Motorized Vehicle Core Methcx:lology

V~ion6.0 Page 20-27
Highway Capacify Manual: A Guide for Multimoda/ Mobi/ify Analysis

Step 10: Final Capacity Adjustments

Step lOa: Shared-Lane Capacity of Minor-Street Approaches
Where two or more movements share the same ¡ane and carmot stop side by
side at the stop line, EquaBon 20-59 is used to compute shared-Iane capacity.
¿yVy
CSH = v
I...2..
YCm.y

where
eSI! eapacity of the shared lane (veh/h),
vy f10wrate of the y movement in the subjeet shared lane (veh/h), and
c",.y movement capacity of the y movement in the subject shared lane
(veh/h).

Step lOb: Flared Minor-Street Lane Effects

To estimate the capacity of a flared right-tum lane (such as in Exhibit 20-17),
the average queue length for each movement sharing the right lane on the minar-
street approach must first be computed.

Exhibit 20-17
capadty of a Aared-lane
Approach

•
n. QUEU£
L£"IGTH

This computation assumes the right-tum movement operates in one lane,

and the other traffie in the right lane (upstream of the flare) opera tes in another,
separate lane, as shown by Equation 20-60.
dsepvsep
Equation 20-60 Qsep = 3,600
where
Q"", average queue length for the movement eonsidered as a separate lane
(veh),
d><p control delay for the movcment considered as a separatc lane (as
described in Step 11),and
v"'P = flow rate for the movement (veh/h).
Next, the rcquired length of the storage area sueh that the approaeh would
operate effectively as scparate lanes is computed with Eguation 20-61. This value
is the maximum value of the queue lengths computed foc each separate
movement plus one vehiele.

Motorized Vehicle Core Methodology Chapter 20/Two-Way STOP-Controlled lntersectioos

Page 20-28 versJOn6.0
 Highway Capacity Manual: A Guide far Multimadal Mability Analysis

Equation 20.61

where
11,\1'" length of the storagc area sueh that the approaeh would operate as
separatc lanes;
Q"'I"¡ average gucue length for movement i eonsidered as a separate lane;
and
round round.off operator, rounding the quantity in parenthescs to the nearest
intcgcr.
Next, the eapacity of a separate Jane eondition c"'I' must be eomputed and is
assumed to be the eapacity of right-turning traffie opcrating as a separate lane
and the eapacity of the other traffic in the right lane (upstream of the fiare)
operating as a separate lane. The eapacity of a separate lane eondition is
ealrulated aeeording to Eguanon 20-62,

csep = min leR (1 + VL+TH).

VR
CL+TH (1 + ~)lVL+TH
Equation 20.62

where
c"'I' sum of the capacity of the righHurning traffie operating as a separate
Jane and the capacity of the other traffie in the right lane (upstream of
the fiare) operating in a separate lane (veh/h),
cR eapacity of t:heright-turn movement (veh/h),
CL+TH eapaeity of the through and leH-turn movemcnts as a shared lane
(veh/h),
VR right-tum movement flow rate (veh/h), and
VL+TH through and left-tum movement combined flow rate (veh/h).
Final1y,thc capaeity of the lane is computed, taking into aeeount the fiare,
The eapacity is interpolated as shown in Exhibít 20-17.A straight line is established
by using the values of two points: (e"", /lM",) and (CSII' O).The intcrpolated value of
the actual value of the flared-lane eapaeity cR is eomputed with Eguation 20-63.

e Csep - CSH) ~ + CSH ir nR ::;;n",ax

= n"'ax Equation 20-63
CR
{
Csep ir nR > n",ax
where
CR actual eapacity of the flaced lane (veh/h),
c"" eapacity of the lane jf both storage areas were infinitely long (rcfcr to
Eguation 20-62) (veh/h),
CSII eapacity of the lane when al! traffie shares one lane (veh/h), and
IIR actual storage arca for right-turning vehicles as defined in Exhibít 20-17.
The actual eapacity CR must be greater than CSII but less than or egual to '"'1"

Chapter 20/Two-Way STOP-Controlled intersections Motorized Vehide Core Methodology

Version 6.0 Page 20-29
Highway Capacity Manual: A Guide for MultimodaJ Mobility Analysis

Step 11: Compute Movement Control Delay

The delay experienced by a motorist is related to factors such as control type,
geometrics, traffic, and incidents. In the TWSC intersection methodology, only
that portion of delay aUributed to the STOP-controlaspect of the intersection,
referred to as control delay, is quantified.
Control delay ineludes delay due to deceleration to a stop at the back of the
queue from free.flow speed, move.up time within the queue, stoppcd delay at
the front of the queue, and delay due to acceleration back to free-f1ow speed.
With respect to field measurements, control delay is defined as the total time that
elapses from the time a vehiele stops at the end of the queue to the time the
vehiele departs from the stop lineoThis total elapsed time ineludes the time
required for the vehiele to travel from the last.in.queue position to the first-in-
queue po~ition, ineluding deceJeration of the vehiele from free.flow speed to the
speed of vehicles in the queue.

5tep 11a: Compute Control Delay to Rank 2 Through Rank 4 Movements

Average control deJay for any particular minor movement is a function of
the capadty of the approach and the degree of saturation. The anal)'tical model
used to estimate control deJa)' (Equation 20-64) assumes demand is less than
capacity for the period of analysis. If the degree of saturation is greater than
about 0.9, average control delay is significantly affected by the length of the
analysis periodo In most cases, the recommended analysis period is 15 minoIf

Equation 20-64
demand exceeds capadt)' during a 15-min period, the delay results computed by
the procedure may not be accurate. In this case, the period of analysis should be
lengthened to inelude the period of oversaturation.
•
where
d control delay (s/veh),
D" f10wrate for movement x (veh/h),
cm." capadty of movement x (vehlh), and
T analysis time period (0.25h for a 15-min period) (h).
A constiJnt value of 5 S/veh is The constant 5 s/veh is included in Equation 20-64 to account for the
tJsed ro retlect deJay during
deceleration ro and deceleration of vehicles from free-flow specd to the speed of vchiclcs in the
acceIeralion from él stop.
queue and the acceleration of vehieles from the stop line to free-flow speed.

Step llb: Compute Control De/ay to Rank 1Movements

The eHect of a shared lane on the major.street approach where left-turning
vehicles ma)' block Rank 1 through or righHurning vehicles can be significant. If
no exclusive left.turn pocket is provided on the major street, a delayed left-
turning vehicle may block the Rank 1 vehicles behind jt. This will delay not only
Rank 1 vehicles but .lIso lower.ranked movements. While the delayed Rank 1 •

Motorized Vehicle Core Methcxlology Chapter 20fTwo-Way SToP-Controlled Intersections

Page 2Q.30 V~6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Anolysis

vehieles are discharging from the queue formed behind a Icft-turning vehide,
they impcde lower-ranked conflicting movements.
Field observations have shown that such a blockage eHect js usually very
small, because the major street usually provides enough space for the blocked
Rank 1 vehiele to bypass the left-turning vehicle on the right. At a minimum,
incorporating this eHect requires estimanng the proportion of Rank 1 vehicles
being blocked and computing the average delay to the major-street left-tuming
vehicles that are blocking through vehicles.
In the simplest procedure, the proportion of Rank 1 major-street vehieles oot
being blacked (Le.,in a gueue-free state) is giveo by P~,iin Equation 20-43 (P~.I
should be substituted for the major lefHum factor PO.¡ in Equatian 20-43 in
computing the capacity of Jower-raoked movements that conflict). Therefore, thc
proportion of Rank 1 vehieles being blocked is 1 - l'~,i'
The average deJay to Raok 1 vehieles is computed with Eguatioo 20-65.

dRllnkl
_{(l-
-
p;,,)dMLT
Vi,l + Vi.2
(ir) N> 1 Equation 20-65

(l-p;,,)dM,LT N= 1
where
ditlnkl delay to Rank 1 vehieles (s/veh),
N number of through lanes per direction on the major street,
P~.i proportion of Rank 1 vehieles not blocked (from Eguation 20-43),
dM.lT delay to major-street left-turning vehic!es (from Eguatian 20-64)
(s/veh),
V¡,I major-street through vehicles in shared lane (veh/h), and
V;,2 major-street turning vehic!es in shared lane (veh/h).
On a multiJane road, only the major-street volumes in the lane that may be
blncked should be used in the computation as V',I and V;,2' On multilane roads, jf
jt is assumed blocked Rank 1 vehicles do not bypass the blockage by moving into
other through lanes (a reasonable assumption under conditions of high major-
street £Iows),then V¡,l = v2/N. Because of the unigue characteristics associated with
each site, the decision 00 whether tn account for this eHect is left to the analyst.

Step 12: Compute Approach and Intersection Control Delay

The control deJay for all vehieles on a particular approach can be computed
as the weighted average of the control delay estimates for each movement on the
approach. Eguanon 20-66 is uscd for the computanon.
drvr + dtvr + dlv¡ Equation 20-66
dA =
Vr + Vt + VI

where
control delay 00 the approach (s/veh);
computed control delay for the right-tum, through, and ¡eft-tum
movements, respectively (s/veh); and

Oiapter lO/Two-Way SToP-Controlled intersectioos Motorized Vehicle Core Methodology

VerslOn 6.0 Page 20-31
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

v" V" VI = volume or flow rate of right-turn, through, and left-turn traffic on
the approach, respectively (veh/h).
Similarly, the intersection control delay dI can be computed with Equation
20-67.
+ dA.ZvA,Z + dA,3VA,3 + dA,4VA.4
Equation 20-67 d ,-_ dA,lVA,l

VAl + VA,Z + VA,3 + VA4

where d".x is the control delay on approach x (s/veh), and VAI is the volume or
flow rate 00 approach x (veh/h).
In applying Equation 20-66 and Equation 20-67, the delay for all Rank 1
major-street movements is assumed to be O s/vch. LOS is not defined for an
overall intersection because major-street movemcnts with O s of deJay typically
result in a weighted average deJay that is extremely low. As such, total
intersection control de1ay calculations are typically used only when comparing
control delay among difierent types oí traffic control, such as two-way STOP
control versus all-way STOPcontrol.

5tep 13: Compute 95th Percentile Queue Lengths

Queue 1ength is an important consideration at unsignalized intersections.
Theoretical studics and empirical observations have demonstrated that the
probability distribution oí queue lengths for any minor movement at an
unsignalized interscction is a function of the capacity of the movement and the
volume of traffic being served during the analysis periodo Equation 20-68 can be
used to estimate the 95th percentile queue length for aoy minor movement at an
unsigoalized intersection during the peak 15-min pcriod 00 the basis of these
two parameters (10).

3.600) ( v, )
Equation 20.68
v,
--1+
cmx
.2.- _
( cm.x
1)' + c,;;:;-
(
150T
c;;;:; (~)
3,600

where
Q.~ 95th percentile queue (veh),
VI flow rate for movement x (veh/h),
c"'.x capacity oí movement x (veh/h), aod
T analysis time period (0.25 h for a 15-min period) (h).
The mean queuc leogth is computed as the product of the average delay per
vehicle and the flow rate for the movemeot of interes!. The expected total de1ay
(vehicIe hours per hour) equals the expected number oí vehicIes in the average
queue; that is, the total hourly delay and the average queue are numerically
identical. For example, four vehicle hours per hour of delay can be used
interchangeably with ao average queue length of four vehicles during the hour.

Motorized Vehicle Core MethOOoIogy Olapter 20fTwo-Way SToP-Cootro11edIntersectiOns

Page 20-32 Vt¥SiOn 6.0
 Highway Capacify Manual: A Guide for Multimodal Mobility Analysis

4. EXTENSION TO THE MOTORIZED VEHICLE

METHODOLOGY

INTRODUCTION
This section presents the details of incorporating pedestrian effects on
motorized vehicle capacity into the motorized vehicle methodology. The steps
below replace Steps 6 through 9 frorn Section 3.

REPLACEMENT STEPS TO INCORPORATE PEDESTRIAN EFFECTS ON

MOTORIZED VEHICLE CAPACITY
5tep 6: Compute Rank 1 Movement Capacities
Rank 1 major-street movements are assumed to be unirnpeded byany
movements of lower rank. Major-street movements of Rank 1 are not expected to
incm delay or slowing as they travel through the l1'VSC intersection. Empirical
observations have shown that such delays occasionally occur, and they are
accounted for by using adjustments provided later in this procedure.
For the purposes of this proccdure, major-street rnovements of Rank 1 are
assumed to be unimpeded by pedestrians at a lvVSC intersection, even though
research indicates sorne degree of Rank 1 vehicular yielding to pcdestrians (sec
the pedestrian rnethodology in Section 5). The assumption that pcdestrians do
not impede Rank 1 major-strcet movements is a known limitation in the
procedure.

5tep 7: Compute Rank 2 Movement capacities

Movements of Rank 2 (Ieft turos from the major street and right turns from
the minar street) must yield to conflicting major-strect thraugh and righHurning
vehicular movements of Rank 1 as well as conflicting pedestrian movements of
Rank 1. The movement capacity of each Rank 2 movement is egual to its
potential capacity, factored by any impedance due to pedestrians.

Step 7a: Pedestrian Impedance

Minar vehicular movements must yield to conflicting pedestrian movements
at a TWSC intersection. A factor accounting for pedestrian blockagc is computed
by Equation 20-69 on the basis of pcdestrian volume, pedcstrian walking speed,
and width of thc lane the minor movcment is negotiating into.
w
vxxs Equation 20-69
fpb = 3,60;
where
fpo pedestrian blockage factor or proportion of time that one lane on an
approach is blocked during 1 h;
Vz number of groups of pedestrians, where x is Movement 13, 14, 15, or 16;
w width of the lane the minor movcment is negotiating into (ft); and
Sp pedestrian walking specd, assumed to be 3.5 ft/s.

Chapter 20fTwo-Way SToP-Controlled Intersections Extension to tlle Motorized Vehicle Metllodology

VersiOn 6.0 Page 20-33
Highway Capadty Manual: A Guide for Multimodal MobiJjty Analysis

The pedestrian impedance factor for pedestrian movement x, PP.x, is

computed by Eguation 20-70.
Equation 20-70 PP-X = 1- fpb

Exhibit 20-18 shows that Rank 2 movements V1 and v~must yield to

pedestrian movements V16 and VIS> respl-"Ctively.Exhibit 20-18 also shows that
Rank 2 movement V9 must yield to pedestrian movements VI5 and vl4J and Rank 2
movement V12 must yield to pedestrian movements VI6 and v13• Rank 2 U-turn
movements Vw and V4U are assumed to not yield to pedestrians crossing the
major street, consistent with the assumptions stated previously for Rank 1
vehicles.

Exhibit 20-18 Vehicular Movement Must Yield to Pedestrian Impedance Factor tor
Relative Pedestrian-Vehide (vx> Movement Pedestrians (PA.)
Hierarchy for Rank 2
Movements " ',. Po.16

Po.15

(Po.IS)(PA\4)
(PP,\6)(P••.ll)

Step 7b: Movement Capacity for Major-Street Left-Tum Movements

Rank 2 major-street left-tum movements can be impeded by conflicting
pedestrians. The movement capacity c.••.
¡ for major-strcet left-turn movements is
computed with Equation 20-71.
Equation 20-71 Cm,) = 'p,i X Pp,i

where j denotes movements of Rank 2 priority, j denotes movcments of Rank 1

priority, and cp•i is thc potential capacity of movement j.

Step 7c: Movement CapaCity for Minor-Street Right- Turn Movements

The movement capacity cm,; for Rank 2 minor-street right-tum Movements 9
and 12 is impeded by two conflicting pedestrian movements, The capacity
adjustment factors are denoted by 19 and/12 for minor-street right-turn
Movements 9 and 12, respectively, and are given by Equation 20-72 and Equation
20-73, respectively.
Equation 20-72 f9 = PP,lS X Pp,14
Equation 20-73 112 = Pp,16 X PP,13

where
I'JI 112 capacily adjustment factor for Rank 2 minor-street right-turn
Movements 9 and 12, rcspedively; and
Pp,i "= probability that conflicting Rank 2 pedestrian movernent j will operate
in a queue-free sta te.
The movement capacity for rninor-street right-turn movernents is then
computed with Equation 20-74.
Equation 20-74 Cm,i = 'p,i X ti

Extension to the Motorized Vehicle Methodology Chapter 20fTwrrWay STOP-Controlled Intersections

Page 20-34 Version 6.0
 Highway Capacity Manual: A Guide tor MultimOdal Mobility Ana!ysis

where
c"'.i movemcnt capacity for Movements 9 and 12,
ep•i potential capacity far Movements 9 and 12 (from 5tep 5), and
f, capacity adjustment factor for Movements 9 and 12.

Step 7do'Movement Capacity for Major-Street U-Turn Movements

This step is the same as 5tep 7c in Section 3.

Step 7eo'Effect of Major-Street Shared Through and Left-Turn Lane

This step is the samc as 5tep 7d in Section 3.

Step 8: Compute Rank 3 Movement Capacities

Rank 3 minor-street traffic rnovements (minor-street through movements at
four-Ieg intersections and rninor-street left tums at three-leg intersections) must
yield to conflicting Rank 1 and Rank 2 rnovements. Not aH gaps of acceptable
length that pass through the intersection are normally available far use by Rank 3
movements because sorne of them are likely to be used by Rank 2 movements.
If the Rank 3 movement is a two-stage movernent, the movement capacity
for the one-stage movement is computed as an input to the two-stage calculation.

Step $a: Pedestrian Impedance

Exhibit 20.19 shows that Rank 3 movements Vs and Vil must yield to
pedestrian movements VIS and V16•

Vehicular Must Yield to Pedestrian Impedance Factor for Exhibit 20-19

Movement ( v~) Movement Pedestrians (PA-~) Relative Pedestrian-Vehide

• (PAlS)(PAl6)
Hierarchy for Rank 3
Movements
'" (P".IS)(P""6)

The pedestrian impedance factor for Rank 3 movements is computed

according tú Equation 20-69 and Equation 20-70.

Step Sbo'Rank 3 Capacity for One-Stage Movements

This step is the same as 5tep Sa in Section 3, except that the capacity
adjustmcnt factor l. far aH movements k and foc all Rank 3 movements is given
by Equation 20-75.

fk = n
i
PO.j X Pp.r Equation 20-75

where n indica tes the product of a series of terms, and

l'O.i probability that conflicting Rank 2 movement j will opera te in a queue-
free state,
flr.~ probability of pedestrian mavements af Rank 1 or Rank 2 priority,
k Rank 3 movements, and
x = 13,14,15, or 16 (pedestrian movements ofboth Rank 1 and Rank 2).

O1apter 20fTwo-Way $TOP-Controlled Intersections Extension to the Motorized Vehicle Methodology

Vef5ion 6.0 Page 20-35
Highway Capacity Manual; A Guide tor Multimodol MobWty Anolysis

Step Be: Rank 3 Capacity for Two-Stage Movements

This step is the same as 5tep 8b in Section 3.

Step 9: Compute Rank 4 Movement capacities

Rank 4 movements (Le., only the minor-street left tums at a four-leg
intersection) can be impeded by al! higher-ranked movements (Ranks L 2, and 3).

Step 9a: Pedestrian Impedance

Exhibit 20-20 shows that Rank 4 movemcnt V7 must yield to pl.>dcstrian
movements VIS and VIJ, and Rank 4 movement VIO must yield to pedestrian
movements VI6 and Vu-

Exhibit 20-20 Vehicular Must Yield te Pedestrian Impedance Factor for

Relative Pedestrian-Vehide Movement ( vx) Movement Pedestrians (Pax)
Hierarchy for Rank 4 (p,\1~)(h13)

"""
Movements (PV16)(Pc")

The pedestrian impedance factor for Rank 4 movements is computed

according to Equation 20-69 and Equation 20-70.

Step 9b: Rank 4 Capaa"ty for One-Stage Movements

This step is the same as Step 9a in Section 3, except that the capadty

•
adjustment factor for the Rank 4 minor-street left-tum movement can be
computed by Equation 20-76.
Equation 20-76 ¡, = pi X PO) x Pp.x
where
minor-street lcft-tum movement of Rank 4,
J conflicting Rank 2 minor-street right-tum movement, and
Pp,x values shown in Equation 20-70 (the variable po.¡ should be induded
only jf movement j is identified as a conflicting movement).

Step 9c: Rank 4 Capacity for Two-Stage Movements

This stcp is the same as Step 9b in Section 3.

Extension to the Motorized Vehide Methodology

Page 20-36
Qlapter 20fTwo-Way SrOP-Controlled Intersections
Ver-siOn 6. O
•
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

5. PEDESTRIAN MODE

SCOPE OF THE METHODOLOGY

This methodology applies to nvsc intersections and midblock crossings at
which pedestrians cross up to four through lanes on the major street. It is applied
through a series of steps requiring input data related to vehiele and pedestrian
volumes, geometric conditions, and motorist yield rates to pedestrians.

Spatial and Temporal Limits

This scction's methodology applies to pedestrian crossings across an
uncontroJ1ed approach of a TWSC intersection or at a midblock location. The
recommended length of the analysis period is the HCM standard of 15 min
(although longer periods can be examined).

Performance Measures
This methodology produces the following performance measures:
• Average pedestrian delay, and
• LOS based on average pedestrian deJay.

Limitations of the Methodology

The pedestrian methodology's limitations differ from the Jimitations of the
motorized vehiele mode because the methods were developed in separate
research efforts. The pedestrian methodology does not appJy to undivided streets
with more than four lanes, although it can accommodate up to four lanes in each
direction separated by a median. lt does not account for interaction effects of
upstream signalized interscctions, and it assumes random arrivals on the major
street and equal directional and ¡ane distribution on the major street.
The methodology does not take into account pedestrian cross flows (i.e.,
pedeslrian flows approximately perpendicular to and crossing another
pedestrian stream), and it assumes the pedestrian will reach the crossing without
delay from pedestrians traveling parallel to the major 5treet. Under high
pedestrian volumes, this assumption may not be reasonable.
Thc method is for steady statc conditions (i.e., the demand and capacity
conditions are constant during the analysis period); it is not designed to evaluate
how fast or how often the facility transitions from one demand or capacity statc
to anothcr.

Alternative Tool Considerations

This section offers a mcthod for estimating thc delay and LOS for pedestrians
crossing a major street at a TWSC intersection or midblock location. Sorne
simulation-bascd tools offer a more detailed treatment of the arrival and
departure of vehieles and their interaction with pcdestrians, but for most
purposes the HCM procedure produces an acceptable approximation.
The identified limitations for this chapler are shown in Exhibit 20-21, along
with the poten tial for improved treatmcnt by alternative tOol5.

Chapter 20fTwG-Way STOP-Dlntrolled Intersectioos Pedestrian Mode

Version 6.0 Page 20-37
Highway Capacity Manual: A Gujde fer Multimedal MebUity Analysis

Exhibit 20-21 Limitation Potential for Improved Treatment by Altemative Tools

Umitations of ttle HCM TWSC Undivided streets with more Simulation tools may be able te accommodate larger lane
Pedestrian Procedure ttlan four Ianes configurations.
Effects of upstream Simulation tools can inelude an unSignalized intersection explícitly
intersections wittlin a signalized arterial or network.
Pedestrian cross fIows parallel
to ttle major street ttlat Simulatioo tools that model pedest:rian f10ws explicitly may be
impede pedestrian crossings able te capture ttlis effect.
across ttle major street
Non-steady state oonditions Most altemative tools provide for multiperiod variation of demand
for demand and capacity and, in sorne cases, capacity.

REQUIRED INPUT DATA ANO SOURCES

Exhibít 20-22lists the information necessary to apply the pedestrian
methodology and suggests potential sources for obtaining these data. It also
suggcsts default values for use when specific informanoo is not available.

Exhibit 20-22 Suggested

Required Input Data, Potential Required Data and Units Potential Data Source{s) Default Value
Data Sources, and Detault Geometric Data
Values for TWSC Pedestrian
Number of lanes on ttle majOf street Design p1ans, road inventery Must be provided
Analysls
Crosswalk. lengttl (ft) Design plans, road inventory Must be províded
Crosswalk width (ft) Design plans, rcad inventory Must be provided
Presence of a raised median te allow a
Deslgn plans, road inventory Must be provided
two-stage crossing
Posted speed limit Design plans, road inventory Must be provided
[)emanc!Data
Pedestrian ftow rate (pis) Field data, modeling Must be provided
Presence of pedestrian platoooing Reld data, modeling Must be provided
Conflicting vehicular f10w rate (veh/s) Reld data, modelíng Must be provided
Average pedestrian walking speed (ft/s) Reld data 3.5 ft/s
Pedestrian start-up time and erIÓ c1earance Reld data
time (s)
3,
Mean motorist yielding rate to pedestrians FlE!lddata, literature Must be provided.
Note: 4 Samplevaluesfrorn t:heliterature are provided In tt1issection (see Exhibit 20-2").

COMPUTATIONAL STEPS
The required steps are ilIustrated in Exhibit 20-23.

Exhibit 20-23
5tep 1: Identify Two-Stage Crossings
TWSC Pedestrian
MethOOoIogy

5tep 2: Determine Critical Headway

5tep 3: Estimate Probability of a Delayed Crossing

5tep 4: CaIQJlateAverage Delay to Wait ter Adequate Gap

5tep 5: Estimate Delay Reduction due to Yielding Vehides

Pedestrian Mode
Page 20-38
Step 6: CaIQJlateAverage Pedestrian Delay
and Detennine LOS

Chapter 20/Two-Way SrOP-Controlled Intersectioos

Version6.0
•
 Highway Capacity Manual: A Guide lor Multimodal Mobility Analysis

Step 1: Identify Two-Stage Crossings

When a raised median refuge island is available, pedestrians typicalIy cross
in two stages, similar to the two-stage movement desceibed foe motorized
vehic1es eaeHer in this chaptee. Determination of whether a median refuge exists
may require engineering judgment. The main issue to determine is whether
pedestrians cross the traffic streams in one or two stages. When pedestrians ceoss
in two stages, pedestrian delay should be estimated separately for each stage of
the crossing by using the procedures described in Steps 2 through 6, separating
the conflicting vehicular volume accordingly. To determine pedestrian LOS, the
pedestrian delay for each stage should be summed to establish the average
pedestrian delay associated with the entire crossing. This service measure is used
to determine pedestrian LOS for a nvsc intersection with two.stage crossings.

Step 2: Determine Critical Headway

Critical headway for
The procedure for estimating critical headway for pcdestrians is similar to pedestrians fs similar lo mtíeiJ/
that described far motorized vehicles. The critical headway is the minimum time headway for motorlzed
interval in seconds below which a pedestrian will oot attempt to begin crossing vehides.
the street. Pedestrians use their judgment to determine whether the available
headway between conflicting vehicles is long enough for a safe crossing. If the
available headway is greater than the critical headway, it is assumed the
pedestrian will cross, but if the available headv .•ray is less than the critical
headway, it is assumed the pedestrian will not crass.
Far a single pedestrian, critical headway is computed with Equation 20-77.
L
tc=S+ts Equation 20.77
p

where
te critica! headway for a single pedestrian (s),
Sp average pedestrian walking speed (ft/s),
L crosswalk length (ft), and
t, pedestrian start-up time and end dearance time (s).
Groups o, pedestri<Jns require
If groups of pedestrians are observed crossing in the field (Le., a platoon, or
computation o{ theír spaaa!
more than one pedestrian crossing at a time), then the spatial dislribution of distJibution.
pedestrians should be computed with Equation 20-78. The spatial distribution of
pedestrians represents the number of rows of pedestrians waiting to cross, with
the first row in position to cross and subsequent rows lined behind the first row.
1£the crosswalk is wide enough to accommodate a group of pedestrians traveling
side-by-side without needing to also travel behind one another, theo the spatial
distribution oi pedestrians equals ooe. lf 00 pedestrian grouping is observed, the
spatial distributioo oi pedestrians is assumed to be ooe.
. [S.O(N, -
Np = mt -----
1)] + 1 Equation 20.78
W,

Chapter 20fTwo-Way STop-Controlled Intersections Pedestrian Mode

Version 6.0 Page 20-39
Highway Capacify Manual: A Guide tar Multimadal Mobilify Analysis

where
N, : spatia! distribution of pedestrians (p),
N, total number of pedestrians in the crossing platoon (from Equation
20-79) (p),
W, crosswalk width (ft), and
8.0 defau!t c1ear effective width used by a single pedestrian to avoid
interference when passing other pedestrians (ft).
To compute spatial distribution, the analyst must make field observations or
estimate the platoon size by using Equation 20-79.
v eVpte + ve-vte
Equation 20.79
N -~p--- __ -
e - (vp +v)e(vp-v)te

where
Ne total number of pcdcstrians in the crossing platoon (p),
vp pedestrian flow rate (pIs),
v '" conflicting vehicular flow rate (veh/s) (combined flows for one.stage
crossings; separate flows for two-stage crossings), and
te '" single pedestrian critical headway (s).
The group critical headway is the critica! headway needed to accommodate a
group oí pedestrians. The group critica! headway is determined with Equation
20-80.
Equation 20.80 tC•G = te + 2(Np - 1)
where
(,G group critical headway (s),
( critical headway for a single pedestrian (s), and
Np '" spatial distribution of pedestrians (p).

5tep 3: Estimate Probability of a Delayed Crossing

On the basis of the calculation of the critical headway te,e- the probability that
a pedestrian will not incur any crossing delay is egua! to the likelihood that a
pedestrian will encounter a gap greater than or equa! to the critical headway
immediately upon arrival at the intersection.
Assuming random arrivals of vehic1es on the major street and egual
distribution of vehic1es among aH through lanes on the major street, the
likelihood that a gap in a given !ane does not exceed the critica! headway is as
shown in Equation 20-81, Because traffic is assumed to be distributed
independently in each through lane, Equation 20-82 shows the probability that a
pedestrian incurs nonzero delay at a TWSC crossing.
Equation 20-81 -te,GV
Pb = 1- e --¡¡¡:-
Equation 20-82 Pd = 1 - (1 - Pb)NL

Pedesbian Mode Olapter 20fTwo-Way STOP-Controlled intersections

Page 20-040 Version 6.0
 Highway Capadty Manual: A Guide for Multimodal Mobility Analysis

where
Pb probability of a blocked lane,
Pa probability of a delayed crossing,
NL number of through Janes crossed,
tc,G group critical headway (s), and
v conflicting vehicular flow rate (veh/s) (combined flows for one-stage
crossings; separate flows for two-stage crossings).

Step 4: Calculate Average Delay to Wait for Adequate Gap

Research indica tes average deJay to pedestrians at unsignalized crossings,
assuming no motor vehides )'ield and the pedestrian is forced to wait for an
adeguate gap, depends on the critical headway, the vehicular flow rate of the
subject crossing, and the mean vehicle headwa)' (11).The average dela)' per
pedestrian to wait for an adcquate gap is given b)' Equation 20-83.
1
dy =_(evt',G_vt
v e,<-1)
Equation 20-83

where
dg average pedcstrian gap dela)' (s),
tc•G group critical headway (s), and
v conflicting vehicular flow rate (veh/s) (combined flows for one-stage
crossings; separate flows for two-stage crossings).
The average delay for any pedestrian who is unable to cross irnmediately
upon reaching the intersection (e.g., any pedestrian experiendng nonzero deJay)
is thus a function of PJ and dg, as shown in Equation 20-84.
dg
d yd --
- P Equation 20-84
a

where
dga average gap delay far pedestrians who incur nonzero delay,
dg average pedestrian gap dela)' (s), and
Pd probability of a delayed crossing.

Step S: Estimate Oelay Reduction due to Yielding Vehicles

When a pedestrian arrives at a crossing and finds an inadequate gap, that
pedestrian is delayed until one of two sihlations occurs: (a) a gap greater than the
critical headway is available, or (b) motor vehides yield and allow the pedestrian
to cross. Equation 20-83 estimates pedestrian delay when motorists on the major
approaches do not yield to pedestrians. When motorist )'ield rates are
significantly higher than zero, pedestrians will experience considerably less
detay than that estimated by Eguation 20-83.
In the United States, motarists are legally required to yield to pedestrians,
under mast circurnstances, in both marked and unmarked crosswalks. However,
actual rnotorist yielding hehavior varies considerably. Motorist )'ield rates are

Chapter 20fTwo-Way STOP-COntrolled Intersections Pedestrian Mode

Version 6.0 Page 20-41
Highway Capacify Manual: A Guide lar Multimodal Mobility Anolysis

influenced by a range of factors, including roadway geometry, travel speeds,

pedestrian crossing treatments, local culture, and law enforcement practices.
Research (12, 13) provides information on motorist responses to typical
pedestrian crossing treatments, as shown in Exhibit 20-24. The exhibit shows
results from two data collection methods. Staged data were collected with
pedestrians trained by the research team to maintain consistent positioning,
stance, and aggressiveness in crossing attempts. Unstaged data were collected
through video recordings of the general population. The values shown in Exhibit
20-24 are based on a limited number of sites and do not encompass the full range
of available crossing treatments. As always, practitioners should supplement
these values with local knowledge and engineering judgment.

Exhibit 20-24 Unstaged

Effect of Pedestrian Crossing Staged Pedestrjans Pedestrians
Treabnents an Motorist Yleld No. No. Mean Yield
Rate< Crossinn Treatment of Sites
Mea: ~~~~d
Rate Ojo of Sites Rat~ i~¡;'"
Overhead f1ashing beacon (push-button activatian) 3 47 4 49
Overhead f1ashing beacon (passive activatian) 3 31 3 67
Pedestrian crossing f1ags 6 65 4 74
In-street crossing signs (25-30 rni/h) 3 87 3 90
High-visibility signs and markings (35 mi/h) 2 17 2 20
High-visibility signs and markings (25 mi/h) 1 61 1 91
Rectannular ranid-fIash beacon NA NA 17 81
Note: NA '" not available.
SOurce: Fitzp<ltrick et <11.(1.l¡ <lnd Shurbutt et <11.(J.J;.

Depending on (he aossing It is possibJe for pedestrians to incur less actual delay than dg because of
treatment and other faetors, yielding vehicles. The likeJihood of this situation occurring is a function oi
motorist behavior vades
significantfy. vchicle volumes, motorist yield rates, and number of through lanes on the major
street. Consider a pedestrian waiting for a crossing opportunity at a TWSC
intersection, with vehicles in each conflicting through Jane arriving every h
seconds. On average, a potential yielding event will occur every h scconds. As
vehicles are assumed to aerive randomly, each potential yielding event is
considered to be independent.
For any given yielding event, each through lane is in one of two states:
1. C1ear-no vehicles are arriving within the critical headway window, or
2. Blocked-a vehicle is arriving within the critical headway window. The
pedestrian may cross only if vehicles in each blocked Iane choose to yield.
lf vehicles do not yield, the pedestrian must wait an additional h seconds for
the next yielding evento On average, this process will be repeated until the wait
exceeds the expe<:teddeJay rcquired for an adequate gap in traffic (dgd), at which
point the average pedestrian will reccive an adequate gap in traffic and will be
able to cross the strcct without having to depend on yielding motorists.
Average pedestrian deJay can be calculated with Equation 20-85, where the
first term in the equation represents expected delay from crossings occurring
when motorists yield, and the second term represents expected delay from
crossings when pedestrians wait for an adequate gap.

Equatlon 20-85
dp ~ t h(i - O.5)P(Y¡) + (Pd - t P(Y¡)) dgd

Pedestrian Mode Chapter 20{Two-Way STOP-Controlled lntersectioos

Page 20-42 Ve~6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

where
dr average pedestrian delay (s);
I = crossing event (i = 1 to 11);
h average head1,','ayfor each through lane = (NJv) (s);

P(Y,) probability that motorists yield to pedestrian on crossing event i;

probability of a delayed crossing; and
1I = average number of crossing events befare an adequate gap is available
= int(d •.Jh).
Equation 20-85 reguires the calculation of P(Y¡). The probabilities P(Y¡) that
motorists 1,','illyield for a given crossing event are considered below for
pedestrian crossings of one, t1,','o,three, and four through lanes.

One-Lane Crossing
Under the scenario in which a pedestrian crosses one through lane, P(Y,) is
found simply. When i = 1, P(Y¡) is egual to the probability of a delayed crossing
Pd multiplied by thc motorist yield rate My• For i = 2, P(Y¡) is equal to My
multiplied by the probability that the second yielding event occurs (Le., that the
pcdestrian did not cross on the first yielding event), P/(1 - My). Equation 20-86
gives P(Y¡) for any i.
Equation 20-86
P(Y,J = PdMy(l- Myt'
where
My motorist yield rate (decimal), and
crossing event (i = 1 to ll).

Two-Lane Crossing
Far a two-lane pedestrian crossing at a TWSC intersection, P(Y,) requires
either (a) motorists in both lanes to yield simultaneously if both lanes are blocked
or (b) a single motorist to yield if onIy onc lane is blocked. Because these cases
are mutually exclusive, where i = 1, P(Y¡) is given by Equation 20-87.
Equation 20-87
P(Yl) = 2Pb(I - Pb)My + P;M;
where Po is the probability of a blocked lanc.
Equation 20-88 shows P(Y.) where i is greater than one. Equation 20-88 is
equivalent to Equation 20-87 if P(Yo) is set to equal zero.

P(Y,) = [Pd - ~ P(ljJ] [(2Pb[1 - PbJ;'y) + (P~M~)l Equation 20-88

Three-Lane Crossing
A three-Iane crossing follows the sarne principies as a two-Iane crossing.
Eguation 20-89 shows the calculation far P(Y,).

Chapter 10/Two-Way SToP-Controlled lntersections Pedestrian Mode

Version 6.0 Page 20-43
Highway Capacify Manual: A Guide for Multimodal Mobility Analysis

P(Y,) = Pd

where P(Yo) ~ O.
[ -
~¿ P(lí) 1 [PfM~+ 3Pt(1-
j~O
Pb)M;
P
d
+ 3Pb(l- Pb)2My]

•
Four-Lane Crossing
A four-lane crossing follows the same principies as aboye. Equation 20-90
shows the calculation for P(Y;).

Equation 20-90 P(Y,) = [pr ~ P(lí)]

X [P:M: + 4Pf(1-Pb)M~ + 6Pip:1- Pb)2M; + 4Pb(l- Pb)3My]

where P(Yo) '" O.

Step 6: Calculate Average Pedestrian Delay and Determine LOS

The delay experienced by a pedestrian is the service measure. Exhibit 20-3
lists LOS criteria for pedestrians at TWSC intersecnons based on pedestrian
delay. Pedestrian delay at TWSC intersections with two-stage crossings is equal
to the sum of the delay for each stage of the crossing.

Pedestrian Mode Chapter 20fTwo-Way STOP-Controlled Intersections

Page 20-44 VeiWn6.0
 Highway Capacity Manual: A Guide for Multimodal MobiJity Analysis

6. BICYCLE MOOE

As of the publication of this edition of the HCM, no methodology spedfic to

bieyclists has been developed to asscss the performance of bieyclists at TWSC
interseetions, as few data are available in the United States to support model
ealibration or LOS definitions. Depending on individual eomfort level, ability,
geometrie eonditions, and traffic conditions, a bicyclist may travel thraugh the
intersection eithcr as a motor vehicle or as a pcdestrian. Critical headway
distributions have becn identified in the research (14, 15) for bicycles crossing
two.lane major streets. Data on critical headways for bicycles under many
circumstances are not readily available, however. Bieycles also differ fram motor
vehicles in that they normally do not queuc linearly at a STOP sign_ Instead,
multiple bicycles often use the same gap in the vehicular traffie stream. This
practiee probably affeets the determination of bicycIe follow-up time. This
phcnomenon and others described in this seetion have not been adequatcly
researched and are not explicitly induded in the methodology.

Chapter 20/Two-Way STOP-Controlled intersections Bicycle Mode

V~6.0 Page 20-45
H;ghway Capacity Manual: A Guide for Multimodal Mobility Analys;s

7. APPLICATIONS

TYPES OF ANALYSIS
The methodology of this chapter can be used in three types of analysis:
operational analysis, design analysis, and pJanning and preJiminary engineering
analysis.

Operational Analysis
The methodology is most easily applied in the opcrational analysis mode. In
operational analysis, aH traffic and geometric charaderistics oi the analysis
segment must be spedfied, including analysis-hour demand volumes ior each
turning movement in vehicles per hour, percentage of heavy vehicles for each
approach, peak hour factor for aH demand volumes, lane configurations, speciiic
geometric conditions, and upstream signa! in(ormation. The outputs oi an
operational analysis are estimates oi capacity, control deJay, and queue lengths.
The steps oí the methodology, described in this chapter's methodology section,
are iollowed diredly without modification.

Design Analysis
The operational analysis described earHer in this chapter can be used for
design purposes by using a given set of traffic flow data and iteratively
determining the numbcr and configuration of lanes that would be required to
produce a givcn LOS.

Planning and Preliminary Engineering Analysis

The operational analysis method described earHer in this chapter provides a
detailed procedure far evaluating the performance of a lVVSCintersection. To
estimate LOS íor a (uture time horizon, a planning analysis based on the
operational method is used. The planning method uses all the geometric and
traffic flow data required for an operational analysis, and the computations are
identical. However, input variables for percentagc of heavy vehicles and peak
hour factor are typically estimated (or defaults are used) when planning
applications are performed.

EXAMPLE PROBLEMS
Scction 2 of Chapter 32, STOp.Controlled lntersections: Supplemental,
provides five example problems that illustrate each of the computational steps
¡nvolved in applying the motorized vehicle method:
1. Analyze a TWSC intersection with thrce legs,
2. Analyze a pedestrian crossing at a nvsc intersection,
3. Analyze a nvsc intersection with flared approaches and median storage,
4. Analyze a nvsc intersection within a signalized urban street segment,
and
5. Analyze a 1VVSCintersection on a six-Iane street with U-tums and
pedestrians.

AppliCatiOns Chapter 20/Two-Way STOP-COOtrollecrIntersections

Page 20-46 V~6.0
 Highway Capacify Manual: A Guide for MultimodaJ Mobility Analysis

EXAMPLE RESULTS
Analysis o[ TWSC intersections is commonly periormed to determine
whether an existing intersection or driveway can remain as a nvsc intersection
or whether additional treatments are necessary. These treatments, including
geometric modifications and changes in traffic control, are discussed in other
references, including the presentatian of traffic signal warrants in the Mallua/ 011
Uniform Traffie Control O('Vices for Streets und Highways (16). This section discusses
two common situations analysts face: the analysis of shared versus separate lanes
and the interpretation of LOS F.

Analysis of Shared Versus Separate Lanes

Sorne movements, most often lefHum movements, can sometimes have a [ntefpretation of (he effects of
shared fanes should conSider
poorer LOS when given a separa te lane than when they share a lane with another both de/ay associated Wlth
movement (usually a through movement). This is not inconsistcnt in terms of the individUiJI movement'i and
delayassoeiated with a!/
stated crHeria. Left-turn movements will generall)' experiencc longer control vehides on a given approach.
dela)'s than other movements because uf the natUfe and priority o[ the
movement. The control delay for leh turns in a shared lane ma)' be less than the
control delay fur le[t turns in a separate lane. However, if dela)' for all vehieles
on the approach or at the intersection is considered, providing separate lanes will
rcsult in lower total delay.

Interpretation of LOS F
LOS F occurs when there are nut enough gaps of suitable size to allow
minor-street vehieles to enter or cross through traffic on the major street; this
results in long average control delays (greatcr than 50 s/veh). Depending on the
demand on thc approach, long queues on the minor approaches may result. The
method, howevcr, is based on a constant critical headway.
LOS F may also appear in the form of drivers on the minor street selecting
smaller-than-usual gaps. In such cases, safety issues ma)' occur, and some
disruptian to the majar traffic stream ma)' resultoWith lawer demands, LOS F
may nat always result in long queues.
At nvsc intersections, the critical movement, oftcn the minor-street left
turn, may control the overall performance af the intersection. The lawer
threshold for LOS F is set at 50 s of dela)' per vehiele. In sorne cases, the delay
equations will predict delays greater than 50 s for minor-strcct movements under
very low-volume conditions on the minor street (fewer than 25 veh/h). On the
basis of the first term of the delay equation, the LOS F threshold is reached with a
movement capacity uf approximatcly 85 veM1 or less, regardless of the minar-
street movement valume.
This analysis procedure assumes random arrivals on the major street. For a
typical majar street with two lanes in each direction and an average traffic
volume in the range of 15,000 to 20,000 veh/day (roughly equivalent ta a peak
hour fIow rate of 1,500 to 2,000 veM1), the delay equatiun will predict greater
than 50 s of delay (LOS F) for many urban TWSC intersections that allow minor-
street left-tum movcments. LOS F will be predictcd regardless of the volume of
minor-street left-turning traffie. Even with a LOS F estimate, most low-volume

Chapter 20/Two'Way STOP-Controlled Intersections Applications

Version 6.0 Page 20-47
Highway Capacity Manual: A Guide for Munimodal Mobility Analysis

minor-street approaches would not mect any of the volume or delay warrants for
signalization noted in the Manual 011 Ul1iform Traffic Control Devices (16). As a
result, analysts who use the HCM LOS thresholds as the sole measure to
determine the design adequaey of TWSC intersections should do so with caunon.
In evaluating the overall performance of TWSC intersections, it is important
to consider measures of effectiveness such as volume-to.capadty ratios for
individual movements, average queue lengths, and 95th percentile queue lengths
in addition to considering delay. By focusing on a single measure of effectiveness
for the worst movement only, such as deJay for the minor.street left tum, users
may make less effective traffic control dedsions.

Applications Chapter 20fTwo-Way SToP-Controlled Intersections

Page 20-48 Version 6.0
 Highway Capacity Manual: A Guide tor Multimodal Mobility Analysis

8. REFERENCES

1. Kyte, M., Z. Tian, Z. Mir, Z. Hamredmansoor, W. Kittclson, M. Vandehey, B. Sorne of these references are
avaifable in the TeetmiCal
Robinson, W. Brilon, L. Bondzio, N. Wu, and R. Troutbeck. NCHRP Web Reference Library in VoIume 4.
Dacument 5: Capacíty and u'vd oJ Service at Unsignalized lntersectiolls: Final
Report, Vo/ume 1- Two-Way Stap-Controlled lIrterseetiolls. Transportabon
Research Board, National Rcsearch Council, Washington, D.C., 1996.
http://www.nap.edu/books/nch005/html. Accessed March 19, 2010.
2. Brilon, W., and M. Gro15mann. Aktua/isiertes Bereeh/ll.l.I/gsverJallrenfiir
Kllatellpunkte olme Lichtsiglla/anlagen (Updated Calclllation Method Jor
Unsignalized 11lterseefiO/ls). Forschung Strassenbau und
Strassenverkehrstechnik, Heft 596, Bonn, Germany, 1991.
3. Kittelson, W. K., and M. A. Vandehey. Delay Effects on Driver Gap
Acceptance Characteristics at Two-Way Stop.Controlled lntersections. In
Transportation Rcsearch Record 1320, Transportation Research Board, Natíonal
Research Counci\, Washington, D.C., 1991, pp. 154-159.
4. Troutbeck, R. £stimating the Critical AeceptanCf? Gap frOIll Traffic Movements.
Research Repart 92 5. Queensland University of Technology, Brisbane,
4

Australia, March 1992.

5. Zegeer, J. D., M. A. Vandchcy, M. Blogg, K. Nguyen, and M. Ereti. NCHRP
Report 599: DeJault Values Jor Highway Capacify Q/ld Leve! oJ Service AIlQlyses.
Transportation Research Board of the National Academies, Washington,
D.C., 2008.

6. Líu, P., T. Pan, J. J. Lu, and B. Cao. Estimating Capacity oi U.Turns at

Unsignalized Intersections: Conflicting Traific Volume, Impedance Effects,
and Left Turn Lane Capacity. In Trallsportation Researcl, Record: ¡auma/ oJ the
4

Transportatioll Research Board, No. 2071, Transportation Research Board of the

National Academies, Washington, D.C., 2008, pp. 44-51.
7. Hardcn;, J. Die Leistll1lgsJaehigkeit Ilicht sigllalgeregelter staedtischer
Verkeltrsknoten (Tlle Capacífy o/ Ullsigllalized Urballlnterseclions). Series
Strassenbau und Strassenverkehrstechnik, Vol. 76,1968.
8. Robertson, D. 1.Coordinating Traffic Signals to Reduce FueI Consumption.
In Proceedil/gs o/tite Royal Socíefy Series A, Vol. 387, No. 1792, 1983, pp. 1-19.
9. Wu, N., and W. Brilon. Modcling Impedance Effccts of Left Turners from
Major Streets with Shared Short Lanes at Two-Way Stop-Controlled
Intersections. In Trallsportafion Resenrelt Record: ¡ournal oJ tIJe Transporfatioll
Rescarelt Board, No. 2173, Transportation Research Board of the National
Academies, Washington, D.C., 2010, pp. 11-19.
10. Wu, N. An Approximation for the Distribution of Queue Lengths at
Unsignalized lnterscctions. In Proceedillgs, 2nd JI/ternatiollal Symposium 011
Higltway Capacíty, Vol. 2, Australian Road Rcscarch Board, Ud., Melbourne,
Australia, Aug. 1994.

Chapter 20{Two-Way STOP-Controlled Intersections References

VersiOn 6.0 Page 20-49
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

11. Gerlough, D. L., and M. J. Hubcr. Special Report 165: Traffic Flow Theory: A
MOllograph. Transportation Research Board, National Research Coundl,
Washington, D.C., 1975.
12. Fitzpatriek, K, S. M. Turner, M. Brewer, P. J. Carlson, B. Ullman, N. D. Trout,
E. S. Park, J. Whitaere, N. Lalani, and D. Lord. NCHRP Report 562: Improving
Pedestrian Safety at Unsignalized Crossillgs. Transportation Research Board of
the National Academies, Washington, D.C., 2006.
13. Shurbutt, J., R. G. Van Houten, and S. M. Turner. Analysis of Effeetsof
Stutter Flash LEO Beacons to Inerease Yielding to Pedestrians Using
MuItilane Crosswalks. Presented at 87th Annual Meeting of the
Transportation Research Board, Washington, D.C., 2008.
14. Opiela, K S., S. Khasnabis, and T. K Datta. Determination of the
Characteristics of Bicyele Traffie at Urban Intersections. In Transportatioll
Research Record 743, Transportation Researeh Board, National Research
Counol, Washington, D.C., 1980, pp. 30-38.
15. Ferrara, T. C. A Study ofTwo-Lane Intersections and Crossings Under Combined
Motor Vehicle and Bicyc1e Demands. Final Report. Report No. 75-5. University
ofCalifornia, Davis, Dec.1975.
16. Manual on UlIiform Traffic Control Devices Jor Streets and Highways. Federal
Highway Administration, Washington, D.C., 2009.

References Chapter 20/Two-Way 5T0J>-C0ntrolled Intersections

Page 20-SO V~6.0
 Highway Capacity Manual: A Guide for Mu/tjmodal MobUity Analysis

CHAPTER 21
ALL-WAY STOP-CONTROLLED INTERSECTIONS

CONTENTS

1. INTR O O U CTI O N ......................•...•...•...••..••..••...........•...•...•...•...................•.......•. 21-1

Chapter Organization 21-1
Related HCM Content 21-1

2. CONCEPTS 21-2
Phase Patterns . 21-2
Dcparture liead way . 21-4
Capacity Concepts 21-5
Level-of-Scrvice Criteria 21-9

3. MOTORIZED VEHICLE eORE METHOOOlOGY ..............•...•................. 21.10

Scopc of the Methodology 21-10

Requircd Input Data and Sources...................... . 21-10
Computational Steps 21-11

4. EXTENSION TO THE MOTORIZED VEHlCLE METHODOLOGY 21.20

Introduction 21-20
Rcplacement Steps for Three-Lane AWSC intersection Approaches 21-20

5. PEO ESTRIAN M O O E .........................•.....................................................•...•...• 21.22

Vehicular Volumes 21-22
Number of Approach Lanes 21-23
Proportion of Turning Traffic. 21-23
Pedestrian Volumcs 21-23

6. B1CYCl E M O D E ...........••...•................................................•...••.......................... 21-24

7. A PPLI CA TI O N S .....................•...•...••.......•...........................................••...•.......... 21-25

Types of Analysis 21-25
Example Problems 21.25
Example Results 21-25

8. R EF ER EN CES .•....••••••••••••..•...•......•...•...•...•...•...•....•••.•••.•••••••••••.•••••...••..•...•....•..•.. 21-27

Chapter 21/AII-Way STOP-Controlled Intersections Contents

Versfon 6.0 Page 21.¡
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

LIST OF EXHIBITS

Exhibit 21-1 Analysis Cases for AWSC Intersections 21-2

Exhibit 21-2 Operation Pattems at AWSC Intersections 21-4
Exhibit 21-3 AWSC Configuration: Formulation 1 21-5
Exhibit 21-4 AWSC Configuration: Formulation 2 21-6
Exhibit 21-5 Probability of Degree-of-Conflict Case 21.7
Exhibit 21-6 Degree-of-Conflict Cases for Two-Lane Approaches 21-8
Exhibit 21-7 Degree-of-Conflict Cases for lhree-Lane Approaches 21-8
Exhibit 21-8 LOS Criteria: Motorized Vehide Mode 21-9
Exhibit 21-9 Required Input Data, Potential Data Sources, and Default
Values for AWSC Motor Vehide Analysis 21-11
Exhibit 21-10 AWSC Intersection Methodology 21.12
Exhibit 21-11 Geometry Groups 21-13
Exhibit 21-12 Saturation Headway Adjustments by Geometry Group 21-13
Exhibit 21-13 Probability of Q¡ (Two-Lane Approaches) 21-14
Exhibit 21-14 Probability of Degree-of-Conflict Case: Multilane AWSC
Intersections (Two-Lane Approaches, by Lane) 21-15
Exhibit 21.15 Saturation Headway Values by Case and Geometry Group 21.17
Exhibit 21.16 Probability of Q¡ (Three.Lane Approaches) 21.20

Contents Chaptef 21jAJI-Way STOP-Controlled Intersections

Page 21-ji Version 6.0
 Highway Capadty Manual: A Guide for Multimodal Mobility Analysis

1. INTRODUCTION

AII-way STOP-controlled(AWSC) inlersections are cornmon in the United VOLUME 3: INTERRUPTED FLOW
16. Urban 5treet F~dlities
States. They are characterized by having all approaches control1ed by STOI'signs 17. Urban Street Reli~bility ~nd
without any street having priorily. Streets intersecling at AWSC intersections can ATDM
18. Urb~n 5treet Segments
be public streets or private driveways. This chapler presenls concepts and 19. Sigrlalized lntersections
procedures for analyzing these types of intersections (1). A glossary and lis! of 20. TWSC Intersections
21. AWSC Intersections
symbols, ineluding Ihose used for AWSC inlersections, is provided in Chapler 9. 22. Roundabouts
23. R~mp Terminals ~nd A1tem~tive
The inlerseclion analysis boundaries for an AWSC analysis are assumed lo Intersections
be those of an isolated intersection; that is, no upstream or downstream effects 24. Off.Street Pedestri~n ~nd Bicycle
Facilities
are accounled for in the analysis. The present methodology is limited lo motor
vehieles, although qualitative guidance is provided foc analysis of pedestrians
and bicyelists.

CHAPTER ORGANIZATION
This chapter is ocganizcd into the following sections:
e Seclion 1 (this section) introduces the chapter.
e Section 2 presents basic concepts of Ihe AWSC methodology, ineluding a
descriplion of the phase-patlern concept, general capacity concepts, and
level-of.service (LOS) crHeria.
• Section 3 describes the melhodological details of Ihe proccdurc, ineluding
a step-by-step description of the analysis steps, a discussion of limitations
of the method, and required dala.
e Section 4 addrcsses extensions of Ihe analysis methodology foc motor
vchieles specifically related to analysis of three-Iane AWSC intersection
approaches.
e Section 5 and Section 6, respectively, present pcdestrian and bicyele
evaluation considerations for AWSC intersections.
• Section 7 describes example problems included in Volume 4, suggests
applications for alternalive loo[s, aod provides guidance 00 intcrpreting
analysis rcsults.

RELATED HCM CONTENT

Other HCM content related to this chapter ineludes the following:
• Chapler 32, Stop-Controlled lntersections: SupplementaL provides
example problems for the AWSC methodology and additional
melhodological details.
e Scction M, STOP-Controlled Intersections, in Ihe Plmmiflg Qlld Preliminary
Engineering Aplllicatiofls Cuide lo fhe HCM in Volume 4, provides guidance
on analyzing AWSC iotersections in the context of a planning study.

Chapter 21{AII.Way Srop..Controlled Intersections lntroduction

VelWn6.0 Page 21-1
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

2. CONCEPTS

AWSC íntersections are a type of unsignalízcd intersection that require

drivers on aH approaches to stop at the intersection before proceeding. Because
each driver rnust stop, the decision to proceed into the íntersection ís a function
of traffic conditíons on the other approaches. If no traffic is present on the other
approaches, a driver can proceed irnrnediately after stopping. If there is traffíc on
one or more of the other approaches, a driver proceeds only after determining
that no vehicles are currently in the intersection and that it is the driver's tum to
proceed.
Although giving priority to the driver on the right is a recognized rule in
sorne areas, it is not a good descriptor of actual intersection operations. Drivers
develop a consensus of right-of-way that alternates between the drivers on the
intersection approaches, a consensus that depends primarily on the íntersection
geometry and the arrival patterns at the stop line.
The methodology analyzes each intersection approach separately. The
approach under study is called the subject approach. The opposing approach and
the conflícting approaches create conflícts with vehides on the subject approach.

PHASE PATTERNS
Fíeld observations índícate that standard four-Ieg AWSC intersections
operate in either a two-phase or a four-phase pattern, based prirnarily on the
complexity of the intersection geornetry. Flows are determined by a consensus of
right-of.way that aiternates between the north-south and east-west streams (for
a single-Iane approach) or proceeds in tum to each intersection approach (for a
multilane-approach intersection).
If traffic is present on the subject approach only, vehides depart as rapidly as
individual drivers can safely accelerate into and clear the jntersection. This case
is ilIustrated as Case 1 in Exhibit 21-1.

Exhibit 21-1
Analysis cases for AWSC
Intersections
~ L ~~ L ~ L
I DI I ~I
CMe 1: Vehide(5)on Case 2: \IeIlide!¡ on 5Ubject
--,- Ó
I
case 3: VeNde5 on 5lbJed: I
...tljed apprQadl oruy an:l oppo5irIg approaches a<ld contbdirl\l appr0ad'e5

~~ L ~~ .1- 1

--,-
~I --,-01
case 4: Vehide5 on 5UbJed: case s: VefIidel¡ on al approache5
af1d two od1er approaches

Concepts Olapter 21/A1I-Way Sl"OP-Ccntrolled Inter5ections

Page 21-2 Vet5ion 6.0
 Highway Capacity Manual: A Gujde tor Multimoda/ Mobility Analysis

If traffic is present on the othcr approaches, as well as on the subject

approach, the saturation headway (the time between subsequcnt vehicle
departurcs) on the subject approach will increase somewhat, dcpending on the
degree of conflict that results between the subject approach vehicles and the
vehicles on the other approaches. In Case 2, sorne uncertainty is introduced with
a vehicle on the opposing approach, and thus the saturation headway wiU be
greater than for Case 1. In Case 3, vehicles on one of the conflicting approaches
further restrict the departure rate of vehicles on the subject approach, and the
saturation headway will be longer than for Case 1 or Case 2. In Case 4, two
vehicles are waiting on opposing or conflicting approaches, and saturation
headways are even ¡onger. When vehicles are present on all approaches, as in
Case 5, saturation headways are the ¡ongest of any of the cases because the
potential for conflict between vehicles is greatest. The increasing degree of
potential conflict translates direct!y into longer driver dedsion times and longer
saturation headways. Because no traffic signal controls the stream movement or
allocates the right-of-way to each conflicting traffic stream, the rate of departure
is controIled by the interactions between thc traffic streams.
The CiJpadty of an A WSC
Capadty is defined as the maximum throughput on an approach given the intersection CiJn be detenrIJned
fIow rates on the other intersection approaches. In general, the capadty at an as a function of saturaWn
AWSC intersection can be determined as a function of a few key time-based heildway, departure headway,
and service time.
terms:
• The saturation headway hsí is the time between departures of successive
vehicles on a given approach for a particular case (case 1), as described
aboye, assuming a continuous queue.
• Thc departure headway hd is the average time between departures of
successive vehicles on a given approach accounting for the probability of
each possible case.
• The service time ts is the average time spent by a vehicle in first position
waiting to depart. It is ('qual to the departure headway minus the time it
takes a vehicle to move from second position into first position (the movc-
up time 111).
A two-phase pattern, as shown in Exhibit 21-2(a), is observed at a standard
four~leg AWSC intersection (one approach lane on each leg), when drivers from
opposing approaches enter the intersection at roughly the same time. Sorne
interruption of this pattcrn occurs when there are conflicts between certain
turning maneuvers (such as a northbound left-turning vehicle and a southbound
through vehicle), but generalIy the north-south streams alternate right-of-way
with the east-west streams. A four-phase pattern, as shown in Exhibit 21-2(b),
emerges at multilane four-Ieg intersections, where development of the right-of-
way consensus is more difficult. Here drivers from each approach enter the
intersection together as right-of-way passes from one approach to the next, and
each is scrved in turno A similar three-phase pattern emerges at multilane three-
leg intersections.

Chapter 21/AJI-Wa'l SToP-COntrolled intersections CoI1cepts

V~6.0 Page 21-3
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Exhibit 21-2
Operation Patterns at AWSC Phase 1 Phase 2
Intersections Phase 1 Phase 3

t--I---
1
Phase 2 Phase 4
(a) Two-phase (single-Iane approaches) (b) Four-phase (muttilane approaches)

OEPARTURE HEAOWAY
Two ases for departure
heafways.
The headways of vehicles departing from the subject approach faJl into one
of two cases. lf there are no vehicles on any of the other approaches, drivers on
the subjeet approach can enter the intersection immediately after stopping.
However, if vehicles are waiting on a conflicting approach, a driver from the
subjcct approach must wait for consensus with the next conflicting driver. The
headways between consecutively departing subjcct approach vehicles will be
shorter in the first case than in the second case. Thus, the headway for a
departing subject approach vehicle depends on the degree of conflict experienced
with vehicles on the other intersection approaches. The degree of conflict
increases with two factors: the number of vehicles on the other approaches and
the complexity of the intersection geometry.
Vehide type and tuming Two other factors affect the departure headway of a subject approach
movement affect depi1rture
headway. Tñese effects are vehicle: vehicle type and turning movement. The headway for a heavy vehicle
captured empirically in the will be longer than that for a passenger caro Furthermore, the headway for a left-
method.
turning vehicle will be longer than that for a through vehicle, which in turn will
be longer than that for a right-turning vehicle.
In summary,
1. Standard four-Ieg AWSC intersections operate in either two-phase or
four-phase patterns, based primarily on the complexity of the interseetion
geometry. Flows are determined by a consensus of right-of-way that
alternates between the north-south and east-west streams (for a single-
lane approaeh) or proceeds in turn to each intersection approach (for a
multilane approaeh).
2. The headways between conserutively departing subject approaeh vehicles
depend on the degree of eonflict between these vehicles and the vehicles
on the other intersection approaehes. The degree of conflict is a funetion
of the number of vehicles faeed by the subject approach vehicle and of the
number of lanes on the interseetion approaehes.
3. Thc headway of a subject approaeh vehicle also depends on its vehicle
type and its turniog maneuver (if any).

Concepts Chapter 21jAlI-Way SToP-Controlled Intersections

Page 21-4 Version 6.0
 Highway Capadty Manual: A Guide tor Multimodal Mobility Analysis

CAPACITY CONCEPTS
The capacity model described in more detail here is an expansion of earlier
work (2). The model is described for four increasing!y complex cases: the
intersection of two one-way strects with no turning movements, the intersection
of two two-way streets with no turning movemcnts, a generalized model for
single-lane sites, and a generalized model for multilane sites. The methodology
describcd later in this chapter is an implementation of the latter and most general
case.

IntersecUon of Two One-Way 5treets

The impad of tuming
The first formulation of the model is based on the intersection of two one- movements is considered
way streets, each SToP-controlled. In this basic model, vehicles on either beIow as part of the
generafized models.
approach travel only straight through the intersection, as shmvn in Exhibit 21-3.

Exhibit 21.3
AWSC Configuration:
Formulabon 1

:_ Conflicting

.
: approach

.--------¡:
I t Subject
approach

The sahiration headway for a vehicle assumes one of two values: hs1 is the
saturation hcadway if no vehicle is waiting on the conflicting approach, and ¡¡s2
is the saturation headway if thc conflicting approach is occupied. The departure
headway for \'Chicles on an approach is the expected value of this bivalued
distribution. For the northbound approach, the mean service time is computed
by Equation 21.1.
Equation 21-1
hd N = hS1 (1 - xw) + hs2xw
wherc Xw is the dcgree of utilization of the westbound approach and is equal to
the probability of finding at least one vehicle on that approach. Thus 1 - Xw is the
probability of finding no vchicle on the westbound approach.
By syrnmetry, the mean scrvice time for the westbound approach is given by
Equation 21-2.
Equation 21.2
hdW = hs1(l- XN) + hS2XN
Because thc degree of utilization x is the product of the arrival rate Aand the
mean departure headway lid, the departure headways for each approach can be
exprcssed in terms of the bivalued saturation hcadways and the arrival rates on
each approach, as shown in Equation 21-3 and Equation 21-4.
hs1[1 + Aw(hs2 - hS1)] Equation 21.3
hd N = ---------
, 1 - AN..1-W(hs2 - hS1)2
Equation 21-4
hst[l + AN(hs2 - hst)]
hd W = ---~-~-~-
. 1 - AN'lW(hs2 - hS1)2

O1apter 21/AII-Way $TOP-COntrolled Intersections Concepts

Version 6.0 Page 21-5
Highway Capacity Manual: A Guide for Mulfimodal Mobility Analysis

lntersection of Two Two-Way Streets

In this simplified mode!, the saruration headway for a vehide assumes one of
two values, hs1 or hsv because vehides are again assumed to pass straight
through the intersection, as shown in Exhibit 21-4. The departure headway for
vehides on an approach is the expected value of this bivalued distribution. A
northbound vehide will have a saturation headway of hs1 if the eastbound and
westbound approaches are empty simultaneously. The probability of this event
is the product of the probability of an empty westbound approach and the
probability of an empty eastbound approach. The departure headway for the
northbound vehide is computed with Eguation 21-5.
Equation 21-5
hdN = h,,(l- xE)(l- xw) + h,,[l- (1 - xE)(l- xw))

Exhibit 21.4
AWSl:. Configuratlon:
Fonnulation 2
~
POSi""
approadl ¡
---------
L:-N
: :- Conflictit19 approach
Conflicting approach : ,from right
fromleft_

¡---------r=: t Subject
'pp""",

Unlike Formulation 1, it is not possible to solve directly for the departure

headway in terms of a combination of arrival rates and the bivalued saturation
headways. The departure headway on any approach depends on, or is directly
coupled with, the traffie intensity on the two conflicting approaches. This
eoupling prevents a direct solution. However, it is possible to solve for the
dcparture headway on each approach in an iterative marmer, by using a system
of equations similar in fonn to Equation 21-5.

Generalized Model for Single-Lane Sites

Copacity ÍS determined by an
iterative procedure.
The generalized model is based on five saturation headway values, each
refleeting a different level or degree of oonflict faced by the subject approach
driver. Exhibit 21-5 specifies the conditions for eaeh case and the probability of
occurrenee of each. The probability of occurrence is based on the degree of
utilization on the opposing and conflicting approaches. The essenee of the rnodel,
and its complexity, are evident when one realizes that the traffie intensity on one
approaeh is computcd from its eapacity, which in tum depends on the traffic
intensity on the other approaches. The interdependence of the traffic flow on aH
intersection approaches creates the need for iterative ealculations to obtain stable
estimates of departure headway and service time, and thus eapacity.

Co"",pts Chapter 21/AII-Way STOP<Ontro11ed Intersections

Page 21-6 V~6.0
 Highway Capacity Manual: A Guide far Multimadal Mability Analysis

ADDroach Exhibit 21-5

Degree-of-Conflict Conflicting Conflicting Proba bility of Probabílíty of Degree-of-
ca", OODOsina Riaht Occurrence Conflict Case
L""
1 N N N (1 xo'l(1 xa)(l- xCR)
2 y N N (xQl(I- xa)(1 - xCR)
3 N Y N (1- xQl(xa)(l- xCR)
3 N N Y (1- xQl(1 - xa)(xCR)
4 y N Y (xQl(I- xa)(xCR)
4 y Y N (xQl(xa.)(I- xCR)
4 N Y Y (1 - xQl(xa)(x('Rl
5 y Y Y (xo'l(xa)(xCR)
NOtes: N '" no; y '" ves.

The probability P(C¡) for each degrec-of-conflict case given in Exhibit 21-5
can be computed with Equation 21-6 through Equation 21.10. The degrecs of
utilization on the opposing approach, the conflieting approaeh from the left, and
the conflicting approach fram the right are given by Xo< Xcv and XCIV respectively.
P(C,) = (1 - xo)(l - xeL)(l - XCR) Equation 21-6

P(C,) = (xo)(l - xc¿)(1- XCR) Equation 21-7

P(C,) = (1 - xo)(xc¿)(l - XCR) + (1 - xo)(1- XeL)(XCR) Equation 21-8

P(C.) = (xo)(l - XCL)(XCR) + (xo)(xcL)(l - XCR) + (1 - XO)(XCL)(XCR) Equation 21-9

P(Cs) = (xo)(XC¿)(XCR) Equation 21-10

The departure headway for an approaeh is the expected value of the

saturation headway distribution, computed by Equation 21.11.
s Equation 21.11
h. = Ií=l
P(C,)h,¡

where P(C¡) is the probability of the degree-of-conflict case ej, and hsi is the
saturation headway for that case, given the traffie stream and geometric
conditions of the intersectian approach.
The capadt)' is computed by incrementally ¡ncreasing the volume on the C1p.:JCity is determined by
inaeasing voIume on !he
subjed approach until the degree of utilizatian exceeds 1.0.This flow rate is the subject approadl untif the
maximum possible flow or throughput on the subject approach under the degree of utifiution exceeds
1.0 (i.e., x > 1.0).
conditions used as input to the analysis.

Generalized Model for Multilane Sites

Saturation headways at multilane sites are typically longer than at single-
lane sites, aH other conditions being equal. This situation is the result of two
factors:
• A larger intersection (i.c., grcater number of lanes) requires more travel
time through the intersection, thus increasing the saturation headway;
and
• Additionallanes also result in an increasing degree of conflict with
opposing and conflieting vehicles, increasing driver decision time and thc
saturation headway.

O1apter 21/AII-Way STOP-Controlied Intersections

Concepts
Vers,U? 6.0
Page 21-7
---------------------
Highway Copocity Manual: A Guide for Multimodal Mobility Anolysis

However, sorne movements may nol conflict with c3ch other to the same
degree al multilane sites as at single-lane sites. For example, a northbound
vehicle tuming right may be able to depart simultaneously with ao castbound
through movement ii the twa vehicles are able to occupy separate receiving lanes
whcn departing tú the east. Consequently, in sorne cases, the saturation headway
may be lower at rnultilane sites.
The theory dcscribed aboye propases that the saturation headway is a
funcHon of the diredional movernent oi the vehicle, the vchicle type, and the
degree oí conflict laced by the subject vehicle. This theory is extended here for
multilane sites with respect to the concept of degree of conflict: saturation
headway is affected to a large extent by the number of opposing and conflicting
vehicles faced by the subject driver. For example, in degree-of-conflict Case 2, a
subject vehicle is faced only by a vehicle 00 the opposing approach. At a two-
lane approach intersection, there can be either ooe or two vehides on the
opposing approach. Each degrce-of-conflict case is expanded to consider the
number of vehicles present 00 each of the opposing and conflicting approaches.
The cases are delined in Exhibit 21-6 and Exhibit 21.7 for two-Iane and three-Iane
approaches, respcctively.
For multilane sites, separate saturation headway values are computed for the
number of vehicles faced by the subject vehicle for cach degree.of.conflict case.
This calculation requires a further cxtension of the service time model to account
for the increased number oi subcases. These combinations can be further
subdivided ii a vehicle can be present on any lane on a given approach.

Exhibit 21-6 Approaches with Vehicles No. of Opposing

Degree-of-<:Onflict Cases f()(
Degree.of. Conflicting and Conflicting
con,~~~¡ng
Two-Lane Approaches Conflict Case Onnncinn Left Ri ht Vehicles
1 O
2 ,
, 1 2
3 , 1,2
, ,
4 , , 2,3,4
, ,
5 , , , 3 4 S 6

Exhibit 21-7 ApDroaches with Vehicles No. of Opposing

Degree-of-Confliet Cases for Degree-of~ Conflfcting and Conflicting
Three-Lane Approaches con,~~~ing
Conflict Case Onnnsinn Left Ri ht Vehicles
1
2 , O
1 2 3
,
3 , 1,2,3
, ,
, ,
4
, , 2,3,4,5,6

5 , , , 3,4, S, 6, 7, 8, 9

Concepts Chapter 21{AII-Way STOP-ControUed Intersections

Page 21-8 Ver.Wn 6.0
 Highway Capacity Manual: A Gujde for Multjmodal Mobility Ana!ysjs

The probability of a vehicle being al the stop line in a given lane is x, the
degree of utilization. The product of the six degrees of saturation, encompassing
each of the six lanes on the opposing or conflictíng approaches (two Janes on the
opposing approach and two lanes on each of the conflicting approachcs), gives
the probability of any particular combination occurring.
The iterative proccdure to compute the departure hcadways and capacities
for cach approach as a function of the departure headways on the other
approaches is thc same as described carlier. However, thc additional subcases
c1early increase the complexity of this computation.

LEVEL-OF-SERVICE CRlTERIA
LOS critena for AWSC intersections are given in Exhibit 21.8. As the exhibit
notes, LOS F is assigned if the volume-to-capacity (v/c) ratio of a lane exceeds 1.0,
rcgardless of the control delay. For assessment of LOS at the approach and
intersection leveIs, LOS is based solely on control delay.

lOS by Volume-to-Capacity RatiQ~ Exhibit 21-8

Control Delay
v/c'5.1.0 v/e>1.0 LOS Oiteria: Motorized
s veh
Vehide Mode
0-10 A F
> 10--15 B F
>15-25 e F
>25-35 D F
>35-50 E F
>SO F F
Note: • fQr approaches and intersectionwide assessment, lOS IS defined
solely by control delay.

Chapter 21/AII-Way 5T~-COntrQlled Intersections Concepts

Version 6.0 Page21-9
---------------------
Highway Copacity Manual: A Guide for Mulfimodol Mobility Analysis

3. MOTORIZED VEHICLE CORE METHODOLOGY

SCOPE OF THE METHOOOLOGY

This section facuses on the operation oi AWSC intersections. This version of

the AWSC intersection analysis procedures is primarily a result Di studies
conducted in National Cooperative Highway Research Program Praject 3-46 (1).

Spatial and Temporal Limits

The methodology assumes the AWSC intersection under investigation is
isolated. If interaction effects (e.g., queue spillback, demand starvation) are likely
between the subject AWSC intersection and other intersections, the use oi
alternative too15 may resuIt in a more accurate analysis. Analysis boundaries
may also ¡ndude different demand scenarios related to time oi day or to different
deve!opment scenarios that produce different demand flow rates.
The recommended length of the analysis period is the HCM standard of 15
min (although longer periods can be examined).

limitations of the Methodology

The methodologies in this chapter apply to iso!ated AWSC intersections with
up to three lanes on each approach. They do not account for interaction effects
with other intersections. The methodologies do not apply to AWSC intersections
with more than four approaches. In addition, the effect of conflicting pedestrians
on motor vehides is not considered in this procedure. Conflicting pedestrian
movements are likely to increase the saturation headway of affected vehicular
movements, but the magnitude of this effect is unknown as of the publication of
this edition of the HCM.

Use of Alternative Tools

Except for the effects of interaction with other interseetions, the limitations of
the methodology stated aboye have minimal potential to be addressed by
altemative tools. Therefore, insufficient experience with altemative tools is
availab!e as of the time of publication of this edition of the HCM to support the
development of useful guidance for their application to AWSC interseetions.

REQUIRED INPUT OATA ANO SOURCES

Exhibit 21.91ists the infonnation necessary to apply the motor vehide
methodology and suggests potential sources for obtaining these data. It also
suggests defau!t values for use when intersection~specific information is not
available.
A comprehensive presentation of potential default values for interrupted-
fIow facilities is available elsewhere (3). These defauIts cover the key characteristics
of peak hour factor and percentage of heavy vehides. Recommendations are
based on geographic region, population, and time of day. AIl general default
values for interruptcd-flow facilities may be applied to analysis of AWSC
intersections in the absence of field data or projections of conditions.

Motorized Vehlcle Core Methodology O1apter 21/AlI-Way SrOP-Cootrolled Intersections

Page 21-10 Version 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Suggested Exhibit 21.9

Default Value Required Input Data, Potential
Required Data and Units
Data Sources, and Default
Values fer AWSC Motor
Number and contiguration of lanes of each approach Must be provided Vehicle Analysis

DemandData
Hourly tuming movement demalld volume (veh/h) AND
peak hour factor
Field data, modeling Must be provided
OR
Hourly tuming movement demand flow rate (veh/h)
Analysis penod length (min) 5et by analyst 15 min (0.25 h)
Peak hour factor (decimal) Fleld data 0.92
Heavy-vehide percentage (%) Fleld data 3%

COMPUTATIONAL STEPS
The AWSC intersection methodology for the motor vehicle mode is applied
through a series of steps that relate to input data, saturation headways, departure
headways, service time, capacity, and LOS. The steps are listed in Exhibit 21.10.

5tep 1: Convert Movement Demand Volumes to Flow Rates

Flow rates for each turning movement at the intersection must be converted
from hourly volumes in \'Chicles per hour to peak IS-min flow rates in vehicles
per hour as given in Equation 21-12.
Vi
v---
1 - PHF
Equation 21-12

where
Vj demand flow rate far movement i(veh/h),
V¡ demand volume far movement ¡(veh/h), and

PHF intersection peak hour factor.

If PHF is used, a Single
If peak hour factors are used, a single peak hour factor fur the entire Intersedioflwide PHF should be used
intersection is generally preferred to decrease the Iikelihood of creating demand rather than movement-specirlc {)('
approach-speciñc PHFs. If IndIvidual
scenarios with conflicting volumes that are disproportionate to the actual approaches {)(' movements peak at
volumes during the lS-min analysis periodo If peak hour factors far each different times, a series of 15-mlf'l
analysts periods that encompass the
individual approach or movement are used, they are Iikely to generate demand
peakJng should be consfdered.
volumes from one lS-min period that are in apparent conflict with demand
volumes from another lS-min period, but in reality these peak volumes do not T11elJ5e of a peak 15-min traffic
occur at the same time. Furthermore, to determine individual approach or CO¡.Jntmultiplied by tour is preferred
for exlsting conditions when traffic
movement peak hour factors, actual IS-min count data are like!}'available, counts are avallabfe. T11euse of a l-h
permitting the determination of actual IS-min demand and avoiding the need to demiJnd voIume divided by a peak
hour factor is preferreé with
use a peak hour factor. In the event individual approaches ar movements are projected voIumes or with projected
known to havc substantially different peaking characteristics or peak during votumes that have been added te
Oirrent voIumes.
different IS-min periods within the hour, a series of lS-min analysis periods that
encompass the peaking should be considered instead of a single anal}'sis period
using a single peak hour factor for the intersection.

O1apter 21/AJI-Way ST()I>oContro11ed

Intersections Motorized Vehicle Core Methodology
VersiOn 6.0 Page 21-11
 Highway Capacity Manual: A Guide tor Multimodal Mobility Analysis

Step 5: Determine Initial Departure Headway

The process oE determining departure headways (and thus service times) for
each of the lanes on each of the approaches is iterative. For the first iteration, an
initial departure headway of 3.2 s should be assumed. For subsequent iterations,
the cakulated values of departure headway from the previous iteration should
be used if the calculation has not converged (see Step 11).

Step 6: Calculate I"itial Degree of Utilization

By using the lane f10w rates from Step 2 and the assumed initial departure
headway from Step 5, the initial degree of utilization x is computed with
Equation 21-14. lE it is not the final iteration, and the degree oE utilization exceeds
1, then the degree oE utilization should be reset to 1.
vhd
Equation 21.14
X = 3,600
Step 7: Compute Probability States
The probability state oE each combination i is found with Equation 21.15.
Equation 21.15 P(i) = n J
peal)

where n represents the product of a series oE tenns, and

j = 01 (opposing approach, Lane 1), 02 (opposing approach, Lane 2), CLl
(conflicting left approach, Lane 1), CL2 (conflicting left approach, Lane
2), CRl (conflicting right approach, Lane 1), and CR2 (conflicting right
approach, Lane 2) for a two-lane, two-way AWSC intersection;
P(a,) probability of ay computed on the basis of Exhibit 21.13, where Vi is the
lane flow rate; and
al 1 (indicating a vehicle present in the lane) oc O(indicating no vehicle
prescnt in the lane) (values of Q¡ for each lane in each combination ¡are
listed in Exhibit 21-14).

Exhibit 21-13
Probability of aj (Two-Lane al
1
'Í'
O
P(ap
O
_es)
O O 1
1 > O Xj
O >0 l-X
i
Note: xis the degree af utITizationdefu1edin Equation 21-14.

Tables for th,-ee.tane Exhibit 21.14 provides the 64 possible combinations when alternative lane
approaches are given In
Chapter 32, SroP-ControlJed occupancies are considered Eor two-lane approaches. A 1 indicates a vehicle is in
¡ntersections: Supplemental. the lane, and a Oindicates a vehicIe is not in the lane. A similar table for three
lanes on each approach is provided in Chapter 32 in Volume 4.

Motorlzed Vehicle (.ore MethodoIogy Chapter 2l/AlI-Way STIY-COntrolled Intersections

Page 21-14 Versioo 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Conflicting Left Conflictlng Rlght Exhibit 21-14

DOC No.of 0Rmina Aooroach Aoorpach ADOroach Probability of Degree-of.
¡ ca•• Vehicles Lane 1 Lane 2 Lane 1 Lane 2 Lane 1 Lane 2 Conflict Case: Multilane AWSC
1 1 o o o o o o o Intersections (Two-Lane
2 1 o o o o o Approaches, by Lane)
1 o o o o
o 1
-1-
4
2
2 1 1 o o o o
o o o
,5
1
o
o
o
o
1
o 1
o
o o
o
7
3
o o o 1
o o o o
+
,
10
2
o
o
o
o
o
1
o
1
o
o
1
1
o
1
11 o o o 1 o 1
12 o o 1 o o 1
13 o o 1 o 1 o
14 o o o 1 1 o
15 o 1 o 1 o o
16 2
1 o 1 o o o
17 o 1 o o 1 o
18 1 o o 1 o o
19 o 1 1 o o o
20 o 1 o o o 1
2l 1 o o o 1 o
1 o o o o 1
-4-
23 o o o 1 1 1
24 4 o o 1 1 o 1
25 o o 1 1 1 o
1 o 1 1 o o
"
27
28
1
1
1
1
1
o
o
o
o
1
o
o
3 o o o 1
29 1 1
30 o 1 1 1 o o
31 1 o o o 1 1
32 o o 1 o 1 1
33 1 1 o 1 o o
-ª-"
35

37
4
o
o
1

1
1
1
o
1
o
o
1
1
o
o
1
1
1
1
1
o
1
1
1
o
38 o 1 o 1 o 1
39 1 o o 1 1 o
o 1 1 o 1 o
""
41 o 1 o 1 1 o
3 o o
42 o 1 1 1
43 1 o 1 o o 1
44 1 o o 1 o 1
1 o 1 o 1 o
~ o o 1 1 1
•••
47
1
o 1 1 1 1 o
48 o 1 1 1 o 1
4' 1 o 1 o 1 1
50 1 o 1 1 1 o
51 5 4
o 1 o 1 1 1
1 1 1 o o 1
"53
54 o
1 o
1
1
1 o
1 o
1
1
1
55 1 1 o 1 1 o
56 1 1 o 1 o 1
1 1 1 o 1 o
--¥-
58 1 o 1 1 1 1
59 1 1 o 1 1 1
60
5
1 1 1 o 1 1
61 o 1 1 1 1 1
1 1 1 1 1 o
~" , 1 1 1 1 o 1
64 1 1 1 1 1 1
Notes: DOC = degree-of-eonnict; No. of Vehides = total number of vehicles 00 opp::lSiog aod connicting
approaches.

Chapter 21/AII-Way STOP-Controlled Intersections Motorized Vehicle Core Methodology

VersiOn 6.0 Page 21-15
 Highway Capacify Manual: A Guide for Multimodal MobiUty Analysis

5tep 8: Compute Probability Adjustment Factors

The probability adjustment is computed with Eguation 21-16 throllgh
Eqllation 21-20 to account for the serial correlation in the previous probability
computation. First, the probability of each degree-of-conflict case must be
determined (assllming the 64 cases presented in Exhibit 21-14).
Equation 21-16
P(C,) = P(l)

Equation 21-17
P(C,) = L•
1=2
P(i)

10
Equation 21-18 P(C,) = L P(i)
i=S

Equation 21-19 P(C.) = L"

i=11
P(i)

P(C,) = L••
í=38
P(i)

The probability adjustment factors are then computed with Eguation 21-21
through Equation 21-25.
Equation 21-21
AdjP(l) + 2P(C,) + 3P(C.) + 4P(C,)J/l
= a[P(C,)
Equation 21-22 AdjP(2) through AdjP(4) = a[P(C,) + 2P(C.) + 3P(C,) - P(C,)]/3
Equation 21-23 AdjP(S) through AdjP(10) = a[P(C.) + 2P(C,) - 3P(C,)1/6
Equation 21-24 AdjP(ll) through AdjP(37) = a[P(C,) - 6P(C.))/27
Equation 21-25 AdjP(38) through AdjP(64) = -a[10P(C,)J/27
where aequals 0.01 (or 0.00 if correlation among saturation headways is not
taken into account).
111eadjllsted probability P'(l) for each combination is simply the sum of P(i)
and AdjP(¡), as given by Eguation 21-26.
Equation 21-26 P'(i) = P(i) + AdjP(i)
5tep 9: Compute Saturation Headways
111esaturation headway hsi is the sum of the base saturation headway as
presentcd in Exhibit 21-15 and the saturation headway adjustment factor from
5tep 4.lt is shown in Equation 21-27.
Equation 21-27
hsi = hhase + hadj

Motorized Vehicle CUre Methodology Chapter 21/AII.way STOP-Control1ed Intersections

Page 21-16 Version6.0
 Highway Capacity Manual: A Gujde for Multimadal Mobility Analysis

Base Saturation Headway, 4._ (s) Exhibit 21-15

No.of Group Group Group Group Group Group Group Group SaturationHeadwayValuesby
Ca",
1
Vehicles
O
1
3.9
2
3.9
3a
4.0
3b
4.3
4a
4.0
4b
4.5
S
4.5
•
4.5
case andGeometryGroup

1 4.7 4.7 4.8 5.1 4.' 5.3 5.0 '.0

2 2 6.2 6.8
>3 7.4
1 5.8 5.8 5.9 6.2 5.9 6.4 6.4 6.6
3 2 7.2 7.3
>3 7.8
2 7.0 7.0 7.1 7.4 7.1 7.' 7.6 8.1
3 7.8 8.7
4
4 9.0 9.6
>5 12.3
3 9.6 9.' 9.7 10.0 9.7 10.2 9.7 10.0
4 9.7 11.1
5
5 10.0 11.4
>6 11.5 13.3

5tep 10: Compute Departure Headways

The departurc headway hd of the lane is the expected value of the saturation
headway distribution as given by Equation 21-28.

hd ~ I"
i:: 1
P'(i)h"
Equation 21.28

where j represents each combination of the five degree-of-conflict cases, and hsi is
the saturation headway for that combination.

5tep 11: Check for Convergence

The calculated values of hd are checked against the initial values assumed for
hd' If the values change by more than 0.1 s (or a more precise measure of
convergence), 5teps 5 through 10 are repcated until the values of departure
headway for each lane do not change significantly.

5tep 12: Compute Capacity

The capacity of each lane in a subject approach is computed under the Cilpadty iS estJmated for él
stated set of opposing and
assumption that the flows 00 the opposing and conflicting approaches are conflicting voIumes.
constant. The given flow rate on the subject lane is increased, and the departure
headways are computed for each lane on each approach until the degree of
utilization for the subject lane reaches 1. When this degree of utilization occurs,
the final value of the subject I.me flow rate is the maximum possible throughput
or capacity of this laoe. The following stcps iIIustrate this process.

Step 12a: Se/ect a Subject Approach and Step 12b: Estab/ish a Tria/ Va/ume for
the Subject Approach
Jf the degree of utilization calculated for the subject approach is less than 1,
then the trial volume for the subject approach should be increased. lf the degree
of utilization ealculated for the subject approach is greater than 1, then the trial
volume for the subject approach should be decreased.

Chapter21/AII-Way
STOP-Controlled
Intersecbons MotorizedVehicleCoreMethodology
Vef"SiofJ 6.0 Page 21-17
Highway Capacity Manual: A Guide tar Multimadal Mobility Analysis

5tep 12c: Compute !he Degree ofUtilization Using 5teps 5 Through 11 and 5tep
12d: Check the Degree of Utilization x
If the calculated degree of utilization x is not 1, returo to Step 12b and use a
different trial volume. When the degree of utilization equals 1, the trial volume
that was selected is the capacity of the subject approach.

5tep 12e: Repeat 5teps 12a Through 12d for the Other Approaches

Step 13: Compute Service Times

The service time ts required to calculate control delay is computed on the
basis of the final calculated departure headway and the move-up time with
Equation 21-29.
Equation 21.29 tS=hd-m
where hd is the departure headway, and m is the move-up time (2.0 s for
Geometry Groups 1 through 4; 2.3 s for Geometry Groups 5 and 6).

Step 14: Compute Control Delay and Determine LOS tor Each lane
The delay experienced by a motorist is related to factors such as control,
geometrics, traffic, and inddents. Control delay is the difference between the
travel time that is actually experienced and the reference travel time that would
result during conditions in the absence of traffic controlar conflicting traffic.
Equalion 21-30 can be used to compute control delay foc each lane.

hdx
Equation 21.30 d ~ t, + 900T (x - 1) + (x - 1)' + 450T + 5
[
where
d average control delay (s/veh),
x = Vhd/3,600 = degree of utilization (unitless),
ts service time (s),
hd departure headway (s), and
T length of analysis period (h).
The LOS for each approach and the LOS for the intersection are determined
by using the computed values of control delay, with Exhibit 21-8.

Step 15: Compute Control Delay and Determine LOS tor Each Approach
and the Intersection
The control delay for an approach is calculated by computing a weighted
average of the delay for each lane on the approach, weighted by the volume in
each lane. The calculation is shown in Equation 21-31.
"dv
Equation 21.31 d ~-"-'-'
a LV¡

Motorized Vehide (ore Methodology Chapter 21/A1I-Way SroP-Controlled Intersections

Page 21-18 VeI3'ion 6.0
 Highway Capacity Manual: A Guide for Mulfimoda/ Mobility Analysis

where
d. control deJay for the approach (s/veh),
d¡ control delay for lane i (s/veh), and
V¡ flow rate for lane i (veh/h).
The control delay for the interscction as a whole is similarly calculated by
computing a weighted average oí the delay for each approach, weighted by the
volume on cach approach.lt is shown in Equation 21-32.
¿dava Equation 21-32
dinteT1iectinn = ¿ va
where
control delay fOIthe entire intcrsection (s/veh),

d. control delay for approach a (s/veh), and

v. flow rate for approach a (veh/h).

The LOS for each approach and for the intcrscction are determined by using
the computcd values of control delay, with Exhibit 21-8.

Step 16: Compute Queue lengths

Research (4) has dctermined that the methodology for predicting queues at
TWSC interscctions can be applied to AWSC intersections. As such, the mean
queue length is computed as thc product of the average delay per vehide and the
flow rate lor the movemcnt of interest.
Equation 21-33 can be used to calculate the 95th percentile queue for each
approach lane.

900T [ Equation 21-33

Q9S~-- (x-1)+
hd

where
Q9S 95th pcrcentile queue (veh),
x = vhd/3,600 = degree of utilization (unitless),
hd departure headway (s), and
T length of analysis period (h).

Chapter 21jAII-Way STOP-Controlled Intersections Motorized Vehide Core Methodology

VersiOn 6.0 Page 21-19
Highway Capacify Manual: A Guide for Multimodal Mobility Analysis

4. EXTENSION TO THE MOTORIZED VEHICLE

METHODOLOGY

INTRODUCTION
This seetion provides details for a procedure to analyze three-Iane
approaehes at AWSC intersections. This procedure is an extension oí the
methodology described in the previous section, and the same general eapaeity
eoneepts apply.

REPLACEMENT STEPS FOR THREE-LANE AWSC INTERSECTION

APPROACHES
The methodology for three-Iane AWSC approaehes is fundamentally the
same as for two-Iane approaehes. The process used to determine geometry
groups (Step 3) will typieally result in Geometry Group 5 or 6 (per Exhibit 21-11),
whieh affeets the saturation headway adjustmcnts in Step 4. Steps 7 and 8
inerease in eomplexity to accornrnodate three-Iane approaehes and are replaeed
with the following steps. The remaining steps remain as presented in Seetion 3.
Step 7: Compute Probability States
The probability state of each combination j is found with Equation 21-34.

Equation 21-34 P(O = n J

pral)

where n represents the produet of a series of terms, and

j = 01 (opposing approach, Lane 1), 02 (opposing approach, Lane 2), 03
(opposing approoeh, Lane 3), CLl (eonflieting left approach, Lane 1),
CL2 (conflicting left approaeh, Lane 2), C1..3(eonflicting left approach,
Lane 3), CRl (conflicting right approaeh, Lane 1), CR2 (eonflicting
right approach, Lane 2), and CR3 (eonflicting right approaeh, Lane 3)
for a three-lane, two-way AWSC intersection;
P(a) probability of ay eomputed on the basis of Exhibit 21-16, where Viis the
Jane f10w rate; and
aj 1 (indicating a vehide present in the lane) or O (indicating no vehide
present in the lane) (values of ni for eaeh lane in eaeh eombination j are
listed in Exhibit 32.15 in Chapter 32, SlOP-Controlled Intersections:
Supplemental).

Exhibit 21-16 F(ap

al VI
Probability of al (Three-Lane
Approaches)
1
• •• •
1
1 >. XI

Note;
• >.
)( is the degree of utilizaban defined in Equatiorl21-14.
1-XI

Extension ro the Motorized Vehicle MethOdology O1apter 21jAII-Way STOP-Controlled IntersectiOfls

Page 21-20 Ve~6.0
 Highway Capacity Manual: A Gujde tor Multimodal Mobility Analysis

Tables for th,-ee.-Iane

For three-Iane AWSC approaches, the number oí possible combinations oí approaches are gNen in
degree-oí-conflict cases when alternative lane occupandes are considered Chapter 32, SroP-Controlled
Intersections: Supptemental
increases from 64 cases to 512 cases. These 512 cases are presented in tabular
form in Exhibit 32-15 in Chapter 32, STOP-Controlled Intersections:
Supplemental.

Step 8: Compute Probability Adjustment Factors

The probability adjustment is computed with Eguation 21-35 through
Eguation 21-39 to account for the serial correlation in the previous probability
computation. First, the probability of each degree-of-conflict case must be
determined (assuming the 512 cases presented in Exhibit 32-15 in Chapter 32,
STOI'-Controlled Intersections: Supplemental).
P(C,) = P(l) Equation 21-35
8

P(C,) = I
i=2
P(i) Equation 21-36

P(C,) = I"
i=9
P(i) Equation 21-37

'"
P(C,) = I
i=Z3
P(i) fquation 21-38

su
P(Cs) = I
i=170
P(i) Equation 21-39

The probability adjustment factors are then computed with Equation 21-40
through Eguation 21-44.
AdjP(l) + ZP(C,) + 3P(C,) + 4P(Cs)]/1
= a[P(C,) Equation 21-40

AdjP(Z) through AdjP(8) = a[P(C,) + ZP(C,) + 3P(Cs) - P(C,)]/7 Equation 21-41

AdjP(9) through AdjP(Z2) = a[P(C,) + 2P(Cs) - 3P(C,)]/14

AdjP(23) through AdjP(169) = a[P(Cs) - 6P(C,)]/147 Equation 21-43

AdjP(170) = -a[10P(Cs)]/343
through AdjP(S12) fquation 21-44

where aeguals 0.01 (or 0.00 if correlation among saturation headways is not
taken into account).
The adjusted probability P'(i) for each combination is simply the sum of P(i)
and AdjP(i), as given by Eguation 21-45.
P'(i) = P(i) + AdjP(i) Equation 21-45

Chapter 21jAI1-Way STOP-COntrolled Intersections Extension to tI1e Motorized Vehicle Methodology

Version 6.0 Page 21-21
Highway Copacify Manual: A Guide for Multimodal Mobility Analysis

5. PEDESTRIAN MODE

Di1ta CDlfection and res&NdI

are needed ro determine an
Applying the LOS procedures used to determine pedestrian delay at TWSC
appropriate LOS methodoIogy interscctions to AWSC intersections does not produce intuitive or usabJe results.
for pedestrians atA WSC
intersections.
The TWSC deJay calculations apply only for crossings where conflicting traffic is
not STOP-controlled(i.e., pedestrians crossing the major street at a TWSC
intersection). Approaches where conflicting traffic is SfOP-controlled (Le.,
pedestrians crossing the minor street at a TWSC intersection) are assumed to
result in negligible delay for pedestrians, as vehicles are required to stop and
wait for conflicting vehicle and pedestrian traffic before proceeding.
Consequently, applying the TWSC methodology to pedestrians at AWSC
intersections results in negligible delay for all pedestrians at all approaches. The
reality of AWSC intersection operations for pedestrians is much different,
however, and generally resuIts in at least sorne deJay for pedestrians. The
amount of delay incurred will depend on the operating and geometric
charaderistics of the intersection in question. Although no quantitative
methodology accounting for these factors is available, several of the most
important factors are discusscd qualitatively below.
The operational characteristics of AWSC intersections for pedestrians largely
dcpend 00 driver behavior. In most cases, drivers are legally required to yield to
pedestrians crossing or preparing to cross AWSC intersections. Howevcr, it
should be cxpected that operations differ significantly dcpcnding on cnforccmmt
levels, region oC the country, and location (e.g., urban, suburban, or rural).

VEHICULAR VOLUMES
At intersections with relatively low vehicular volumes, there are generally no
standing queues oC vehicles at AWSC approaches. lo these cases, pedestrians
atternpting to cross an approach oC the intersedion wiII typically experience Httle
or no delay as they will be able to proceed almost immediately after reaching the
intersection.
At AWSC intersections with higher vehicular volumes, there are typically
standing queues of motor vehieles on each approach. These intersedions operate
in a two-phase or four-phase sequence, as described aboye and depicted in
Exhibit 2].2. In these situations, the arrival of a pedestrian does not typically
disrupt the normal phase operations of the intersection. Rather, the pedestrian is
often forced to wait until the phase arrivcs for vehicles in the approach moving
adjacent to the pedestrian.
Under a scenario in which the intersection fundions under the operations
described aboye for pedestrians, average pedestrian delay might be expected to
be half the time needed to eyele through aH phases for the particular intersection,
assuming random arrival of pedestrians. However, several other factors rnay
affect pedestriao delay and operations at AWSC intersections, as described
below.

Pedestrian Mode Chapter 21jAU-Way SToP-Controlled Intersectioos

Page 21-22 Vef5ion6.0
 Hjghway Capacity Manual; A Gujde tor Multimodal Mobility Analysis

NUMBER OF APPROACH LANES

As the number of approach lanes at AWSC intersections increases,
pedestrian crossing distance increases proportionally, resulting in significantly
longer pedestrian crossing times compared with single-Iane intersections. In
addition, vehicles already in the intersection or about to enter the intersection
take longer to complete their movement. As a result, pcdestrians at multilane
AWSC intersections may wait longer before taking their turn to cross.

PROPORTION OF TURNING TRAFFIC

The ability of a pedestrian to cross at an AWSC intersection may also depend
on the proportion of through motor vehicle traffic to turning motor vchicle
traffic. As described aboye, pedestrians may often cross during the phase in
which adjacent motor vehicle traffic traverses the intersection. Howevcr, when
an adjacent motor vehicle is turning, that vehicle will conflict with pedestrians
attempting to cross. Because of thc additional conflicts with pedestrians created
by turning vehicles at AWSC intcrscctions, pedestrian delay may be expccted to
rise as the proportion of turning ,'ehicles increases, similar to the effect that
turning proportion has on vehicular dc1ay.

PEOESTRIAN VOLUMES
Under rnost circumstances, there is adequate capacity for aH pedestrians
queued for a given movernent at an AWSC intersection to cross simultaneously
with adjaeent motor vehicle traffic. Howevcr, in locations with very high
pedestrian volurnes, this may not be the case. The total pedestrian eapaeity oí a
particular AWSC interseetion phase is Iimited by both the width of the crosswalk
(how many pedestrians can eross simultaneously) and driver behavior.
In situations in which not all queued pedestrians may eross during a
particular phase, pedestrian delay will inerease, as sorne pedestrians will be
forced to wait through an additional cycle of intersection phases before crossing.
However, pedestrian volumes in this range are unlikely to occur often; rather,
intersections with pedestrian volumes high enough to cause significant delay are
typically signalized.

Chapter 21/AII-Way STOP-COntrolled Intersections Pedestrian Mode

Version 6.0 Page 21-23
Highway Capacity Manual: A Guide for Multimoda/ Mobility Ana/ysis

6. BICYCLE MODE

The proeedures deseribed to estimate motor vehicle delay can be applied to

bieyeles that queue with motor vehieles on AWSC approaehes. However,
bicyeles differ from motor vehieles in that thcy do not qucue Iinearly at STOP
signs. lnstead, multiple bicyeles often cross at the same time as the adjacent
vehicular traffic stream. This phenomenon has not been researehed as of the time
of publication of this edition of the HCM and is not explicitly induded in the
methodology.
On an AWSC approach that provides a bieydc lane, bieycle delay will be
signifieantly different and, in general, lower than motor vehide deJay. The
exception is bieycles intending to tum left; those cyclists will typieally queue
with motor vehides. Where bieyde lanes are available, bicycles are able to move
unimpcdcd until reaching thc stop line, as the bike lane allows the cydist to pass
any queued motor vehicles on the right. In this situation, bicydes will still incur
delay upon reaehing the intersection.
In most cases, bicydes wiII be ablc to travel through the interseetion
concurrently with adjacent motor vehicle traffie. This practice, in cffect, results in
multilane operations, with the bike lane serving as the curb lane, meaning that
bicycles will be delayed from the time of arrival at thc interseetion until the
adjaecnt motor vehicle phase oecurs. As noted aboye, multiplc bicyclcs will
likely be able to cross simultaneously through the intersection. Finally, even
where bicycle lanes are not available, many cydists still pass queued motor
vehicles on the right; this results in lower effeetivc bicycle delay compared wilh
motor vehicle delay.

Bicyde Mode O1apter 21fAII-Way STOP-Controlled Intersections

Page 21-24 Versicn 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

7. APPLICATlONS

TYPES OF ANALYSIS
The methodology of this chapter can be used in three types of analysis:
operational analysis, design analysis, and planning and preliminary engineering
analysis.

Operational Analysis
The methodology is most easily applicd in the operational analysis mode. In T7Je operaticnaf ana/ySiS
methodology for A WSC
operational analysis, aH traffic and geometric characteristics of the analysis intersedions can also be used
segment must be spedfied, including analysis-hour demand volumes for each for design analysis and
p/iJnning and prelimfnary
turning movement (vehicles per hour), heavy-vehicle percentages for each engineering analysis.
approach, peak hour factor for aHdemand volumes, and lane configuration. The
outputs of an opcrational analysis are estimates of capacity, control delay, and
queuing. The steps of the methodology, describcd in Section 3, are followcd
directly without modification.

Design Analysis
The operational analysis described earHer in this chapter can be used for
design purposes by using a given set of traffic £Iowdata and iteratively
dctermining the numbcr and configuration of lanes that would be required to
produce a given LOS or other desired performance measures.

Planning and Preliminary Engineering Analysis

The operational analysis method describcd carlier in this chapter provides a
detailed procedure for evaluating the performance of an AWSC intersection. To
estimatc LOS for a future time horizon, a planning analysis bascd on the
opcrational mcthod is used. The planning method uses all the gcometric aod
traffic flow data requircd for an operational analysis, and the computations are
identical. However, input variables for heavy-vehicle percentage and peak hour
factor are typically cstimated (or defaults used) when planning applications are
performcd.

EXAMPLE PROBLEMS
Section 5 of Chapter 32, STOP-ControlledIntersections: Supplemental,
providcs two example problems that iIlustrate each of the computational steps
involved in applying the motor vehicle method:
1. Analyze a single-Iane, threc-Ieg, AWSC intersection; and
2. Analyze a multilane, four-Icg, AWSC intersection.

EXAMPLE RESULTS
The computations discussed in this chapter result in the estimation of control
delay and LOS for each lane, for each approach, and for the entire intersection.
When capacities are calrulated with the iterative method dcscribed in this
chapter, they also produce a voiume-to-capacity ratio for cach lanc. This section
provides sorne uscful interpretations of these performance rneasures.

Chapter 21jAII-Way STOP-controlled Intersections Applications

Version 6.0 Page 21-25
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Level of SelVice
In general, LOS indicates the general acceptability of deIay to drivers. In this
regard, it should be remembered that what might be acceptable in a large city is
not necessarily acceptable in a smaller city or rural area.
As with other intersection types and controls, intersection LOS must be
interpreted with caunon. It can suggest acceptable operanon of the intersection
when, in reality, certain lanes or approaches (particularly those with lower
volumes) are operating at an unacceptable LOS but are masked at the
intersection level by the acceptable performance oí higher-volume lanes or
approaches. The analyst should verify that each lane or approach is providing
acceptable operation and consider reporting the LOS for the poorest-performing
lane or approach as a means oí providing context to the interpretation of
intersection LOS.

Volume-to-Capacity Ratio
The interpretation of volume-to-capacity ratios for AWSC intersections
requires care due to the interdependence of the movements at the intersection.
As discussed in the calculation of capacity in Step 12 of the methodology, the
capacity of a subject approach is dependent on the performance of adjacent and
opposing approaches, each of which depend on each other and the subject
approach. As a result, unlike other procedures in which capacity is estimated
directIy, for AWSC intersections, capacity is estimated indirectly by holding the
adjacent and opposing f10ws constant and loading the subject approach to the
point of failure (a degree of utilization of 1.0). In addition, the degree of
utilization, x, is used in the deJay or queue equations rather than the capacity.
In general, a vo!ume-to-capacity ratio greater than 1.0 is an indication of
actual or potential breakdown. Assuming turning movement volumes are fixed,
improvements that might be considered indude the following:
• Basic changes in geometry (Le., change in the number ar use of lanes). The
addition of lanes to an AWSC intersection approach typically changes the
geometric groups for all movements with a resulting increase in
departure headways, so a change to the subject approach to improve its
performance may reduce the performance of othcr approaches; and
• Conversion to another typc oí intersection or control, or both (e.g., signal
control or a roundabout).
Local guidelines should be consulted before potential improvements are
developed.

Applications Chapter 21/AJI-Way STop-Controlled Intersections

Page 21-26 VetSion6.0
 Highway Capacity Manual; A Guide for Multimodal Mobility Analysis

8. REFERENCES

SOrne of these references iJfe

1. Kyte, M., Z. Tian, Z. Mir, Z. Hameedmansoor, W. Kittelson, M. Vandehcy, B. aVilllitble in the Techniaf
Robinson, W. BrUon, L. Bondzio, N. Wu, and R. Troutbeck. NCHRP Web RefererK:e Ubrary in VoIume 4.
Document 6: Capacity and Leve! olService at Ullsignalized Intersectiolls: filial
Report, VO/rlme 2 -AIl-Way Stop-Controlled lntersections. Transportation
Research Board, National Research CounciL Washington, D.C., April1996.
2. Richardson, A. J. A Delay Model for Multiway Stop-Sign Intersections. In
Transportation Research Record 1112, Transportation Research Board, NahonaI
Research Council, Washington, D.C., 1987, pp. 107-114.
3. Zegeer, J. D., M. A. Vandehey, M. Blogg, K. Nguyen, and M. Ereti. NCHRP
Report 599: DelauU Va/ues lor Highway Capacity mzd Leve! olService Analyses.
Transportation Research Board of the National Academics, Washington,
D.C., 2008.
4. Tian, Z., and M. Kyte. Queue Length Models for All-Way Stop-Controlled
Intersections. In Transportation Researc/¡ Record: JOllrna/ 01the Transportation
Research Board, No. 1988, Transportation Research Board of the National
Academies, Washington, D.C., 2006, pp. 63--66.

Olapter 21/AlI-Way STOP-COntrolledintersections References

Version 6.0 Page 21-27
 Highway Capacity Manual: A Guide for MultimodaJ Mobility Analysis

CHAPTER 22
ROUNDABOUTS

CONTENTS

1. 1NTRO O UCTI O N 22.1

Chapter Organization... . 22.1
Related HCM Content . 22-1

2. CON CEPTS 22.2

Capacity Concepts . 22-2
Level-of.Service Criteria . 22-9

3. MOTORIZED VEHIClE CORE METHODOlOGY ............•...•...•....•...•......22.10

Scope of the Methodology 22-10
Required Input Data and Sources 22-14
Computational Stcps 22-15

4. EXTENSION TO THE MOTORIZEO VEHICLE METHOOOWGY 22.25

introduction 22-25
Calibration of Capacity Models 22.25

5. PEDESTRIAN MODE •....•..•...•...•...•...•...•.........................................................•.

22-26

6. B I CYCL E M O DE .•..••..••...•.......................................................•...........••.............. 22.27

7. APPLI CATI ONS ......................•......•...•...•...•...•....................................................22.28

Types oí Analysis 22-28
Example Problems 22-28
Example Results 22-28

8. REFERENCES ........................•......••..•...•...............................................................22-30

Chapter 22jRoundabouts Contents

VetSiQn 6.0 Page 22-i
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

LIST OF EXHIBITS

Exhibit 22-1 Analysis on One Roundabout Leg 22-3

Exhibit 22-2 Example of One-Lane Entry Conflicted by One Circulating
Lane 22-4
Exhibit 22-3 Example of Two-Lane Entry Conflicted by One Circulating
Lane 22-6
Exhibit 22-4 Example of One-Lane Entry Conflicted by Two Circulating
Lanes 22-6
Exhibit 22-5 Example of Two-Lane Entry Conflicted by Two Circulating
Lanes 22-7
Exhibit 22-6 Capacity of Single-Lane and Multilane Entries 22-7
Exhibit 22-7 Right-Turn Bypass Lanes 22-8
Exhibit 22-8 LOS Criteria: Motorized Vehicle Mode 22-9
Exhibit 22-9 Required Input Data, Potential Data Sources, and DefauIt
Values for Roundabout Motorized Vehide Analysis 22-14
Exhibit 22-10 Roundabout Methodology 22-15
Exhibit 22-11 Passenger Car Equivalencies 22-16
Exhibit 22-12 Calculation of Circulating Flow 22-17
Exhibit 22-13 Calcula non of Exinng Flow 22-18
Exhibit 22-14 Assumed (de facto) Lane Assignments 22-19
Exhibit 22-15 Volume Assignments for Two-Lane Entries 22-19
Exhibit 22-16 Capacity Equations for Entry Lanes 22-19
Exhibit 22-17 Capacity Equations for Bypass Lanes 22-19
Exhibit 22-18 Model of Entry Capacity Adjustment Factor for Pedestrians
Crossing a One-Lane Entry (Assuming Pedestrian Priority) 22-20
Exhibit 22-19 Illustration of Entry Capacity Adjustment Factor íor
Pedestrians Crossing a One-Lane Entry (Assuming Pedestrian
Priori ty) 22-20
Exhibit 22-20 Model oí Entry Capacity Adjustment Factor for Pedestrians
Crossing a Two-Lane Entry (Assuming Pedestrian Priority) 22-21
Exhibit 22-21 l1Iustranon of Entry Capacity Adjustment Factor for
Pedestrians Crossing a Two-Lane Entry (Assurning Pcdcstrian
Priority ) 22.21
Exhibit 22-22 Illustration of Capacity for Low Entry F10ws and High
Ci rculating Flows 22-29

I
I
Con'_ OIapter 22JRoundabouts
Page 22.ji Vt'fSiOn 6.0

1
 Highway Capacity Manual: A Guide for Multjmodal Mobility Ana/ysis

1. INTRODUCTION

Roundabouts are interSt'ctions with a generally circular shape, charactcrized VOLUME 3: lNTERRUPTED FLOW
16. Urban Street Facilities
by yield on entry and circulation (counterclockwise in the United States) around 17. Urnan Street Re!iability and
ATOM
a central island. Roundabouts have been used successfully throughout the world
lB. Urban Str"eet 5egments
and are being used increasingly in the United States, especially since 1990. 19. Signalized Intersections
20. TWSC Intersections
This chapter presents concepts and procedures for analyzing these 21. AWSC Intersections
intersections. A Federal Highway Administration-sponsored project (1) has 22. Roundabouts
23. Ramp Terminals and
provided a comprehensive database of roundabout operations for U.S. Attemative Intersectioos
conditions that is bascd on a study of 24 approaches at single-lane roundabouts 24. Qff.Street Pedestrian and
Bi(yde Facilities
and 37 approaches at multilane roundabouts. This study updates work
conducted for National Cooperative Highway Research Program (NCHRP)
Project 03-65 (2). The proecdures that follow are based on these studies'
recommendations. These procedures allow the analyst to assess the operational
performance of an existing or planned one-lane or two-Iane roundabout given
traffie demand levels.

CHAPTER ORGANlZATION
This chapter is organized into the following scctions:
• Section 1 (this section) introduces the chapter.
• Seetion 2 presents basic concepts of the roundabout methodology,
induding capacity coneepts and level-of-service (LOS)criteria.
• Seclion 3 describes the methodological details of the proeedure, which
indude a step-by-step description of the analysis steps, a discussion of
limitations of the method, and required data.
• Section 4 addresses cxtensions of the motorized vehicle analysis
methodology spccifically related to ealibration of the model.
• Section 5 and Seclion 6, respectively, prcsent pcdestrian and bicyde
cvaluation considerations for roundabouts.
• Section 7 describes types of analysis, cxample problems included in
Volume 4, and examplc results.

RELATEO HCM CONTENT

Other HCM content related to this chapter indudes the following:
• Chapter 23, Ramp Terminals and Alternative Intersections, discusses the
analysis of interchange ramp terminals that are roundabouts.
• Chapter 33, Roundabouts: Supplemental, provides example problems of
the roundabout methodology and additional methodological details,
including model calibration.
• Section N, Roundabouts, in Part 2 of the Plmmiflg and Prelimillflry
ElIgilleerillg Applicatiolls Cuide lo the HCM, describes how to incorporate
this chapter's methods and performance measures into a planning
eHort.

Chapter 22{Roundabouts Introcluction

Version 6.0 Page 22-1
Highway Capacify Manual: A Guide for Multfmodal Mobility Analysis

2. CONCEPTS

This chapter presents procedures for analyzing roundabouts, introduces the

unique characteristics of roundabout capadty, and presents terminology specific
to roundabouts.
lnterscction analysis models generally fal! into two categories. Regression
mode1s use {jeld data to develop statistically derived relationships between
geometric features and performance measures such as capacity and delay.
Analyticalmodels are based on traffic fiow theory combined with Held measures
of driver behavior, resulting in an analytic formulation of the relationship
between those field measures and performance measures such as capacity and
delay.
7ñe pnxeáure in this chapter Both types of rnodels are applicable to roundabouts. Gap acceptance models,
uses a combination of
regression and ana!ytiCill an example of an analytical model, are cornmonly applied for analyzing
modeis. unsignalized intersections because they capture driver behavior characteristics
directly and can be made site-specific by custom-tuning the values used for those
parameters. However, simple gap acceptance models may not capture all
observed behavior, and more complex gap acceptance models that account for
limited priority or reverse priority are diHicult to calibratc. Regression modcls
are often used in these cases in which an understanding of driver behavior
characteristics is incomplete. The procedure presented in this chapter, which is
based on a recent analysis of US. field data, incorporates a combination of
simple lane-based regression and gap acceptance models for both single-Iane and
double-Iane roundabouts. As a result, thc capadty models in this chapter focus
on one entry of a roundabout at a time. The roundabout is considered in its
entirety only in the determination of conflicting flow for thc entry under
consideration.

CAPACITY CONCEPTS
The capacity of a roundabout approach is directly infiucnced by fiow
patterns. The three fiows of interest, the entering flow v" thc drculating fiow v"
and the exiting flow Va' are shown in Exhibit 22-1.
7ñe capacity of a roundabout The capacity of an approach decreases as the conflicting flow incrcascs. In
approach decreases as the
drcufating tJow Il1CTeases. general, the primary conflicting flow is the circulating flow that passes directly in
front of the subject entry. The circulating fiow directly confiicts with the entry
fiow, but the exiting flow may also aHect a drivcr's decision to enter the
roundabout. This phenomenon is similar to the eHect of the right-tuming stream
approaching from the left side of a two-way STOI'-controIledinterscction. Until
these drivers complete their exit maneuver or right tum, there may be sorne
uncertainty in the mind of the driver at the yield or stop line about the intentions
oí the exiting or tuming vchide. However, even though this eHect may have an
influence in some cases, induding it did not signiHcantly improve the overaIJ fit
of the capacity models to the data (1), and consequently it is not induded in this
chapter's models.

Coocepts Chapt€r 22jRoundabouts

Page 22-2 Version 6.0
 Highway Capacity Manual: A Guide tor Multimodal MobUity Analysis

Exhibit 22-1
Analysis on One Roundabout
''9
-
v,-
V"\F7r .
\\Vl/
\\\ ,;;
n
When the conflicting flow rate approaches zero, the maximum entry flow is
given by 3,600 s/h divided by the follow-up headway, which is analogous to the
saturation flow rate for a movement receiving a green indication at a signalized
intersection. At high levels of both entering and conflicting flow, limited priority
(in which circulating traffic adjusts its headways to allow entering vehicles to
enter), priority reversal (in which entering traffic forces drculating traffic to
yield), and other bchaviors may occur. In these cases, more complex analytical
models or regression models, such as those incorporated into sorne of the
alternative tools discussed later in this chapter, may give more realistic results.
When an approach opera tes over capadty during the anal)'sis period, a
condition known as capacíty cOllstraínf ma)' occur. During this condition, the
actual drculating £low downstream of the constrained entr)' will be less than
demando The reduction in actual drculating flow ma)' thcrefore increase the
capadty of the affected downstream entries during this condition.
In addition, it has becn suggestcd that origin-destination pattcrns have an
influence on the capadty of a given entry (3, 4). This effect was not identified in a
more recent study (2) and has not becn incorporated into this chapter's models.
Both roundabout design practices and the public's use of roundabouts
continue to mature in the United States. Research at the time oí writing found
U.S. drivers presently seem lo
variation in capadties throughout the United States. Such variation contributes display more heSJtiltion and
to considerable spread in the data; more detail can be found in Chapter 33, cautiOn in llSing roundabouts
lhan drivers in other countries,
Roundabouts: Supplemental. A likely source for this variation is differences in which results in Iower obsetVed
driver behavior in vacious regions, which may be in£luenced by the local driving capaaties.

culture and the density of roundabouts in an area. Other sources for the variation
may inelude geometric features. Research in the United states was not able to
isolate spedfic geometric factors rciative to variations in driver behavior (1),
although some international researeh has identified specific geometric
contributions (S,6). Fadors that may explain differences observed in the United
States compared with other countries inelude limited use of turn indicators at
roundabout exits, differences in vehiele t)'pes, and the effect the comparatively
low use of YIElD-eontrolledintersections has on driver behavior.

Chapter 22/Roundabouts Concepts

Vet3'lon 6.0 Page 22.3
Highway Capadfy Manual: A Guide tor Multimodal Mobility Analysis

U.5. research at the time of Data coHection in 2010 as part of national research resulted in a much larger
wntJi'Ighas no( found
significant increases in capacity and more saturated data set that exhibited higher capacities than previously
OVIY time, but sorne reported in the United States. The capacities presented here are believed to be
geographic areas showed
significantly higher capacities higher primarily due to the larger and more saturated data set and not primarily
than the nationaf average. due to an increase in capacity over time. Although it has generally been assumed
roundabout capacity values in the United States will increase as drivers become
more experienced with roundabouts, it has not becn possible to provide direct
evidence of this characteristic from the available data. Data examined at two
roundabouts observed under saturated conditions in 2003 and 2012 revealed no
significant change in observed capacities. However, Carme!, Indiana, a city with
a large number of roundabouts, had significantly higher roundabout capacity
values than average for U.5. conditions (1).

Single-Lane Roundabouts
The capacity of a single cntey lane conflieted by one circulating lane (e.g., a
single-lane roundabout, iJlustrated in Exhibit 22~2)is based on the conflieting
flow. The equation for estimating the capacity is given as Equation 22-1.
Equation 22-1 ee,pce -- 1380e(-l.02XI0-3)Vc,pc~
,

where
'"pa lane capacity, adjusted for heavy vehicIes (pc!h); and
vt,pa conflicting flow rate (pc/h).

Exhibit 22~2
Example of One-Lane Entry
Conflicted by One Orculating
Cme

The capacity model given above reflects observations made at U.5,

roundabouts in 2012 (1). Considerable variation in capacity was observed in
various regions of the country and with different sites within a region. Therefore,
local calibration of the capacity models is recommended to best reflect local
driver behavior. This topie is discussed later in this chapter and in Chapter 33.

Concepts Olapler 22/Roundabouts

Page 22-4 Version6,O
 Highway Capacity Manual: A Guide for Multjmodal Mobility Analysis

Multilane Roundabouts
Multilane roundabouts have more than one lane on at lcast one entry and at
least part of the circulatory roadway. The number of entey, circulating, and
exitíng lanes may vary throughout the roundabout. Because of the many possible
variations, the computational complexity is higher for multilane roundabouts
than single-lane roundabouts.
The definition oí headways and gaps for multilane facilities is ,lIso more
complicated than for single-lane facilities. lf the circulating roadway truly
functions as a multilane facility, then motorists at the entry perceive gaps in both
the inside ,lnd outside lanes in sorne integrated fashion. Sorne drivers who
choose to cnter the roundabout vi,l the right entry lane will yield to all traffic in
the circulatory roadway due to their uncertainty about the path of the circulating
vehicles. This uncertainty is more pronounccd at roundabouts than at other
unsignalized intersections due to the curvature of the circulatory roadway.
However, sorne drivers in the right entry lane will enter next to a vehicle
circulating in thc inside lane if the drculating vehicle is not perceived to conflict.
In addition, the behavior of circulating vehicles may be affech:>dby the presence
or absence oí lane markings within the circulatory roadway. As a result, the gap
acceptance behavior of drivers in the right entry ¡ane, in particular, is imperfect
and difficult to guantify with a simple gap acceptance model. This difficulty
leads to an inclination toward using a rcgression-based model that implicitIy
accounts for these factors. More detail on the nuances of geometric design,
pavement markings, and their relationship with operational performance can be
found elsewhcre (7).
For roundabouts with up to two circulating lanes, which is the only type of
rnultilane round,lbout addressed by the ,lnalytical methodology in this chapter,
the entries and exits can be either one or twa lanes wide (plus a possible right-
turo bypass lane). The capacity model given bclow refleds observatians made at
VS. roundabouts in 2012 (1). As with single-lane roundabouts, local calibratian
of the capacity models (presented later in this chaptee) is recommended to best
refled local driver behavior.

Capacity for Two-Lane Entries Conf/ict:ed by One Circu/ating Lane

Eguatian 22-2 gives the capacity of each entry lane conflided by one
circulating l,lne (illustrated in Exhibit 22-3).
e -
e.pce -
1, 420e(-O.91)(10-J)Vc.Pc~ Equation 22-2

where al1variables are as defined previously.

Chapter22/RoundabouG Concepts
Version 6.0 Page 22-5
Highway Capacity Manual: A Guide for Mulfimodal Mobility Analysis

Exhibit 22-3
Exampre of Two-lane Entry
HWW'm B'M \ ~ ••••••••••• m

~~
Conflicted by One Orculating
....
",oe . ". ~
. .. rt
)
f •••
,

CapaCity for One-Lane Entries Conflided by Two Circulating Lanes

Equation 22-3 gives the capacity oí a one-Iane roundabout entry conflicted by
two drculating lanes (illustrated in Exhibit 22-4).
Equation 22-3 ee.pce -- 1420e(-O.85XIO-3)Vcpce
, '

where all variables are as defined previously (ve."", is thc total oí both lanes).

Exhibit 22-4 .:.:':-;.:;:-:.:

.. -,',",-,-.',"
.... ..
_

Example of One-lane Entry

Confllcted by Two Orculating
..
"'"'"

Capacity for Two-Lane Entries Conflict:ed by Two Circulating Lanes

Equation 22-4 and Equation 22-5 give the capacity of the right and left lancs,
rcspectively, of a two-Iane roundabout entry conflicted by two circulating lanes
(iIlustrated in Exhibit 22-5).
Equation 22-4 e - 14Z0e(-O.8SXIO-3)Vc,pce
e,R.pce - ,

Equation 22-5 e - 1350e(-O.92xl0-3)Vq,ce

e.L.pce - ,

where
The mpacity of the left lane of capacity of the right entry lane, adjusted for heavy vehicles (pc/h);
a roundabout approach is
Jower tMn the CiJpacity of the
right lane.
capacity of the left entcy lane, adjusted foc heavy vehicles (pc/h); and
conflicting flow cate (total of both lanes) (pc/h).

Concepts Qlapter 22/Roundabouts

Page 22-6 V~6.0
 Highway Capacity Manual: A Guide for Multimodol MobjUty Analysis

4
"V
....:.:--:-:
~\ .
N '--
Exhibit 22-5
Exampleof Two-LaneEntry
ConflictedbyTwoOrculating
",oo,

The calculated capacities for each lane in passenger car equivalents per hour
are adjusted back to vehicles per hour, as describcd later in this section.
Exhibit 22-6 presents a plot showing the cntry capacity cquations (Equation
22.1 through Equation 22-5).

Exhibit 22-6
Capacityof Single-Lane
and
MultilaneEntries
',"'"
1,200

:€ 1,000

.e
••
'0
[
"'" capadty <:J e'lher la"" <:J tw<:>-lane
entry agaill'it one connlCting l¥Ie

a "'" Ciopadty <:J one entry lan\, I ._

<tga'll'it one connlCllng la~ ¡ _.. ..~._. _. _
"'" capildty oflel't larll' of two-lane eotry /
against two confIlCting l/mes
.__ .... ::: .• -
-, .::.:
'00

o
o '00 600 800 1,000 1,200 1,400 1-600 1,800 2,000 2,200

Conflicting Flow Rate (pc/h)

Right- Turn Bypass Lanes

Two common types of right-tum bypass ¡anes are used at both single-Iane
and multilane roundabouts. These are il1ustrated in Exhibit 22-7.
The bypiJss lene capacity
Thc following sectiuns describe each type uf bypass lane. In the United prrxedure does not Ifldude /:he
Sta tes, drivers in both types of bypass lanes would generally be required to yield effect of drivers yieldl{)g to
pedestrians.
tu pedestrians crussing the bypass lane. The capacity eHect af drivers yielding tu
pcdestrians has not been includcd in this analysis procedure.

Chapter22jRoundabouts Concepts
Version 6.0 Page22-7
Highway Capacity Manual; A Guide for Multimodol Mobility Anolysis

Exhibit 22-7
Right.Tum Bypass lanes

,
\ Type 1 (yielding)

tri=====
...--------
Type 1: Yie/ding Bypass Lane
A Type 1 bypass lane terminates at a high angle, with right-turning traffie
yielding to exiting traffie. Right-turn bypass lanes were not explicitly evaluated
in the most reeent national researeh (1). However, the eapaeity of a yield bypass
lane may be approximated by using one of the eapaeity formulas given
previously by treating the exiting flow from the roundabout as the circulatory
flow and treating the flow in the right-tum bypass lane as the entry flow.
The eapacity for a bypass lane opposed by one exiting lane can be
approximated by using Equation 22-6.
Equation 22-6 ebypass.pce = 1' 380e(-1.02><10-3)vex.pce

The eapacity for a bypass lane opposed by two exiting lanes can be
approximated by using Equation 22-7.
Equation 22-7 ebypass.pce -- 1, 420e(-O.B5xlO-3)Vu,pce

where
c!>¡"I" ••. p<t= eapaeity of the bypass lanc, adjusted for heavy vehicles (pe/h); and
v"".p<t = eonflicting exiting flow rate (pe/h).

Type 2: Nonyie/ding Bypass Lane

A Type 2 bypass lane merges at a low angle with exiting traffie or forms a
new lane adjaeent to exiting traffie. The eapacity of a merging bypass lane has
not becn asscssed in the United States. lts eapacity is expectcd to be relatively
high due to a merging operation bctween two traffie streams at similar speeds.

Exit Capacity
German researeh (8) suggests that the eapacity of an exit lane, aecounting for
pedestrian and bicyde traffic in a typical urban area, is in the range of 1,200 lo
1,300vehides per hOUf.However, the analyst is eautioned to also evaluate exit
lane requirements on the basis of vehicular lane numbers and arrangements. For
example, a double.lane exit might be required to receive two through Janes to
provide basic lane continuity along a corridor, regardless of the exit volume.
Further guidance can be found in an NCHRP report (7).

Concepts Chapter 22/Roundabouts

Page 22-8 Version 6.0 I
I
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

lEVEL-OF-SERVICE CRITERlA
LOS criteria for motorized vehides in roundabouts are given in Exhibit 22-8.
As the table notes, LOS F is assigned if the volume-to-capacity ratio of a lane
exceeds 1.0 regardless of the control delay. For assessment of LOS at the
approach and intersection levels, LOS is based solely on control delay.
The thresholds in Exhibit 22-8 are based on the judgment of the
Transportation Research Board Committee on Highway Capacity and Quality of
Service. As discussed later in this chapter, roundabouts share the same basic
control delay formulation with two-way and all-way sTOP-controlled
intersections, adjusting for the eHect of YIELDcontrol. However, at the time of
publicaban of this edition of the Highway Capacify Manual (HCM), no research
was available on traveler perception of quality of service at roundabouts. In the
absence of such rescarch, the service measure and thresholds have been made
consistent with those for other unsignalized intersections, primarily on the basis
of the similar control del ay formulation.

Control Delay lOS by Volume-to-CaDacitv Ratio4 Exhibit 22-8

s veh v/cs. 1.0 v/c> 1.0 lOS eriteria: Motorized
0-10 A F Vehicle Mode
>10-15 B F
>15-25 e F
>25-35 D F
>35-50 E F
>50 F F
Note: • For <Ipproaches <Ind interseetionwide assessment, LOS is defined 50Iely by control delay.

Chapter 22/Roundabouts Concepts

Version 6.0 Page 22-9
Highway Capacity Manual: A Guide far Mulfimodol Mobility Analysis

3. MOTORIZED VEHICLE CORE METHODOLOGY

SCOPE OF THE METHODOLOGY

This section focuses on the operation oí roundabouts. This version of the
roundabout analysis procedures is primarily based on studies conducted for the
Federal Highway Administration (1) llpdating work conducted for NCHRP
Project 03-65 (2).
The methodology does not necessarBy apply to other types oí circular
intersections, such as rotaries, neighborhood traffic ardes, or signalized traffic
cirdes, because these types of circular interscctions usually have geometric or
traffic control elements that deviate from those used in rollndabouts. As a resllit,
their operational performance may differ significantly from that experienced at
roundabouts and thus cannot be accurate1y rnodeled by using the procedures in
this section. More detail on the differentiation betwcen roundabouts and other
circular intersections can be found elsewhere (7).

Spatial and Temporal Limits

The analytical procedure presented in this section assurnes the analysis
boundarics are the roundabout itseif, induding associated pedestrian crosswalks.
Altemative t001sdiscussed in this section can, in sorne cases, expand the analysis
boundaries to inelude adjacent intersections. The methodology presented here
discusses motorized vehieles, pedestrians, and bicyeles.
The recommended length of the analysis period is the HCM standard of 15
min (although longer periods can be examined).

Performance Measures
This method produces the following performance measures:
• Volume-to-capacity ratio,
• Control delay,
• LOS based on control delay, and
• 95th percentile queue length.

Limitations of the Methodology

The procedures presented in this section cover many of the typical situations
a user may encounter in practice. However, for sorne applications alternative
tools can produce a more accurate analysis. The following Iimitations, stated
earHer in this section, may be addressed by using available simulation tools. The
conditions beyond the scope of this chapter that are treated explicitly by
aiternative tools ¡nelude
• Pedestrian signals or hybrid beacons at roundabout crosswalks,
• Metering signals on one or more approaches,
PriQrity revefSill CiJn occur
when entering traffic
• Adjacent signals or roundabouts,
dominates an entry, calJSing
drcufating traffic ro yield.
• Priority reversal under extremely high fiows,

Motorized Vehicle Core Methodology Olapter 22/Roundabouts

Page 22-10 VetSion 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

• High pedestrian or bieycle activity levels,

• More than two entry lanes on an approach, or
• Flared entry lanes. A typical flared entJy is one
tllat widens from one approach
A few of the more common applications of alternative tools to overcome the lane lo two entJy lanes. Other

limitations of the procedures presented in this section are discussed in the

following subsections.
fiareso,
flaring combinations, inc/uding

possible.
lane width, are

lnteradion Effects with Other Traffie Control Deviees

Several common situations can be modeled with alternative tools:
• Pedestrian sigllals or hybrid beaCO/lSat rOlmdabout crosswaiks. These devices,
described in detail in the Manual on Uniform Traffic Control Devices Jor
Streets and Highways (9), can be used in a variety of applications, including
the following:
o High vehicle flows in which naturally occurring gaps in vehicle traffic
or vehicular yielding for pedestrians is insufficient;
o High pedestrian £lows in which unrcstricted pedcstrian crossing
activity may create insufficient rnotorized vehicle capacity; and
o Crossing situations in which pedestrians with vision or othcr
disabilities may not receive equivalent access to the crossing; such
access is a legal requirement in the United States under the Americans
with Disabilities Act and is regulated by the U.S. Access Board (10).
• Metering sig/lals on roundabouf approaches. These signals are somctimcs
used in applications in which a dominant entering flow reduces
downstream entry capacity to zero or nearly zero. A metering signal can
create gaps in the dominant flow at regular intervals or as dictated by
queuing at the downstream entry.
• Signals used to give priority fo other usas. Thesc applications indude at-
grade rail crassings, emergency vehide signals, and others.
• Nearby i/ltersectio/ls or traffic cmltrol devices ut w/¡ich queues or ¡une use effeets
i/lteruct. These nearby intersections can have any type of control,
including signalization, STOPcontrol, or YIElD control (as at anothcr
roundabout). Applications could also indude noninterscction treatments
such as freeway rarnp rneters.
Although sorne deterministic interscction tools can model these situations,
they are oftcn trcated more satisfactorily by using stochastic network models.

Flared Entríes or Short-Lane Applieations

Flared entries or short-lane applications are sometimes used at roundabouts
to add capacity at the entry without substantially widcning the approach
upstream of the entry. Common applications indude £laring from one lane to
two lanes at the entry or fram two lanes to three lanes, although sorne
intemational research has found capacity sensitivity to £laring in sub-Iane-width
increments (5).

Chapter 22/Roundabouts Motorized Vehicle Core Methodology

Version 6.0 Page 22-11
Highway Capacity Manual: A Guide tor Mulfimodal Mobility Analysis

The methodology presented in this section provides a mechanism for

flagging conditions under which queues for a given lane may exceed available
storage or block access to adjacent lanes. Alternative tools may provide more
accurate modcling of these situations.

Three-Lane Roundabouts
Three-Iane roundabouts are not induded in the methodology described in
this section but can be analyzed by a number of aIternative tools. Note that no
data for three-Iane roundabouts are available in the source material (1, 2) for this
chapter's methodology, so the analyst should use care in estimating calibration
parameters.

Use of Alternative Tools

General guidance for the use of alternative traffic analysis tools for capacity
and LOS analysis is provided in Chapter 6, HCM and Altemative Analysis Tools,
and Chapter 7, Interpreting HCM and A1ternative TooI Results. This section
contains specific guidance for the application of altemative tools to the analysis
of roundabouts. The reader should also be familiar with the information and
guidance on the design and evaluation of roundabouts (7).
Two modeling approaches are used in the types of alternati ve tools
commonlyapplied:
• Deterministic illfersectioll mode/s.These models represent vehide flows as
flow rates and are sensitive to various fIow and geometric features of the
roundabout, induding lane numbers and arrangements or specific
geometric dimensions (e.g., entry width, inscribed cirde diameter), oc
both. The majority of these rnodels are anchored to research conducted
outside the United States (e.g., 5, 6, 11, 12). Sorne software
implernentations may indude more than one model or employ extensions
beyond the original fundamental research conducted within a particular
country. Sorne deterministic rnodels can model an entire network of
intersections, but they generally assume no interaction effects between
intersections, thus potentially limiting their application.
• Stochastic "etU'ork models. These models represent vehicle f10ws by
simulating individual vehicles and their car-following, lane choice, and
gap acceptance decisions. The models are based on a variety of
fundamental research studies on driver behavior (e.g., 13, 14). By their
nature, most stochastic models used for roundabouts can model an entire
network oí intersections, thus making them capable of modeling a
broader range of problems. However, their data requirements are
typically more intensive than for the deterministic intersection models.
Most stochastic models are implemented in microsimulation tools.

5trengths of the HCM Procedure

The procedures in this section are based on extensive research supported by
a significant quantity oí field data. They evolved over several years and
represent a body of expert consensus. They produce unique deterministic results
for a given set of inputs, and the capacity of each approach is an explicit part oí

Motortzed Vehide Core MethocIology Chapter 22{ROUndabouts

Page 22-12 VersiOn 6.0
 Highway Capacify Manual: A Guide far Multimodal Mobilify Analysis

the results. Alternative tools based on deterministic intersection models also

produce a unique set of results, including capacitics, for a given set of inputs, but
simulation-based tool5 may produce different results based on diHercnt random
number sequences. Unique results from an analysis tool are important for sorne
purposes such as development impact reviews.

Adjustment of Simulation Parameters to the HCM Results

Calibration of any model uscd to analyzc roundabouts is essential in
producing realistic results that are consistent with field data. Ideally, field data
should be used for calibration. For situations involving the asscssment of
hypothetical or proposed alternatives for which no ficld data exist, alternative
too! results may be made more compatible with HCM results by adjusting
alternativc too! parameters to obtain a better match with the results obtained
from the HCM procedures as follows:
• Determinis(ic intersectiO/llIJodels. Typical calibration parameters for
deterministic models inelude global adjustment factors that shift or shape
the capacity model uscd by the model. Thesc adjustment factors inelude
adjustments to thc intercept and slope of linear models or other shaping
parameters of more complex analytical forms.
• Stochastic /let1.t'orkmode/s. Calibration of stochastic models is more
challenging than for deterministic models because sorne calibration
factors, such as factors related to driver aggressiveness, often apply
globally to aHelements of the network and not just to roundabouts. In
other cases, the specific coding of the model can be fine-tuned to reflect
localized driver bchavior, ineluding look-ahead points Cor gap acceptance
and locations for discretionary and mandatory lane changes.

Step-by-Step Recommendations for Applying Alternative Tools

The following steps should be taken in applying an altemative tool in the
analysis of roundabouts:
1. Identify the limitations of the HCM procedures that dictate the use of
altemative tools.
2. Decide between a microscopic and a macroscopic modeling approach.
3. If possible, develop a simpler configuration that can be analyzed by the
HCM procedures. Analyze the simple configuration by using both the
HCM and the selected alternative tool. Make adjustments to the
alternative tool parameters to obtain a better match with the ¡-KM
results.
4. Perform the analysis of the full configuration using the alterna tive too!.
5. Interpret and present the results.

Sample Calculations IlIustrating Alternative Tool Applications

Chapter 29, Urban Strcet Facilities: Supplemental, ineludes an example of the
application of a simulation tool to assess the eHect of using a roundabout within
a coordinated arterial signal system. The interactions between the roundahout

Qlapter 22/Roundabouts Motorized Vehicle Cofe Methodology

Version 6.0 Page 22.13
 ,

Highway Capacity Manual: A Guide tor Mulfimodal Mobility Analysis

and the arterial system are examined by usiog signal timing plans with difierent
progression characteristics.

REQUIRED INPUT DATA AND SDURCES

Exhibit 22-9lists the ¡nfacmation neccssary to apply the motorized vehicle
methodology an~ suggests potential soueces tor obtaining these data. lt also
suggests default values foc use when intersection-specific ¡nfacmation is not
available.
No dcfault values have been developed specifically foc roundabouts.
However, a comprehensive presentatian of potential default values foc
interrupted-f1ow facilities is available (15) with specific recornmendations
summarized in its Chapter 3, Recommended Default Values. These defaults
cover the key characteristics of peak hour factor and percentage of heavy
vehicles. Recommendations are based on geographical region, population, and
time of day. AH general default values for interrupted-flow facilities may be
applied to the analysis of roundabouts in the absence of field data or projections
of conditions.
Default values for lane utilization on two-Iane raundabout approaches are
not pravided in the aboye reference (15). In these cases, in the absence of field
data, the effect oí lane utilization imbalance can be approximated by using the
suggested deíault values given in Exhibit 22-9.
As the numbcr of dcfault values used in any analysis increases, the analysis
result becomes more appraximatc and may be significantly different fram the
actual outcome, depending on local conditions.

Exhibít 22-9 Required Data and Units Potential Data Source(s) Suggested Default Value
Required Input Data, Potential GeometTiC Data
Data $ources, and Default
Number and configuration of
Values tor Roundabout Design plans, road inventory Must be provided
lanes on each approach
Motorized Vehicle Analysis
DemandData
Hourly ruming movement
demand volume (vehjh) ANO
peak hour factor
Fleld data, modeling Must be provided
OR
Hourly ruming movement
demand fIow rate (vehjh)
Analysis pericxi lengtll (min) 5et by analyst 15 min (0.25 h)
Peak hour factor (decimal) Fleld data 0.92
Heavy-vehide percentage (%) Fleld data 3%
lane utilizatiOn Fleld data Left:-through + through-
right:
% trafflC in Ieft lane: 0.47.
% traffic in right lane: 0.53.

Left-through-right + right:
% trafflC in Ieft laoe: 0.47.
% trafflC in right Iane: 0.53.

Left + left-through-right:
% traffte in left lalle: 0.53.
% trafflC in right Iane: 0.47.
Note: • These values are generally ooosistent Wlth observe<!values for through movement5 at s;gnalized
intersections. These values should be applied with care, particularly under col'lditions estimated to be near
c.apac:ity.

Motorized Vehide Core Methodology Chapter 22jRoundabouts

Page 22-14 Ver.skm 6.0
 Highway Capacity Manual: A Gujde for Mulfimodal Mobility Analysis

COMPUTATIONAL STEPS
The capacity of a given approach is computed by using thc process
illustratcd in Exhibit 22-10.

5tep 1: COn..••
ert mo ..••
ement demand ..••
olumes te f10w rates
Exhibit 22-10
Roundabout Methodology

5tep 2: Adjust f10w rates for heavy ..••

ehicles

5tep 3: Determine drculating and exiting f10w rates

5tep 4: Determine entry f10w rates by lane

5tep 5: Determine the capacity of each entry lane and

bypass lane as appropriate in passenger car equi ..••
alents

5tep 6: Determine pedestrian impedance te vehicles

5tep 7: Convert lane f10w rates and capacities into

vehicles per hour

5tep 8: Compute ttle volume-to-capacity ratio for each lane

5tep 9: Compute the average control delay foc each lane

5tep 10: Determine lOS for each lane on each approach

5tep 11: Compute the a..••

erage control delay and determine
lOS for eaen approach and the roundabout as a whole

5tep 12: Compute 95th percentile queues for each lane

Step 1: Convert Movement Demand Volumes to Flow Rates

For an analysis of existing conditions in which the peak 1S-min period can be
measurcd in the ficld, thc volumes for the peak 1S-min period are con verted to a
peak lS-min demand flow rate by multiplying the peak lS-min volumes by four.
For analysis of projected conditions or when 15-min data are not available,
hourly demand volumes for each movement are converted to peak lS-min
demand flow rates in vehicles per hour, as shown in Equation 22-8, through the
use of a peak hour factor for the intersection.

Chapter 22/Roundabouts Motorized Vehicle Core Methodology

Version 6.0 Page 22.}5
Highway Capacify Manual: A Guide tor Multimodal Mobilify Analysis

V,
Equation 22-8 vi = PHF
where
Vi demand flow rate for movement i (veh/h),
Vi demand volume for movement i (veh/h), and

PHF peak hoor factor.

Ir PHF ÍS useó, a single If peak hour factors are used, a single peak hour factor for the entire
intersedionwide PHF shoufd be
used rather thiJn movement- intersection is gcnerally preferred to decrease the Iikelihood of creating demand
specfflC or approach-spedfic scenarios with conflicting volumes that are disproportionate to the actual
PHFs. Ir individwl approaches
or movements peal< at volumes during the 15-min analysis period.1f peak hoor factors for each
different times, a series of 15- individual approach or movement are used, they arc likely to generate demand
m/n analysis periods that
encompasses the peaking volumes from one 15-min period that are in apparent conflict with demand
should be considered. volumes from another 15-min period, but in reality these peak volumes do not
ocrur at the same time. Furthermore, to determine individual approach or
The use of a peak 15-min movement peak hour factors, actual 15-min count data are likely available,
traffic count muftiplied by tour
is preferred for existing permitting the determination of actual 15-min demand and avoiding the need to
condítions when tTatfíc COl/nts use a peak hour factor. In the event individual approaches or movements are
are available. Tñe use of a l-h
demand voIume divided by a known to have substantially different peaking characteristics or peak during
peak holJr factor is preferred diHerent 15-min periods within the hour, a series of 15-min analysis periods that
wlth projeded voIumes or wffh
projected voIumes that have encompasses the peaking should be considered instead of a single analysis
bren added ro CUITl!fJt period using a single peak hour factor for the intcrsection.
voIumes.

5tep 2: Adjust Flow Rates tor Heavy Vehicles

The flow rate for each movement may be adjusted to account for vehicle
stream characteristics by using factors given in Exhibit 22-11.

Exhibit 22.11 VehldeType Passenger Car Equivalen!,. Er

Passenger Car Equivalencies Passenger car l.0
Heavy vehide 2.0

The calculation to incorporate these values is given in Equation 22-9 and

Equation 22-10.

Equation 22-9
V,
v¡,pce =F
,"v
1
Equation 22-10
fHV = 1 + Pr(Er - 1)
where
V j.p<t demand fIow rate for movement j (pc/h),
Vi dcmand fIow rate for movement j (veh/h),
fiN .: heavy-vchide adjustment factor,
P, proportion of demand volume that consists of heavy vehicIes, and
E, passenger car equivalent for heavy vehicIes.

Motorized Vehicle Core Methodology Chapter 22/Roundabouts

Page 22.16 Version 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

5tep 3: Determine Circulating and Exiting Flow Rates

Circulating and exiting flow rates are calculated for each roundabout leg.
Although the lollowing sedions prescnt a numerical methodology lar a four-Ieg
roundabout, this methodology can be extended tu any number al legs.

Circulating Flow Rate

11Iecirculating flow opposing a given entry is defined as the flow conflicting
with the entry flow (Le., thc flow passing in front of the splitter island next to the
subject entry). The circulating flow rate calculation for the northbound
circulating flow rate is illustrated in Exhibit 22-12 and numerically in Equation
22-11. AII flows are in passenger car equivalents.

Exhibit 22-12
(aICtllation of CirCtllating Flow

'.')Dr
~VWBU

-:::d -
===,,1
-- . D~=='
- =

Equation 22-11
VC.NB.pce = vWBu,pce + vSRL,pce + vSRu.pce + vEBT.pce + vEBL.pce + VEBU,Pce
Exiting Flow Rate
11Ieexiting flow rate far a given leg is used primarily in the calculation of Ir a bypass /ane iSpresent on
!he immediate upstream entry,
conflicting f10wfor righHurn bypass lanes. 11Ieexiting flow calculation for the rhe right-tuming flow using!he
southbound exit is iIIustrated in Exhibit 22.13 and numerically in Equation 22-12. bypass lane is dedueted from
rhe exiting fIow.
lf a bypass lane is present on the immediate upstream entry, the right-turning
f10wusing the bypass lane is deduded fram the exiting flow. AII flows are in
passenger car equivalents.

Chapter22/Roundabo~ Motorized Vehide Core Methooology

Version 6.0 Page 22-17
Highway Capacify Manual: A Guide tar Multimodal Mobilify Analysis

Exhibit 22-13
(alculatioo of Exiting Flow

Ü
"b~'
•
- •,

'~::
I
====
--

Equation 22-12
Vex.5B,pce = VNBU.pce + VWBL,pce + VSBT,pce + VEBR,pce - VEBR,pce,bypass

Step 4: Determine Entry Flow Rates by Lane

For single-Iane entries, the entry f10wrate is the sum of aH movement f10w
rates using that entry. For multilane entries or entries with bypass lanes, or both,
the following procedure may be used to assign f10wsto each lane:
1. If a right-tum bypass lane is provided, the f10wusing the bypass lane is
removed from the cakulation of the roundabout entry f1ows.
2. If only one lane is available for a given movement, the f10wfor that
A de facto fane is one movemcnt is assigncd only to that lane.
designated tor muitipfe
movements but tIlat may 3. The remaining f10ws are assumed to be distributed across alllanes,
operate as an exdusíve IiJne
due ro él dominant movement
subject to the constraints imposed by any designated or de facto lane
demiJnd. A common exampfe ÍS assignments and any observed or cstimated lane utilization imbalances.
a Ieft-/hrough IiJne witIJ a Ieft-
turn ffow rate tIlat greatly Five generalized multilane cases may be analyzed with this procedure. For
exreeds the tlJrough ffow rate. cascs in which a moverncnt may use more than one lane, a check should first be
made to determine what the assumed lane configuration may be. This
configurabon may differ from the designated Jane assignment based on the
specific turning movement patterns being analyzed. These assumed lane
assignments are given in Exhibit 22-14. For intersections with a different number
of legs, the analyst should exercise reasonable judgment in assigning volumes to
each lane.

Motorized Vehicie Core Methodology Chapter 22jRoundabouts

Page 22-18 VetS'on 6.0
 Highway Capacify Manual: A Guide tor Multimodal Mobility Ana/ysis

Deslgnated Lane Assignment Assumed Lane Assignment Exhibit 22-14

If + Vl:> V L, TR (de facto left-turn tane) Assumed (de facto) Lane
VI! r+ VI(.<:
Assignments
left-through, through-right If t'll, > VI!+ t't + VI: LT, R (de facto right-tum lane)
ElseLT, TR
If V[ + vI(.< > Vu + v,: L, TR (de facto through-right lane)
Left, left--through-right
Else L, LTR
If vI! + v1 + v[:> vR.<: LT, R (de tacto Ieft-through lane)
Left-through-right, right
Else LTR, R
Notes: vl.I v" v" and v..,.are the U-tum, Jeft-tum, through, and nonbypass right.tum fIow rates using a given
entr)', respectively. L " 1eft; LT " Ieft-thmogh; TR " through-right; l TR " Jeft-through-right; R " right.

On the basis of the ilssumed ¡ane assignment for the entr)' and the lane
utilization effect described abo\'(', £lo\\' rates can be assigned to each lane by
using the formulas given in Exhibit 22-15.

case Assumed lane Assignment Len lane Right lane Exhibit 22-15
Volume Assignments for Two-
1 Left, through-right Vu+ Vi VI + VI(.<
Lane Entries
2 left-through, right vI! + VI + Vr Vil.

3 left-through, through-right (%LLw. ('foRUv.

4 Left,left-through-right (%LLw, (%RLw,
5 Left-through-right, right (%LUv, (%RLw,
Notes: vl.I v" <lnd VII;. are the U-turn, Ieft-tum, through, and nonbypass righHurn fIow rates using a given
Vr,
entr)', respectively. l " Ieft; lT " Ieft-through; TR " through-right; l TR " Ieft-through-right; R " right;
%RL " percentage of entry traff" using the right Ial'le; %LL " percentage of entry traff" using the 1eft:
!afie. %LL + %RL " 1.

Further disrussion of lane use at multilane roundabouts, including

conditions that ma)' create unequallane use, can be found in Chapter 33,
Roundabouts: Supplcmental, iocated in HCM Volume 4.

Step 5: Determine the Capacity of Each Entry Lane and Bypass Lane as
Appropriate in Passenger Car Equivalents
The capadty of each entry lane and bypass lane is calrulaled by using the
capadty cquations disrusscd above. Capadty equations for entey lanes are
summarized in Exhibit 22-16;capadl}' equalions for bypass lanes are
summarizcd in Exhibit 22-17.

Entering Connicting Exhibit 22-16

Lanes Circulating Lanes Capadty Equation Capacity Equations for Entry
1 1 Equation 22-1
la,,,
2 1 Each lane: Equation 22-2
1 2 Equation 22-3
2 2 Right lane: Equation 22-4; leh lane: Equation 22-5

Connicting Exiting Lanes Capadty Equatlon Exhibit 22.17

1 Equation 22-6 Capacity Equations for Bypass
2 Equation 22-7 1.""

Chapter 22/Roundabouts Motorized Vehicle Core Methodology

Version 6.0 Page 22-19

L_ ..
_
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Step 6: Determine Pedestrian Impedance to Vehicles

Pedestrian traffi~ can reduce the vehicular capacity of a roundabout entry if
suHident pedestrians are present and thcy asscrt thc right-of-way typically
granted pedestrians in most jurisdictions. Under high vehicular conflicting £lows,
pedcstrians typical1y pass between queucd vehides on entry and thus have
negligible additional impact on vehicular entry capacity. However, under low
vehicular con£licting flows, pedestrians can effectively function as additional
conflicting vehides and thus reduce the vehicular capacity of the eotry. The
eHect of pedestrians is more pronounced with increased pedestrian volume.
For one-lane roundabout entries, the model shown in Exhibit 22-18 can be
used to approximate this effect (8).These equations are illustrated in Exhibit 22-
19 and are based on the assumption that pedestrians have absolute priority.

Exhibit 22-18 ca•• One-Lane Entry capadty Adjustment Factor for Pedestrians
Model of Entry Capadty
Adjustment Factor for ifvc,pce > 881 {ped = 1
Pedestrians Crossing a
One-Lane Entry (Assuming
Else ifnped S 101 {ped = 1- 0.000137nped
Pedestrian Priority) 1,119.5 - 0.715vc,pct - 0.644nped + O.00073vc.pcenped
Else
{ptd = 1,068.6 - 0.654vc.pct

where
ff"'J entry capacity adjustmcnt factor for pedestrians,
nf"'J number of conflicting pedestrians per hour (p/h), and
ve.pa conflicting vehicular £low rate in the circulatory roadway (pc/h).

100 SO~
Exhlbit 22-19
IIIustration of Entry Capadty 100 p/h
Adjustment factor foc
Pedestrians Crossing a
One-Lane Entry (Assuming
Pedestrian Priority)
~--

0,00
o 100 200 300 400 SOO 600
COnflicting Circulating Flow (pc/h)
700
•• 900 1,000

For two-Iane entries, the model shown in Exhibit 22-20 can be used to
approximate the eHect of pcdestrians (8). These equations are illustrated in
Exhibit 22-21 and share the assumption as befare that pedestrians have absolute
priority.

Motorized Vehicle Core Methodology Chapter 22JRoundabouts

Page 22.20 Verron 6.0
 Highway Capacity Manual: A Guide tor Multimodal Mobility Analysis

Ca •• Two-Lane Entry Capacity Adjustment Factor lor Pedestrians Exhibit 22-20

Model of Entry capadty
_ . [ nped ( 1,260.6 - 0.329vc.pce - 0.381 x 100) ]
lfnped < 100 Adjustment Factor for
fped - mm 1- 100 1- 1380-0Sv ,1

I
, • C,pc~ Pedestrians Crossing a
Two-lane Entry (Assuming
. [1,260.6 - 0.329vc,pce - 0.381nped
Else fp~d =' mm -----~~---~, 1 Pedestrian Priority)
1,380 - 0.5vc.p,~

where
entry capacity adjustment factor for pedestrians,
!""
number of conflicting pedestrians per hour (plh), and
conflicting vehicular flow rate in the circuJatory roadway (pc/h).

LOO Exhibit 22-21

IIIustration of Entry capacity
Adjustment Factor for
Pedestrians Crossing a
J Two-lane Entry (Assuming
Pedestrian Priority)
~ 0,95

!:= 0,90

~
.s
~
J~ 0,85
•
.:;

000
o '00 400 600 800 1,000 1,200 1,400 L,OOO

Conflicting Circulating Flow (pc/h)

Step 7: Convert Lane Flow Rates and Capacities into Vehicles per Hour
The flow rate for a given Jane is converted back to vehicles per hour by
multiplying the passenger-car-eguivaJent flow rate computed in the previous
step by the heavy-vehicle factor for the ¡ane as shown in Equation 22-13.
v¡ = V¡,PCCfHV,e Equation 22-13

where
Vi flow rate for Jane i (veh/h),
V •.pce flow rate for Jane i (pc/h), and
/HV.. heavy-vehicle adjustment factor for the Jane (see below).
SimiJarJy, the capacity for a given Jane is converted back to vehicles per hour
as shown in Eguatian 22-14.
Equation 22-14

Chapter 22/Roundabouts Motorized Vehicle Core Methodology

V~6.0 Page 22-21
Highway Capacity Manual: A Guide fer Multimedal Mebility Analysis

where
C¡ capacity for ¡ane i (veh/h),
e ¡,Pe£ capacity for ¡ane j (pclh),
f/IV,L heavy-vehicle adjustment factor for the lane (see below), and
fptd pedestrian impedance factor.
The heavy-vehicle adjustment factor for each entry lane can be approximated
by taking a weightcd average of the heavy-vehicle adjustment factors for each
movement entcring the roundabout (excluding a bypass lane if present)
weighted by flow rate, as shown in Equation 22-15.

Equation 22-15
!J - fHV,UVU,PCE + fHV,LVL,PCE + fHV,TVT,PCE + fHv,R.eVR,e,PCE:
HV,e - VU,PCE + VL,PCE + VT,PCE + VR,e,PCE
where
fllv,L heavy-vehicle adjustment factor for the entry lane,
¡/IV.; heavy-vehicle adjustment factor for movement j, and
v~Pe£ demand flow rate for movement i (pc/h).
lf specific lane-use assignment by heavy vehicles is known, heavy-vehicle
adjustment factors can be calculated separately for each tane.
Pedestrian impedance is discussed later in this chapter.

Step 8: Compute the Volume-to-capacity Ratio for Each Lane )

For a given tane, the volume-to-capacity ratio x is calculated by dividing the
lane's calculated capacity into its demand flow rate, as shown in Eguation 22-16.
Both input values are in vehicles per hour.
VI
Xi =-
C,
where
Xi volurne-to-capacity ratio of subject lane j,

Vi dernand flow rate of subject lane i (vch/h), and

C¡ capacity of subject lane j (vehlh),

Step 9: Compute the Average Control Delay for Each Lane

Delay data collected foc roundabouts in the United States suggest control
delays can be predicted in a manner generally similar to that used for other
unsignalized intersections, Equation 22-17 shows the rnodel that should be used
to estimate average control del ay foc each tane of a roundabout approach.

Equatlon 22-17 3,600

d =-c-+900T [
x-l + (x - 1)2 + ~T
C,600) X
1+ 5 x min [x, 1)
71Iethird term of this eqwtion
uses the CiJJcufatedvo/~to-
CiJpacity ratio (JI"1, whichever
is Iess.

Motorized Vehicle 'ore Methodology Chapter 22/Roundabouts

Page 22.22 VeI5k1n 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

where
d average control delay (s/veh),
x = volume-to-capadty ratio of the subject lane,
e = capadty of the subject lane (veh/h), and
T time period (h) (T= 0.2S h for a lS-min analysis).
Equation 22-17 is the same as that for STOP-controlledintersections except
that the "+ 5" term has been modified. This modification is necessary to account
for the YIELD control on the subject entry, which does not require drivers to come
to a complete stop when there is no conflicting traffic. At higher volume-to-
capacity ratios, the likeJihood of coming to a complete stop ¡ncreases, thus
causing behavior to resemble STOPcontrol more dosely.
Average control delay for a given lane is a function of the lane's capadty and
degree of saturation. The model used aboye to estimate average control delay
assumes no residual queue at the start of the analysis periodo If the degree of
saturation is greater than about 0.9, average control del ay is significantly affected
by the analysis periad length. In most cases, the recommended analysis periad is
15 minoIf demand exceeds capadty during a 15.min period, the dclay results
calculated by the procedure ma)' not be accurate due to the likely presence of a
queue al the start of the analysis periodo In addition, the canflicting demand for
movements dawnstream of the mavement operating over capacity may not be
full)' realized (in other words, the flow cannot get past the oversaturated entry
and thus cannot conflict with a downstream entry). In these cases, an iterative
approach that accounts far this eHect and the carryover of queues from one time
pcriod to the next may be considercd, as discussed c1se\'.here (16).

5tep 10: Determine LOS tor Each Lane on Each Approach

LOS for each lane on each approach is determined by using Exhibil 22-8 and
the computed or measured values of control deJay.

Step 11: Compute the Average Control Delay and Determine LOS tor
Each Approach and the Roundabout As a Whole
The control del a)' for an approach is calculated by computing a weighted
average of the delay for each lane on the approach, weighted by the volume in
each lane. The calculation is shown in Equation 22-18. The volume in the bypass
lane should be included in the delay calculation far the approach. LOS for each
approach is determined by using Exhibit 22.8 and the computed or measured
values of control deJay.
dUYLL + dRLVRL + dbypassVbypass Equation 22-18
dapproach = --------~----
VLL + VRL + Vbypass

The control delay for the interscction as a whole is similarly calculated by

computing a weighted average of the delay for each approach, weighted by the
volume on each approach, as shown in Eguabon 22-19. LOS far the intersection is
determined by using Exhibit 22-8 and the computed or measured values of
control dela)'.

Chapter 22JRoundabouts Motolized Vehicle Core Methodology

VersiOn 6.0 Page 22.2]
Highway Capacity Manual: A Guide tar Mu/fimadal Mobility Analysis

¿d¡v¡
Equation 22-19
dintersection = ¿ v¡
where
dinterse<;lKm control delay for the entire intersection (s/veh),

d, control delay for approach j (s/veh), and

v, flow rate for approach i (veh/h).

Step 12: Compute 95th Percentile Queues for Each Lane

The 95th percentile queue for a given lane on an approach is calculated by
using Equation 22.20.

3.600)
Equation 22-20
(1 - x)' +
( -x(C)
1~OT 3.600

where
Q95 = 95th percentile queue (veh),
x = volume-to-capacity ratio of the subject lane,
e = capacity of the subject lane (veh/h), and
T time period (h) (T= 1 for a 1-h analysis; T = 0.25 foc a 15-min analysis).
The queue length calculated for each ¡ane should be checked against the
available storage. The queue in each lane may interact with adjacent lanes in one
or more ways:
• If queues in adjacent lanes exceed the available storage, the queue in the
subject lane may be longer than antidpated due to additional qucuing
from the adjacent lanc.
• If queues in the subject lane exceed the available storage for adjacent
lanes, the adjacent lane may be starved by the queue in the subject lane.
Should one or more of these conditions occur, a sensitivity analysis can be
conducted with the methodology by varying the demand in each lane. The
analyst may also use an altemative tool that is sensitive to lane-by-Iane effeets.

Motorized Vehlcle Core Methodology Chapter 22fRoundabouts

Page 22.24 Version 6.0
 Highway Capacify Manual: A Guide tor Multimodal Mobility Analysis

4. EXTENSION TO THE MOTORIZED VEHICLE

METHODOLOGY

INTRODUCTION
As noted in Section 2, research has found variation in roundabout capacities
throughout the Unitcd Statcs; diffcrences in driver behavior and geometric
factors are potential causes of this variation (1). To address this potential for
variation, this section presents a method for calibrating the HCM capacity
models for local conditions.

CALIBRATION OF CAPACITY MODELS

The capacity models presented in Section 3 can be gencralized as the
Sicgloch madel (17) by using the expressions in Equation 22-21 through Equation
22-23.
epce -- Ae(-BVc) Equation 22-21

A =_3,_60_0 Equation 22-22

',-(',/2)
"
B=---- Equation 22-23
3,600
wherc Field measures of cntiCa}
headway and fot!ow-up
lane capacity, adjusted for heavy vehicles (pc!h); headway can be used lO
calibrate the caMCity modeJs.
Ve conflicting flow (pc/h);
( critical headway (s); and
tI = follow-up headway (s).
With this formulaban, the capacity model can be calibrated by using two
parameters: the critical headway te and the follow-up headway tI'
Research (1) has fuund that a reasonablc calibrabon can be made by using
unly field measurements of follow-up headway to calculate the intercept A and
retaining the value for B.This procedure recognizes thc difficulty in measuring
critical headway in the field.
Examplcs illustrating these calibration procedures are provided in Chapter
33, Roundabouts: SupplementaL

Chapter 22fRoundabouts Extension to ttle Motorized Vehicle Methodology

Vt'l5ioil 6. O Page 22-25
Highway Capacity Manual: A Guide tor Multimodal Mobility Analysis

5. PEDESTRIAN MODE

limited rescarch has been performed in the United States on the operational
impacts of vehicular traffic on pedestrians at roundabouts. In the United States,
pedestrians have the right-of-way either aftcr entering a crosswalk or as they are
about to enter the crosswalk, depending on specific state law. This type of
pedestrian right-of-way is somewhat differcnt from those in other countries that
may establish absolute pedestrian right-of-way in some situations (typically
urban) and absolute vehicular right-of-way in others (typically rural).
Much of the recent research on pedestrians in the United States has focused
on asscssing accessibility for pedestrians with vision disabilities. Research has
found that sorne roundabouts prescnt a challenge for blind and visually
impaired pedestrians relative to sighted pedestrians, thus potcntially bringing
them out of compliance with the Amcricans with Disabilities Act (10). Various
treatments have been or are being considered to improve roundabouts'
accessibility to this group of pedestrians, including various types of signalization
of pedestrian crossings. The analysis of these treatments can in sorne cases be
performed by simple analytical methods presented in the HCM (e.g., the analysis
procedure for the pedestrian mode in Chapter 20). Howevcr, in many cases,
altcmative tools will produce more accurate results. Thesc are discussed in
Section 7, Applications.
Techniques to analyze the operational performance of pedestrians as
provided in Chapter 20, Two-Way STOp-Controlledlnterscctions, can be applied 1
with care at roundabouts. As noted in that chapter, vehicular yielding rates vary
dcpending on crossing treahnent, number of lanes, posted speed limit, and
within individual sites (18). This variation makes modeling of pedestrian
interactions imprecise. As a result, models to analyze vehicular effects on
pedestrian travel should be applied with caution.

Pedestrian Mode Chapter 22jRoundabouts

Page 22-26 Vef5ion 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

6. BICYCLE MOOE

As of thc publication date of this edition of the HCM, no methodology Use a passenger-c;Jr-equiva!ent
factor of 0.5 for bicydes wñen
specific to bieyclists has been developed to assess the performance of bicyclists at treating them as motorized
roundabouts, as few data are available in the United States to support model vehides.

calibration. Depending on individual comfort level, ability, geometric conditions,

and traffic conditions, a bieyclist may either circulate as a motorized vehicle or as
a pedestrian. If bieyclists are circulating as motorized vehicles, their effect can be
approximated by combining bieyclist flow rates with other vehicles by using a
passenger.car-equivalent factor of 0.5 (7). If bicyclists are circulating as
pedestrians, their effect can be analyzed by using the methodology described
previously for pedestrians. Further guidance on accommodating bicyclists at
roundabouts can be found elsewhere (7).

Chapter 22jRoundabouts Bicycle Mode

Version 6.0 Page 22-27
Highway Capacity Manual: A Guide lar Multimadal Mobilify Analysis

7. APPLICATIONS

TYPES OF ANALYSIS
This chapter's methodology can be used in three types of analysis:
operational analysis, design analysis, and planning and preliminary engineering
analysis.

Operational Analysis
OperationaJ afliJlysis takes
The methodology is most easily applied in the operational analysis mode. In
traffic flow data and geometric
contigllrations as input lo operational analysis, aH traffie and geometric eharaderistics of the analysis
determine operationaJ segment must be specified, induding analysis-hour demand volumes for each
pe<fo<ma~.
tuming movement (in vehides per hour), heavy-vehide percentages for each
approach, peak hour factor for aHhourly demand volumes (if not provided as
15-min volumes), and lane eonfiguration. The outputs of an operational analysis
will be estima tes of capacity and control delay. The steps of the methodology,
described in the Methodology section, are followed directly without
modification.

Design Analysis
Design analysis determines the
geometric conflguration ola
The operational analysis methodology described earlier in this chapter can
roundabout lo produce a be used for design purposes by using a given set of traffic flow data to determine
desired operationaf
iteratively the number and configuration of lanes that would be required to
pe<fo<ma~.
produce a given LOS.

Planning and Preliminary Engineering Analysis

Planning and prelimifliJry
engineering anafysis is lJsed lo
The operational analysis method described earlier in this chapter provides a
evafuate roture con(!Jtjoos for detailed procedure for evaluating the performance of a roundabout. To estimate
which assumptions and
estimates mlJ5t be made.
LOS for a futore time horizon, a planning analysis based on the operational
method is used. The planning method uses aH the geometric and traffie flow data
required for an operational analysis, and the computations are identical.
However, input variables for percentage of heavy vehides and peak hour factor
are typically estimated (or defauIts are used) when planning applications are
performed.

EXAMPLE PROBLEMS
Section 3 oE Chapter 33, Roundabouts: Supplemental, provides two example
problems that go through each of the computational steps involved in applying
the motorized vehide method:
1. Analyze a single-Iane roundabout with bypass lanes, and
2. Analyze a multilane roundabout.

EXAMPLE RESULTS
Analysis of roundabouts is commonly performed as part of an alternatives
analysis with STOP-controlledor signalized alternatives to determine the most
appropriate intersection form and control. These treatments, induding geometric
modifications and changes in traffic control, are discussed in úther references,

Applications O1apter 22/Roulldabouts

Page 22-28 Vershn 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

including the presentation of traffie signal warrants in the MmlUaI on Ulliform

Traffie Control Deviees for Streets and Highways (9).
In evaluating the overaIl performance of roundabouts, it is important to
consider measures of effectiveness in addition to control delay, sueh as volume-
to-capacily ratios for individual lanes, average queue Jengths, and 95th pereentile
queue lengths. By foeusing on a single measure of effectiveness for the worst lane
only, usees may make less effective traffic control decisions. The analyst using
HCM or other operational analysis mcthods should carefuIly balance the
operational effect for motor vehicles of a change in lane configuration computed
with other performance measures that may be important in the overall evaluation;
these may indude the operational performance of other modes, safety
performance, impacts to the built and natural environment, and Iife-cyeIe eosts.
An example of this consideration occurs in determining the appropriate lane
configuration for a roundabout. Consider as an example a roundabout at the
intersection of a four~lane major street (such as a state highway) with a two-Iane
minor street (such as a local street). At a roundabout, eaeh minor-street entry
would typically have a single lane opposed by two contlicting circulating lanes.
Under typical traffie tlows for a configuration Iike this, the major-street traffic
may be much heavier than the traffic on the minor street, resulting in low entry
tlows and high circulating flows for each of the minor-street entries. As the right
side of Exhibit 22-22shows, the high conflicting flows result in a low entry capacity.
Consequently, a smalI change in entering tlows can result in a large change in the
volume-to-capacity ratio for the entry, as weH as corresponding inereases in
control dela)' and queues, Inereasing the entry from one to two lanes can reduce
the volume-to-capacity ratio and associated control delays and queues for the
motor vehicle mode for this minor-street entry. However, this lane increase comes
at a potential cost to other performance measures, including safety performance
for all modes, control del ay for pedestrians, aceessibility for pedestrians, and
other costs and impaets. The analyst should carefully balance these trade-offs to
determine the appropriate eourse of action and not rely exclusivel)' on the
operational performance of motor vehides to make this decision.

•••• Exhibit 22-22

II1ustration of capacity fof
I I low Entry Aows alld High
'.'"
,.>00 L~ Circulating Aows

j-
-1- Capaoty of _
l.'Illry"9""'"
lMoe of Iw<>-laoe
""" oonnlCll"lll.'loe

o
o 600 800 1,000 1,200 1,400 1.600 1.800 2,000 2,200
"" Conflicting Flow Rate (pc/h)

Chapter 22{Roundabouts Applieations

Version 6.0 Page 22-29
Highway Capacity Manual: A Guide for Multimoda/ Mobilify Analysis

8. REFERENCES

SOme of these references are 1. Rodegerdts, L. A., A. Malinge, P. S. Marnell, S. G. Beaird, M. J. Kittelson, and
avaifable in the Technicaf
Reference Library in Vofume 4. Y. S. Mereszezak. Assessment o/ ROlmdabouf Capacify Models /or the Highway
Capacity Manual: Volllme 2 o/ Acce1erafing ROlllldabouf Implementation in the
United Statcs. Report FHWA-SA.15.070. Federal Highway Administration,
Washington, D.C., Sept. 2015.
2. Rodegerdts, L., M. Blogg, E. Wemple, E. Myers, M. Kyte, M. P. Dixon, G. F.
List, A. Flannery, R. Troutbeck, W. Brilon, N. Wu, B. N. Persaud, C. Lyon, D.
L. Harkey, and D. Cartero NCHRP Report 572: Roundabollts in the United States.
Transportabon Research Board of the National Aeademics, Washington,
D.e., 2007.
3. Ak~lik, R. Roundabouts with Unbalaneed Flow Patterns. Compendium o/
Technical Papas, Institute of Transportation Engineers 2004 Annual Meeting
and Exhibition, Lake Buena Vista, Fla., 2004.
4. Krogscheepers, J. c., and C. S. Roebuck. Unbalanced Traffie Volumes at
Roundabouts. In Transportation Research Circular E-C018: Fourth International
Symposium 011 Highway Capadty, Transportabon Research Board, National
Research Council, Washington, D.C., 2000, pp. 446-458.
5. Kimber, R. M. The Traffic Capadty o/ ROlwdabOllts. TRRL Laboratory Report
LR 942. Transport and Raad Research Laboratory, Crowthornc, Berkshire,
United Kingdom, 1980.
6. Guichet, B. Roundabouts in France: Development, Safety, Design, and
Capacity. Proc., 3rd International Symposium on Intersections Without Traffic
Sig"a/s (M. Kyte, ed.), Portland, Ore., 1997, pp. 100-105.
7. Rodegerdts, L., J. Bansen, C. Tieslcr, J. Knudscn, E. Myers, M. Johnson, M.
Moule, B. Persaud, C. Lyon, S. Hallmark, H. Isebrands, R. B. Crown, B.
Guichet, and A. O'Brien. NCHRP Report 672: Roundabouts: A" InJormafioflu/
Guide, 2nd ed. Transportabon Rcsearch Board of the National Academies,
Washington, O.c., 2010.
8. Brilon, W., B. Stuwe, and O. Drews. Sicherheit t/lld Leistungsftihigkeit von
Kreisverkehrspliifzen (Sa/ety and Capacity o/ Roundabouls). Research reporto
Ruhr-University Bochum, Germany, 1993.
9. Manual 011 Uniform Traffic Control Devices Jor Streets and Highways. Federal
Highway Administration, Washington, D.C., 2009.
10. United Sta tes Access Board. Pub/ic Rigllts-of-Way Accessíbility Guidelines.
Washington, O.c. Revised draft, 2005.
11. Ak~elik¡ R., and R. J. Troutbeck. lmplementation of the Australian
Roundabout Analysis Method in SIDRA. In Highway Capadty and l.evel o/
Seroice: Proceedillgs o/ (he lnternational Symposium on Highway Capadty (U.
Brannolte, ed.), Karlsruhe, Germany, 1991, pp. 17-34.

References Chapter 22fRouooabouts

Page 22-30 Version 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

12. Brilon, W., N. Wu, and L. Bondzio. Unsignalized Interscctions in Germany-

A State of the Art 1997. Proc., 3rd Internatiollal Symposillm 011 IllterseCfions
Wífhollf Traffic Signals (M. Kyte, ed.), Portland, Ore., 1997, pp. 61-70.
13. Wiedemann, R. Simulafion des Verkchrsflllsses (SíllllllatiOII ofTraffic FIows).
SchriftenreiBe des Instituts für Verkehrswescn, Heft 8, Karlsruhe University,
Germany, 1974.
14. Ouncan, G. Paramícs Technícal Report: Car-FoIlowing, Lane-CllQnging and
¡lInction Modelling. Quadstone, Ud., Edinburgh, Scotland, 1998.
15. Zegeer, J. D., M. A. Vandehey, M. BIogg, K. Nguyen, and M. Ereti. NCHRP
Report 599: Default Vallles for Híghway Capacity and LevcI of Seroíce Analyses.
Transportation Researeh Board of the National Aeademies, Washington,
D.C., 2008.
16. Kimber, R. M., and E. M. Hollis. Traffic Quelles and Delays at Road jIIllCtíOIlS.
TRRL Laboratory Report LR 909. Transport and Road Rcscarch Laboratory,
Crowthome, Berkshire, United Kingdom, 1979.
17. Siegloch, W. Die Ú'istungsermíttlllng all Kllofenpunkten Ohlle
Líc1lfsigllalsfeuerung (Capacífy CalculatiOIlSfor Unsígnaliud lntersections).
Schriftenreihe Strassenbau und Strassenverkehrsteehnik, Vol. 154, 1973.
18. Fitzpatrick, K., S. M. Turner, M. Brcwer, P. J. Carlson, B. Ullman, N. O. Trout,
E. S. rark, J. Whitaere, N. Lalani, and O. Lord. TeRP Report 112/NCHRP
Reporf 562: lmprovillg Pedestrían Safefy af Unsignalized Crossings.
Transportation Rcsearch Board oi the National Academics, Washington,
D.C., 2006.

Chapter 22/Roundabouts References

Version 6.0 Page 22-31
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

CHAPTER 23
RAMP TERMINALS ANO ALTERNATIVE INTERSECTIONS

CONTENTS

Part A: Distributed Intersection Concepts 23-1

l. INTRODUCTION 23-1
Overvie\'.' 23-1
Chapler Organization 23-2
Related HCM Content.. 23-2

2. CONCEPTS 23-4
Types of Interchanges and Altemalive Intersections 23-4
Unique Attribules of Interchanges and Altemative Intersections 23-5
Comparing Interchange and Intersection Evaluations 23-10
Spabal and TemporaI LimiIs 23-11
LOS Frame\\'ork 23-12

Part B: Interchange Ramp Terminal EvaJuation 23-19

1. INTRODUCTJON 23.19
Overview 23-19
Part Organization 23.19

2. CDNCEPTS 23-20
Types of Interchanges 23-20
o-D and Tuming Movcments for Conventional Intcrchanges 23.25

3. CDRE METHODOlOGY 23-26

Scope of the Mcthodology 23-26
Requircd Data and SOurces 23-28
Overview . 23-30
Compu lalional Stcps 23-31

4. EXTENSIONS TO THE METHODOLOGY 23.57

Final Dcsign and Operational Analysis for Interchanges with
Roundabou ts 23-57
Interchanges with Unsignalized lntersections 23-57
Estimating Proportion of Time Blockcd for an Isolated DDI... 23-57
Pedcstrian and Bicycle Analysis 23-60

Chapter 23/Ramp Termínals and A1temative Intersectíons Contents

VersiQn 6.0 Page 23-í
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

5. APPLICA TIONS •..••...•...•...••..••..••....•......................................•...•..•.••..•..•.

23-61
Example Problems 23-61
Final Dcsign and Operational Analysis 23-61
Operational Analysis for lnterchange Type Selection 23-62
0-0 and Turning Movemenl Estimation 23-62
Use of Altemative Tools 23-62

Part C: Altemative Intersection Evaluation 23-65

1. INTRODUCTION ...•...•...••...•...•...••..••..................................................•....
23-65
Overview 23-65
Part Organization 23-65

2. CONCEPTS ...........•........••...•...•...••...•....•..•.................................................
23-66
Reslricted Crossing U-Turn and Median U-Turn lntersections 23-66
Displaced Left-Turo 1ntersections 23-68

3. CDRE METHODDlOGY •...••..•...•....•......................................................23-70

Scope of the Methodology 23-70
Required Data and SOurces 23-71
Overview of the Methodology 23-72
Computational Steps.............. . 23-73

4. EXTENSIONS TO THE METHODOLOGY 23-87

Evalualion of Other Alternative Intersection Forros 23-87
Pedestrian and Bicyclc Analysis 23-88

5. APPLICA TIONS .•...••..••...•...••..••...•...•...•...................................................

23-91
Example Problems 23-91
Example Resulls 23-91
Planning-Leve I Analysis 23-94

6. REFERENCES ....................................•..••..••...•...••..••...•..••..•...•...•...•..••..•...•
23-96

Contents Olapter 23/Ramp Terminals and A1temative Intersections

Page 23-ii V~6.0
 Highway Capacity Manual: A Guide for MultjmodaJ Mobility Analysjs

LIST OF EXHIBITS

Exhibit 23-1 Types of Intersections and Alternative Interchanges 23-4

Exhibit 23-2 Impact of Interchange Type on Lane Utilization 23-8
Exhibit 23-3 Demand Starvation at the Internal Link of a Diamond
Interchange 23-10
Exhibit 23-4 Summary Comparison of Intersection and lnterchange
Proeed ures 23-11
Exhibit 23-5 Example of Spatial Limits for an RCUT Intersection 23.12
Exhibit 23-6 Example of Experienced Travel Time at an RCUT
Intersection 23-13
Exhibit 23-7 II1ustration of the LOS Concept at a Diamond Interchange 23-14
Exhibit 23-8 mustration of the EDIT Concept at a Diamond Interchange 23-14
Exhibit 23-9 llIustration of the EDIT Concept at a ParcIo A.2Q
Interchange 23.15
Exhibit 23.10 LOS Criteria for Each O-O Within Signalizcd Interchangcs 23-16
Exhibit 23-11 Il1ustration of the EIT Concept at a ODI 23-16
Exhibit 23-12 Il1ustration of the EIT Concept at a Displaced Left-Turn
Intcrsection 23.17
Exhibit 23-13 LOS Criteria for Each 0-0 Within Alternative Interscctions 23-17
Exhibit 23-14 LOS Criteria for Each O-O of an Interchange with
Roundabouts 23.18
Exhibit 23-15 Types of Oiamond Interchanges 23-21
Exhibit 23.16 IiIustration of a 001 23-22
Exhibit 23-17 Types of ParcIo Interchanges 23-23
Exhibit 23-18 Single-Point Urban Interchange 23-23
Exhibit 23-19 Diamond Interchanges with Circular Ramp Terminals 23-24
Exhibit 23-20 0-0 Flows for Each Interchange Configuration 23-25
Exhibit 23-21 Summary of Required Input Data for Final Design and
Operational Analysis of Signalized Interchanges 23.29
Exhibit 23-22 Interchangc Ramp Terminals Methodology: Final Design
and Operational Analysis for Interchangcs 23-30
Exhibit 23-23 Adjustment Factor for Traffic Pressure (Jo) 23-33
Exhibit 23-24 Parameters for Lane Utilization Models for the ExternaI
Arterial Approaches of Oiamond and ParcIo Interchanges 23.35
Exhibit 23-25 Five Categories for 001 Lane Utilization 23-36
Exhibit 23-26 Lane Utilization Model Coefficients for 001s 23.36
Exhibit 23-27 Adjustment Factor for Turn Radius (fR) 23-38
Exhibit 23.28 IiIustration of Common Green Times 23-40

Chapter 23/Ramp Termínals and A1temative Intersections Contents

Versicn 6.0 Page 23.iii
Hjghway Capacity Manual: A Gujde for Multimodal Mobility Analysis

Exhibit 23-29 Traffic and Geometric Oata for Additional Lost Time at
OOIs 23-43
Exhibit 23.30 Standard Phasing Scheme at a DOI with Signalized Ramp
Movements 23-45
Exhibit 23-31 Graphical Depiction of the Oistances Needed to Calculate
Off-Ramp Lost Time at OOls 23-45
Exhibit 23-32 Lost Time for DOI Off-Ramp Based on Oistance Terrns 23-46
Exhibit 23-33 IlIustration of Interval with Demand Starvation Potential 23-48
Exhibit 23.34 Conflicting Flow Regimes IIIustrated for DOI 23-51
Exhibit 23-35 Estimation of Blocked Period Duration 23-52
Exhibit 23.36 Default DOI Tum Calibration Parameters 23-53
Exhibit 23-37 Plots of Gap Acceptance Capacity Models 23.53
Exhibit 23-38 Queuing Representation as an Approximation ofTime to
Clear Conflicting Queue for Random Arrivals 23-58
Exhibit 23.39 Queuing Representation as an Approximation of Time to
Clear Conflicting Queue for Coordinated Arrivals 23.58
Exhibit 23-40 Listing of Interchange Example Problems Contained in
Chapter 34 23-61
Exhibit 23-41 Four-Legged RCUT with Signals 23.66
Exhibit 23-42 Four-Legged RCUT with Merges and Diverges 23-67
Exhibit 23-43 Three-Legged RCUT with Signals 23-67
Exhibit 23-44 Four-Legged MUT with Signals 23.67
Exhibit 23-45 Roadway Geometry for Full and Partial DLT Intersections 23-68
Exhibit 23-46 Lane Geometry for DLT Intersections 23-69
Exhibit 23-47 Operational Analysis Framework for Altemative
Intersections 23-72
Exhibit 23-48 Junctions and Extra Travel Time Segments at a Four-
Legged RCUT 23.75
Exhibit 23-49 Junctions and Extra Travel Time Segments at a Three-
Legged RCUT 23-75
Exhibit 23-50 Junctions and Extra Travel Time Segments at an MUT 23-76
Exhibit 23-51 Default Arrival Typcs for RCUT and MUT Movements
Encountering More Than One Signal 23-77
Exhibit 23-52 MUT and RCUT Default Saturation Adjustment Factors for
U-Tum Crossovers 23-78
Exhibit 23-53 Urban Street Layout far a Partial DLT lntcrsection 23.81
Exhibit 23-54 Urban Street Layout for a Full DLT Intersection 23-84
Exhibit 23.55 Example Conversion of Displaced Left Tums to Pseudo-
Right Turns 23-85

Contents Chapter 23/Ramp Terminals and A1temative lntersed;ioll5

Page 23-iv Verron 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Exhibit 23.56 Side Street at the End of an MUT U-Tum Crossover in

Michigan 23-87
Exhibit 23.57 Typical Pedestrian Crossing of a Four-Legged Signalized
RCUT 23-88
Exhibit 23-58 Typical Pedestrian Crossing of a Three-Legged Signalized
RCUT 23-89
Exhibit 23-59 Two-Stage Pedestrian Crossing at a DLT 23-90
Exhibit 23.60 Listing oí Altemative Intersection Example Problems
Contained in Chapter 34 23-91
Exhibit 23-61 Example Results Comparing DLT and Conventional
lntersection Performance 23-91
Exhibit 23.62 Optimized Cyc1e Lengths íor DLT and Conventional
lntersection Configura tions 23-92
Exhibit 23-63 Sensitivity of EDTI and LOS to Distance to U-Tum
Crossover and Free-Flow Speed 23-93
Exhibit 23-64 Sensitivity of LOS to Changes in Minor-Street Demand 23-94
Exhibit 23-65 Sensitivity of ETI to Percentage of Minor-Street Traffic
Tuming Right on Red 23-94

O1apter 23/Ramp Terminals and A!temative Intersections Contents

Version 6.0 Page 23-v
 Highway Capacity Manual: A Guide for Multlmodal Mobility Analysis

Part A: Distributed Intersection Concepts

1. INTRODUCTION VOLUME 3: INTERRUPTED FLOW

16. Urban Street FaCilities
17. Urban Street Reliability and
ATOM
OVERVIEW 18. Urnan Street Segments
19. Signalized Intersections
This chapter presents a methodology for the analysis of interchange ramp 20. TWSC Intersections
21. AWSC Intersections
terminals, alternative intersections, and alternative intcrchanges. The interchange 22. Roundabouts
ramp terminal methodology was developed primarily on the basis of research 23. Ramp Terminals and
Altemative Intersections
conducted through the National Coopcrative Highway Research Program (1-3) 24. Off-Street Pedestrian and Bicyde
and elsewhere (4). The altemative interseetion and interchange methodology was Facilities

bascd on research condueted through the Federal Highway Administration

(PHIV A)(5).

Interchange ramp terminals are critical components of the highway nctwork.

They provide the connection between various highway facilities (freeway-
arterial, arterial-arterial, ete.), and thus their effident operation is essential.
Interchanges are typically designed to work in harmony with the frccway, thc
ramps, and the arterials. In addition, they need to provide adequatc capacity to
avoid affecting the connecting facilities.
Alternative intersections are created by rerouting one or more movements
from their usual places to secondary junctions. Dften, the rerouted movements
are left toms. Altemative interscction and interchange designs have significantly
reduced travel times and delays in many areas, compared with conventional
intersection designs. Sorne designs have substantially reduced the numbcr of
conflict points between vehicles and thus increased overall safety.ln addition,
the altemative designs can often be implcmentcd with minimal disruptions to
existing right-of-way. Given the relatively low cost of implementation far many
of these designs, the combination of improved mobility and safety has produccd
outstanding henefit-eost ratios within economic analyses. By relocaling or
eliminating problematic movements, the alternative designs can efficiently
mitigatc congcstion at surface slreel-freeway interchanges and at signalized
intersections.
Both interchange ramp terminals and altcrnative at-gradc intersections are
discussed in this chapter, because thcy combine multiple intersections in a
cluster. "Distributed intersections" consist of groups of two or more intersections "Distributed intersections
6

cvnsist of groups of two or

that, by virtue of close spacing and displaced or distributed traffic movements, more intersections Chal, by
are operational1y interdepcndent and are thus hest analyzed as a single unit. The virtue of dose spaang and
displaced or distributed tTaffic
mosl common distributed interscctions are interchange ramp terminals, but movements, are operafion.ally
other alternative interscction forms-such as those involving displaced left-tum interdependent and, tf/us, bes!
anafyzed as a singie uni!.
movements-also faHinto this category.
Research has not yet been performed on pcdestrian and bicyclist perceptions
of service quality spcdfic to interchange ramp terminals and alternative
intersections. Therefore, no service measures are provided in this chapter far
those modes. Several FHWA publications (e.g., 6-8) discuss designing for non-
automobile users of alternative interscctions and interchanges.

Olapter 23/Ramp Temlinals and A1temative Intersections Introduction

VetSion 6.0 Page 23-1
Highway Copocity Manual: A Guide tor Multimodal Mobility Analysis

CHAPTER ORGANIZATION
Part A of Chapter 23 provides an overview of altemative intersection and
interchange concepts. Within this part, Section 2 documents and describes a
number of common concepts associated with interchanges and altemative
intersections. This section Iists the unique elements and summarizes the shared
attributes of such facilities. It further discusses the need for translating between
turning movemenf volume demands at each intersection approach and origin-
destjnalioll demands across the enfire intersection or interchange. The section
discusses issues related to distributed intersections and interchanges, induding
an origin-destination framework. To facilitate unbiased comparisons among
distributed intersection types, this section introduces a discussion of experienced
travel time and delay-eonsisting of diverted path delay and control delay.
Part B of Chapter 23 focuses on the evaluanon of surface street-freeway
interchanges. Following the Section 1 overview, Section 2 describes the features
of diamond interchanges, partial doverleafs, single-point urban interchanges,
diverging diamond interchanges, roundabout interchanges, and others. Section 3
disCllSseSthe core evaluation methodology, including scope, required data, and
computational steps. Section 4 describes methodology extensions for
interchanges with roundabouts and interchanges with STOPand YIELDsigns, and
it describes a specific procedure for interchange type selection. Section 5 presents
applications of the Part B methodology, induding example results, analysis
types, and the pros and cons of analyzing surface street-freeway interchanges
with altemanve tools.
Part C of Chapter 23 focuses on the evaluanon of altemative intersections.
After the Section 1 overview, Section 2 describes the features of restricted
crossing U-tum intersections (also known as superstreets), median U-tum
intersections, displaced left-tum intersections (also known as continuous flow
intersections), and others. Section 3 discusses the core evaluation methodology,
induding scope, required data, and computational steps. Section 4 describes
methodology extensions for altemative intersection designs not covered in
Section 3. Section 5 prescnts applications of the Part C methodology, induding
example results, analysis types, and the pros and coos of analyzing altemative
intersectians with altemative tools.

RELATEO HCM CONTENT

Other Highway Capacity Manual (HCM) content related to this chapter
indudes the foIlowing:
• Chapter 4, Traffic Operations and Capacity Concepts, explains sorne of
the fundamental concepts behind capacity analysis of interrupted-flaw
facilities, which encompass all intersections and interchanges. The chapter
discusses capacity, volume, headway, stops, queuing. density, flow rate,
lost time, control delay, saturanon flow, peak hour factors, and different
types oí speeds.
• Chapter 5, Quality and Level-of.Service Concepts, discusses methads for
assessing quality of service and level af service (LOS) for all surface and
freeway facilities.

Introduction Chapter23/RampTerminalsaOOA1ternative
Intersections
Page 23-2 Vef3ion 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

• Chapter 19, Signalized Intersections, contains detailed procedures and

discussions for signalized intersection analysis. Many of the concepts
introduced in this chapter establish the baseline for analysis of complex
intersections and interchanges. These concepts indude Jane group
determination, signal timing determination, saturation f10wrate
adjustrnent, Jane utilization, and control delay estimation.
• Chapter 18, Urban Street Segments, extends the Chapler 19 signalized
intersection procedures and discussions to account for adjacent
interscction effects. Sorne of these concepts are also prerequisites for the
anaJysis of complex intersections and interchanges. These concepts
indude flow profile determination, coordinated signaJ timing. travel time
estimation, and estimation of the proportion of vehides arriving on green.
• Chapter 20, Two.Way STOP-Controlled Intersections, contains detailed
procedures and discussions for two-way STOP-controlledinterscction
analysis. Concepts from this chapter, such as criticaJ headways and
follow-up times, are applicable to STOP-controlledlocations within
altemative intersections and interchanges.
• Chapter 34, Interchange Ramp Terminals: Supplemental, contains
example problems and origin-destination volume worksheets. Chapter 34
also presents a sketch-planning method foc inlerchange type selection.

Chapter 23jRamp Terminals and A1temative Intersectioos Introductlon

Ver.s.0n 6.0 Page 23-3
 Highway Capacity Manual: A Guide for Multimoda/ MObility Analysis

2. CONCEPTS

TYPES OF INTERCHANGES ANO ALTERNATIVE INTERSECTIONS

Exhibit 23-1 iIIustrates the various types of interchanges and altemative
intersections addressed in this chapter. Note that this exhibit only provides one
example of each intersection and interchange type. Many of these intersection
and interchange types have several variations, which wiII be detailed later.

Exhibit 23-1
Types of Intersections and
Altemative Interchanges PARCLO DIAMOND

SPUJ DDI

Dlr RCUT

Mur

Note: Gray hnes represent the freewi1'f, which is not analyled as part of the methodology In this chapter.

eo"",,,,, Chapter 23/Ramp Terminals and Altemative Intersections

Page 23-4 V~6.0
 Highway Capacity Manual: A Guide far Multimodal Mobility Analysis

These interchange and alternative intersection types are as follows:

• Oiamolld interchanges: Ramps from both freeway directions intersect the
cross strect at separa te, but often dosely spaced, intersections.
• Partial cloverleaj(parclo) interchallges: Parclos contain one or two loop
ramps that may intersect the crossroad in a manner similar to a diamond
rampo
• Divergillg diamond interchanges (OOls): Similar in configuration to a
diamond-type interchange, but with a crossover at each intersedion
rearranging traffic on the cross street, to reduce conflicts for left-turn
movements.
• Single-point urban interchanges (SPUls): Allleft turns to and from the
fret'way meet at a single intersection on the cross street.
• Median U-tum (MUT) illtcrsections: At-grade intersections at which major-
and minor-street left turns are rerouted. Minor-stret't through movements
are not rerouted.
• Resfricled crossillg U-tum (RCUT) illtersecfio/ls: At-grade intersections at
which minor-street left-tum and through movements are rerouted. Major-
street left turns are not rerouted.
• Displaced {eJt-tllm (DLT) interseclions: At-grade intersections where left-
tuming vehic1cscross opposing through traffie before reaching the main
intcrsection, thus rcducing conflicts at the main intersection.

UNIQUE ATTRIBUTES OF INTERCHANGES ANO ALTERNATIVE

INTERSECTIONS
Interchanges and alternative intersections sharc a number of characteristics,
and they have a number of important differences. Understanding the similaritics
and differences can be helpful in choosing a configuration or in evaluating
operational efficicncy. This section describes the common attributes and key
diffcrences at such facilities. The common attributes, induding origin--destination
demands, intersection spadng. signal coordination, dcmand starvation, and lane
utilization on the internal and external approaches, are discussed first.

Influence of Configuration on Turning Movements

The interchange configuration or intersection configuration has a major
influcnce on turning movements. Movements that involve a right-side merge in
one configuration may become left tums in another. Movements approaching on
the surfacc facility may be affected by configuration, depending on whether
ramp movements involve left or right tums. Thus, the lane utilization on external
approaches to the interchange varies as a fundion of configuration and the
relative pro portian of tuming movements at the downstream intersection.

Innuence of Interchange Configurations

In selecting an appropriate type of interchange, impacts on thc tuming
movements should be considered. Left-turning movements are usually the most
difficult in terms of efficiency of operation. Left tums from the fret'way to the

Olapter 23/Ramp Terminals and A1tematiye IntersectiOOs Concepts

Version 6.0 Page 2).5
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

cross street at a signalized diamond interchange, for example, conflict with both
directions of the cross street at one ramp terminal and then must traverse the
signal at the other ramp terminal. Selection of a type of interchange that does not
route high-demand movements through mulliple conflict points can enhance the
overall flow of vehicles significantly. However, this cannot always be
accomplished. Right-of-way limitations or agency polides may predude the use
of loop ramps, and economic and environmental constraints may make
multilevel structures impractical. Many of the operational effidency henefits (as
well as safety benefits) derived from alternative interchange designs (such as the
OOI) and altemative intersection designs can be attributed to the relocation of
high-demand left-tuming movements to avoid conflicts.
Because of the influence of interchange type on turning movements and the
nred to compare various interchange types, LOS far interchange ramp terminals
and altemative intersections uses origin-destination (0-0) demands as inputs,
since they are identical regardless of the interchange or intersection type. The
methodologies in this chapter use both O-O demands and turning movement
demands; one set of demands can be derived from the other.

lnfluence of Altemative lntersection Configurations

The deñning charaderistic of
the altemative intersec:tions
The defining characteristic of the alternative intersections presented in this
presented in this chapter is a chapter is a rerouting of one or more movements from the center of the
rerouting of one or more
movements from the renter of
intersection, which reduces conflicting flows at the main intersection. One of the
the intersection, whictl reduces important implications is that the tuming movement pattems change from a
(X)Ilflicting flows at the main
in~.
traditional intersection. For example, at a four-Iegged RCUT, the minor-street
left-tum movement must first tum right at the main junction. Then it must drive
to the U-tum crossover, execute a U-turo, drive back to the main junction, and
finally make a through movement. At that RCUT, the minor-street left-turn
traffic volume will appear in a transformed tuming movement diagram as a
right turn at the main junction on the near side (c1osest to the origin), as a U-tum
at the crossover, and as a through movement at the main junction on the far side.
A similar accounting of aH demands must be made for aH movements at RCUTs,
MUTs, and DLTs. Part e of this chapter shows in detail how to make those
transformations for all designs covered.
Rerouting movements at altemative intersections creates junctions in
addition to the main junction. At a four-legged MUT, for example, there are three
junctions: the main intersection and the ends of the two U-tum crossovers.
Vehicles that are rerouted typically must negotiate more than one junction and
may experience control delay at each junction. The operationaI analysis
methodoIogy for each type of aItemative intersection accounts for each source of
control deIay for each rerouted movement in the calculation of performance
measurcs and LOS estimates.
At DLT intersections, displaced left-tuming vehicles typically experience
continuous flow on arriving at the main intersection because they do not yield to
opposing through vehicles. Instead, they move together with opposing through
vehicles, and signal timing offsets aUow them to arrive during a window of
available green time. For compatibility with Chapters 18 (Urban Street Segments)
and 19 (Signalized Intersections), which do not allow protected left-tuming

00_"
Page 23-6
O1apter 23jRamp Tenninals and Altemative Intersections
VetSioo 6.0
 Highway Capacity Manual: A Guide tor Multimodal Mobility Analysjs

vehides to move simultaneousll' with opposing through vehicles, the DLT

anall'sis procedure in Part C assumes that displaced left-turning demand
volumes at the main intersection are egual to zero.

Operational Effects of Intersection Spacing

lntersection and interchange configurations with closell' spaced signaJized
intersections present unique challenges because the intersections do not operate
in isolation. The distance between intersections can limil the available storage for
internalll' queued vehides. The presence of a downstream gueue mal' reduce or
completell' block the discharge from the upstream intersection. Intersection
spacing also affects the proportion of vehides arriving on green and the
possibility of unused green time at the downstream intersection.

Queue Interaction
Queued vehides within a short
Queued vehides within a short segment (or link) Jimit the effective length of segment (01"iink) limit the
the link, and vehides can travel freell' onll' from the upstream stop line to the effective lengttl of!he iink, and
vehides an travel freeIy onfy
back of the downstream queue. Because this distance mal' be smal1, the impact from the upstream stop line to
on the upstream discharge rate is significant. In this methodologl', the effects of the back of the downstream
q_.
the presence of a queue at the downstream link are considered by estimating the
amount of additionallost time experienced at the upstream intersection. The
additionallost time is calculated as a function of the distance to the downstream
queue at the beginning of the green for each of the upstream phases.
The extent of queuing at the downstream intersection depends on several
factors, induding the signal control at the upstream and downstream signals, the
number and use of lanes at both intersections, and the upstream flow rates that
feed the downstream intersection. Some of these effects mal' also exist at
locations where signalized intersections are dosely spaced, particularly where
heavy left-turn movements existoThis chapler addresses the interactions of This chapter addresses the
inrer-actions of interchange
interchange operations with those of adjacent dosell' spaced signalized operations wi!h those of
intersections. Furthennore, the principies described in this chapter mal' be adjacent doseIy spaceri
sigf1iJfizedinterseetions. me
applied to similar situations in which dosely spaced signalized intersections principies desmbed in thiS
(other than those at interchanges) interact. chapter may be applied ro
sJinilar situations in which
Similar issues mal' exist at interchanges with roundabouts that are near cJoseIyspaced SiglJiJlized
intersections (other than those
signalized intersections, or at other intersection or interchange forms with at interdJanges) interact.
unsignalized movements. When the queue from a signalized intersection reaches
upstream to a roundabout or unsignalized intersection, the upstream operations
might be significantly affected. For a roundabout, this queue spiUback mal' cause
gridlock because all movements through the roundabout must use the
drculating roadwal" The HCM computational procedures do not address
spillback between interchanges and nearbl' facilities.

Signa! Progression and Lost Time

The operational efficiency of distributed intersections is dependent on the
spacing of junctions in several other wal's. First, junction spacing mal' allow
excellent signal progression through multiple signals, or it mal' mean that no
progression is feasible. Second, facility performance will be affected if vehides

Chapter 23{Ramp Tenninals and Altemative Intersections Coo<epts

V~6.0 Page 23-7
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

are required to travellonger distances out of their desired paths. Finally, junction
spacing may affect signallost time, such as at a DDI or MUT.

Lane Utilization Effects

Lane utilization is the extent to which lanes are used equaJly (or unequally)
by drivers. The presence of multiple intersections operating as a single unit can
strongly influence drivers' choice of lanes when they approach an upstream
interscction. At interchanges, this can mean that through-lane utilization at the
upstream intersection reflects desired tom movements at the downstream
intersection. Likewise, at MUT and RCUT intersections, this can mean that dual
righHurn lane utilizations reflect downstream movements, with drivers headed
for the U.turn crossover using the leftmost of the side street righHurn lanes.
For two-intersection signalized interchanges, the lane utilization for external
through movements approaching the interchange on the surface facility is
significantly affected by the direction and demand of turning movements at the
downstream interscction. As shown in Exhibit 23-2, significant left-turning
demand onto the freeway can lead to highly imbalanced lane utilization on the
external approach. Drivers wanting to make a downstream left tum will typically
pre-position their vehicles in the leftmost lane(s), in anticipation of the
downstream movement. Conversely, heavy-volume downstream right turos will
gravitate toward the right side at the upstream intersection. This can aeate lane-
use imbalances excceding those at single intersections.

DIAMOND
Exhibit 23-2
Impact of Interdlange Type
on Lane Utilization

Preferred Iane
fOl'downstream
Ieft tums - =
I
-
PARa.O AB-2Q

=
=
Preferred Iane =
o
fOl' downst:ream I
righttums

Coo<ept; O1apter 23/Ramp Terminals and Altemative Intersections

Page 23-8 Vel5ÍOO 6.0
 Highway Capacity Manual: A Guide for MultimodaJ Mobility Anolysis

This chapter's methodology identifies the highest lane utilization at each of

the upstream external through movements as a function of the interchange type,
the number of through lanes, the distance between the two intersections, and the
O-O demands. This chapter also considers the lane utilization of the arterial
approaches at intersections adjacent to the interchange. Lane utilization at those
intersections may be affeded by tuming movement demands at the interchange.

Traffic Control Considerations

When multiple controls (signals, YIELD signs, STOPsigns) are present within a
single intersection or set of intersections, a hybrid analysis may be required. 5uch
an analysis should utilize principies of the unsignalized and signalized HCM
chapters (Chapters 16 through 24) and should work toward obtaining a common
performance measure. Although unsignalized movements can occur at
signalized intersections, hybrid control combinations are more common at
certain distributed configurations, such as DDIs, RCUTs, and MUTs. At DDIs,
criticaI headways and follow-up times affect the analysis of free-flow and YIELD-
controlled tums. Some distributed intersections have traffic movements that
could be considered unconventional. For example, U-tum movements must be
made at RCUT and MUT intersections to reach certain destinations from certain
origins. These movements produce the need for specific analytical techniques,
which are described in Parts B and C of this chapter.
At distributed intersections with signals, the signal operations differ from Signal operations at distnbuted
conventional intersections in sorne specific ways. Signal operations at distributed intersections are different in
that they can provide, and rely
intersections are different in that they can provide, and rely heavily on, a heavllyon, a maxJlnum of one
maximum of one stop along the major arterial. This minimizes del ay and stop along fiJe majar arterial.
Tñis minimizes deJay and
effectively manages the length and location of vehicle queues. If an alternative effectively manages !:he Jength
intersection junction is signalized, the signal will almost always have two phases. and Iocation of vehiCle queues.

The common exceptions lo this are a U~turn crossover signal serving a busy hvo-
way minor street or driveway and the main signal at a partial DLT, where the
street without left-turn crossovers may need an exclusive lcft-turn phase. The
multip!e signals that are below capacity shou!d have the same eycle length (or a
variation such as half-cycle) to allow progression. At a OLT, the left-turn
crossover intersection almost always has an offset relative to the signal at the
main junction that allows most through drivers on that street to arTive on green.
At an MUT, the main junction almost always has an offset relative to the U-tum
crossover that allows most through drivers on that street to arrive on green. At
an RCUT, signals on each side of the major strcct can be timed independently,
with different cycle lengths if desired. 6ut, along each side of the major street,
signal offsets typically allow maximum bandwidths for through traffie.
Compared with conventional signalized intersection analysis, the important
mode!ing differences in this chapter relate to lane utilization, saturation flow
rate, and signal progression. Lane utilization at RCUT and MUT intersections
may differ from that at conventional intersections, because sorne drivers pre-
position themselves at one junction to get ready for the sccond junction.
Saturation flow rates for U-turns at RCUTs and MUTs differ froro those for lcft
turns at conventional intersections. Signal progression is an important feature at
RCUTs, MUTs, and DLTs, as agencies attempt to progress large portions of

Chapter 23/Ramp Terminals and A1temative Intersecbons Concepts

Version 6.0 Page 23-9
Highway Capacity Manual: A Guide tor Multimodal MobiJjty Analysis

through movements through multiple signals. The methodology presented in

Part C takes aH of these important operational characteristics into account in the
estimation of performance measures and LOS.

Demand Starvation at Conventionallnterchanges

Demand starvation occurs when a signalized approach has adequate
capadty but a significant portion oí the traffic demand is held upstream because
of the signalization pattem. For two-intersection signalized interchanges,
demand starvation occurs when a portion oí the green at the downstream
intersection is not used because the upstream intersection signalization prevents
vehides from reaching the stop lineoThus, portions of the downstream green are
unused while demand is stuck at the upstream intersection. Exhibit 23-3
illustrates the concept of demand starvation for an interchange. As shown, the
intemalleft tum in the eastbound direction is green, blocking aH westbound
vehides from reaching the westbound intemallink. Thus, demand starvation is
experienced by the intemal westbound through movement, where the signal is
green, while the demand for it is blocked upstream.

Exhibit 23.3
Demand5tarvationat the
IntemalLinkofa Diamond ~ GREEN ~
~ ~ Upstream vehlcles
Interchange -t RED U canoat enter

_
____
t- ---~ ~ __IlD0_'_=-"_
+ t-::'7"
t-OCD
W2Z
c::n CI3 ICJ----Jo.
c::w cr:m

ll--~
_ cr:D ~ lO) am ----Jo. lClD

COMPARING INTERCHANGE ANO INTERSECTlON EVALUATlONS

To follow up on the prior discussion of unique attributes of distributed
interscctions, Exhibit 23-4 summarizes their shared elernents and differences.
Exhibit 23-4 shows that multiple interchange types require specific adjustments
for lane utilization, saturation flow rate, critical headway, and lost time.
Conversely, sorne configuration type evaluations have their own unique
adjustments. These indude the assumption of zero displaced left-tuming
demands at the main intersection (DLT), U-turn lost time and saturation flow
rate adjustments (RCUT, MUT), and left turns on red (DDI). Parclo and diamond
interchanges are seen to have the same basic framework for evaluation. The SPUI
rnodeling framework is a subset of the parclo and diamond framework. These
modeling frameworks are detailed in Part B (lnterchange Ramp Terminal
Evaluation) and Part C (Altemative Intersection Evaluation).

Coocept; QlapterZ3/Ramp
Terminals
andA1temative
Intersections
Page23-10 Version 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobjlity Analysis

AdjustmentsBeyond Oia~ Par~ RCUT, Exhibit 23-4

Standard Intersection Anal sis mond do SPUI DOI OLT MUT SummaryComparisonof
VoIume Adlistment intersection and Interchange
Assumedisplacedleft-tum demandvolumeof zera Procedures
y
at the main intersectiOll
Lane UtilizatiOn Ad 'ustments
Upstream(external) through movements Y Y Y Y D
MinQ(-streetturnin movements D
Uns. nalized Control-Based Ad"ustments
Criticalheadway,follow-up time: right turn on red Y Y Y Y Y
Criticalheadway,rollow-uptime: left tum on red Y
Criticalheadway,follow-up time: U-turn on red Y
Yieldingright turn (Addressedin Chapter 18)
Yieldin U-tum y
Si nalized Control-Based AdiJstments
saturation flow rate: traffie pressure Y Y Y y
Saturationflow rate: turn radius Y Y Y y y y
Saturationflow rate: U-turn y
Saturationflow rate: DDI y
Yieldingleft turn y
Downstreamlink queue: internal Y Y y
Demat'ldstarvation: internal Y Y y
Downstreamlink queue: external Y Y y
Demat'ldstarvatiOll:external Y Y y
Additionalall-redjlost time Y Y Y y
S iall desi ned offset s y D
Other Adjustments
Weavefmergeadjustments Y
Notes; Pardo = partial dover1eaf, SPUI = single-point urban interchange, DO! = diverging diamond interchange,
DLT = displaced ~ tum, RCUT '" restr~ crossJng U-tum, and Mur = median U-tumo
y = adjustment generally applicabie and O = applicability depends en site configuration.

SPATIAL ANO TEMPORAL LIMITS

Distributed intersections are doscly spaced, interdependent intersections
best analyzed as a single unit. \'\Thenthe spatiallimits of the analysis are defined,
a cordon linio'is drawn around the area to be studied. For example, at the RCUT
intersection shown in Exhibit 23~5,the U-tum intersections essentially serve as
spatial analysis boundaries along the signalized arterial. However, if the cordon
!ine were to indude nearby intersections beyond thosc U-lum locations, the
Chapler 23 procedures might not be able to analyzc the full arca, and altemative
too[s might be needed.
Despite having multiple junctions, distributed intersections act as single
nodes with only three or four legs, origins, and destinations. The origin (01' 02J
0l< 04) and destination (DI' 02J O", 04) points shown in Exhibit 23-5 are generally
applicable to most intersections and interchanges described in this chapter. Thc
O-D methodologies outlined in Parts B and C are based on this concepto

Oiapter 23/RampTerminalsand A1ternativeIntersectiOlls Coneepts

Version 6.0 Page23-11
Highway Capacity Manual: A Guide for Mu/timodol Mobility Ana/ysis

Exhibit 23-5 0, 0,
Example of Spatial Umits fer
an RaJT Intersection

0,
- -- - •••••••
o.
0,

"- .•.•... - -- - -- ./
0•

With regard to the ternporallirnits of the analysis, thc HCM typically

recommends rnodeling the peak 1S-mio period. Multiple-period aoalyses are
commonly used to study oversaturated conditions and to study residual queues
persisting from one time period to the next. However, the Chapter 23 procedures
do not address queue spillback or queue spillover. Thus, alternative tools might
be required when significant oversaturation exists.

LOS FRAMEWORK
In developing an LOS framework for distributed intersections, consideration
of existing frameworks for similar facilities is informative. For isolated signalized
intersections (Chapter 19), average control delay per vehide is an intuitive
measure for LOS determination. For urhan street segments (Chapter 18), the
average difference between free-flow and actual speed is a fundamental quality-
Chapter 23 requires an LOS of-service indicator. Chapter 23 requires an LOS framework capable of capturing
framework CiJPdble of
CiJpturing spedfic signalized
specific signalized and arterial operations in a way that facilitates unhiased
and arterial operations in a comparisons among types of distributed intersections.
way that facHitiltes unbiased
compariscns among types ot Control delay would not be suitable as the sole measure for determining LOS
distributer:f intersections.
(as in Chapter 19), since it would not account for the diverted-path delay present
at sorne facilities. Travel speed would not be suitable as the service measure (as
in Chapter 18), because it does not describe the efficiency of sequential major-
and minor-street movements. Instead, the distributed intersections are aH
responsible for a certain amount of experienced travel time. More spedfically, each
0-0 path can experience (a) control de/ay at signalized or unsignalized locations
and (b) extra distance trave1 time. Sorne O-O paths may have multiple instances of
one or more of these elements. These elements can be used together to determine
the experienced travel time, and from this the performance measures of Chapter 23
can be derived. Equation 23-1 can be used to compute experienced travel time
(ETIl

Coo<epts O1apter 23/Ramp Terminals and A1temative Intersections

Page 23-12 Version 6.0
 Highway Capacity Manual: A Guide for Multimodal Mability Analysis

ETT= ¿d,+ ¿EDTT Equation 23-1

where d¡ is the control delay at each junction ; encountered on the path through
the fadlity and EDIT is the extra distance travel time.
Exhibit 23-6 illustrates the concept of providing unbiased comparisons
among distributed intersection configurations by using an RCUT intersection
example. The dashed Une denotes the path of a typicallcft turner arriving from
the minor street and entering the major street. Surnmarizing control delays in
accordance with Chapter 19 at aH three interscctions (i.e., westmost, middle,
eastmost) would not capture divcrtcd-path travel times between Points 2 and 6.
Furthermore, average travel speeds (in accordancc with Chapter 18) in the east-
west arterial directions would not consider control delays at Points 2 and 4. An
unbiased comparison betwecn configurations would require consideration of
experienced travel times hetween aHO-D points encircling the system, with the
system and 0.0 points spatially defined as in Exhibit 23-5.

Exhlbit 23-6
Example of Experienced
Travel Time at an ROJT
Intersection

'1
('0---- 0
I
1
10
Signalized Interchanges
The LOS designation is bascd on the opcrational performance of o-o
demands (shown in Exhibit 23-7) through thc interchange. The LOS for each O-O
is based on the average experienced travel time EIT of that demand as it travels
through the interchange. For example, foe the diamond interchange shown in
Exhibit 23-7, EIT for O-O movement O~.Dlis equal to the suro of westbound
through control delay at Point 1 (dv.'BTI)' control delay at Point 3 (dWBU)' and extra
distance travel time that lies roughly hetween Points 2 and 3 (EDITn). Thus, thc
EIT could be expressed as folIows:
ETT41 = d W871 + d W813 + EDTT23 Equation 23-2

Chapter 23/Ramp Terminals alld A1temative Intersections Concepts

Vef5,ion 6.0 Page 23-13
Highway Capadty Manual: A Guide for Mulfimodal Mobilify Analysis

Exhibit 23.7 0,
lIlustration of the LOS
Concept at a Diamond I
Interchange

v_,
:.----- "'---- 0.

0,
,
I ~ 2j 11 0.

\
\
\
-lI)"
0.
/
0.

To compute the exact EDTI foc this O-D movement, free-flow travel time
beyond Point 2 would be compared with the (hypothetical) free-flow travel time
that would occur if it were possible to tum left immediately on reaching the
southbound freeway. This EDTT would be similar to, but not necessarily
identical to, the free-flow travel time between Points 2 and 3.
Exhibit 23-8 illustrates the EDTT calculations given by Equation 23.3 through
Equation 23-10 for various O-D movements. EDTT subtracts hypothetical-path
free-flow travel times, which would occur under 90-degree (Le., right angle)
tums, from actual.path free.f1ow tcavel times. TItis calculation produces positive
EDTTs ioc allleft tums and negative EDITs ioc aH right tums.

Exhibit 23-8 o, o,
IIIustration of the EDTT O O
Concept at a Diamond
Interchange

D, • o,
o, o,

4) o
D, o.

"'Page
",ce""23-14 Olapter 23/Ramp Tenninals and A1temative Intersections
Vemm6.0
 Highway Capacity Manual: A Guide tor Multimodal Mobilify Analysis

Equation 23-3

Equation 23-4

Equation 23-5
EDTTZ4 = TTlOc - TTlc

Equation 23-6

Equation 23-7

Equation 23-8
EDTT31 = TT MK - TT MCK

Equation 23-9

EDTT42 = TTNJ - TTNBJ Equation 23-10

where EDIT13 is the extra distance travel time bctween origín 1 and destination
3, ITLPB is the travei time along path L-P-B in Exhibit 23-8, ITLB is the travel time
along (hypothetical) path L-B, other numbercd subscripts indicate other origin-
destination pairs, and other lettercd subscripts indicate other paths through the
interchange.
Exhibit 23-9 illustrates the concept of EDIT calculation at a Pardo A-2Q
interchange. Vehides traveling from arigin 04 to destination DI' instead of being
able to turn directl)' left at Point B, experience a full1,200 ft of out-of-direction
travel beginning at Point B along the arterial and ending at Point B along the
freewa)'. In contrast, vehides traveling from origin 03 to destinaban DI onl)'
experience 750 + 200 - 375 '" 575 ft of aut-of-direction travel (i.c., along-the-Ioop-
ramp distance, plus end-of-loop-to-overpass distance, minus intersection-to-
overpass distance).

Exhibít 23-9
IIIustration of the EOTT
04 to ~ -+ 1,200 ft of extra travel Concept at a Pardo A-2Q
ll.Jto ~ -+ 575 ft of extra travel 1nterchange

D, t C-B length = 200 ft 0,

0, + 0.
+- -+
A-B length = 375 ft
Inner loop length = 800 ft
loop length fmm A = 750 ft
Distance from Point B (on arterial)
to Point B (on freeway) = 1,200 ft

Q1apter 23/Ramp Terminals and A!temative Intersections Concepts

Version 6.0 Page 23-15
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

The first column of Exhibit 23-10 summarizes the delay-based LOS criteria
for each 0-0 within signalized interchanges. The second and third columns of
Exhibit 23-10 show that LOS F is defined to occur when either the volume-to-
capacity (vlc) ratio or the average queue-to-storage ratio (RQ) for any of the lane
groups that contain this O-D exceeds 1. Storage is defined as the distance
available for queued vehides on a particular movement, and it is measured on a
per lane basis. For example, if the lcft-turning lane group shown in Exhibit 23-7
has vlc > 1, the LOS for the entire O-D movement 04-DI will be F.1f a particular
lane group has vlc > 1, a1l O-Ds that travel through this lane group will operate in
LOS F, regardless of their delay. Similarly, if the average per lane queue in a
particular lane group exceeds its available storage, all O-Os traveling through
this lane group will operate at LOS F, regardless of their delay.
The values presented in Exhibit 23-10 reflect a control delay component
greater by a factor of 1.5 than those for signalized intersections. This reflects the
need for O-O movements to travel through multiple intersections.

Exhibit23-10
Cond'tion
lOS Critena for Each 0-0
v/cS land RqS 1 v/c> 1 Rq> 1
Within Signalized
ETT 5 veh for Eve une Grou for An une Grou for An une Grou
Interchanges
:S15 A F F
>15-30 B F F
>3G-SS e F F
>S5-8S o F F
>85-120 E F F
>120 F F F

As an illustration, consider the DDI shown in Exhibit 23-11. The ETI for O-O
movement 0eD3 is equal to the sum of northbound left-tum control delay at
Point 2 (dNBU)' westbound through control delay at Point 3 (dWBT:l)' and extra
distance travel time between Points 1 and 2 (EDTT1v. Thus, the ETI for a
northbound ¡eft-tom movcment (originating at the northbound freeway off-
ramp) could be expressed as follows:
Equation 23-11 ETT13 = dNBL2 + dWBT3 + EDTT12
Exhibit23-11 01 D,
lIIustration of the ETT
Concept at a DO!

D, D.
0, ) ~' D,

D, D,

Concepts Chapter 23/Ramp Termirlals and Altemative Intersections

Page 23-16 V~k1n6.0
 Highway Capacity Manual: A Guide tor Multimodal Mobility Analysis

Alternative Intersections
As with signalized interchanges, the LOS for alternative intersections is
based on the operational performance of O-O demands through the intersections
(previously shown in Exhibit 23-6). LOS for each O-O movement is again based
on the average El7 for that demand as it travels through thc intersections. In
displaced left-tum cases where the extra distance travel times are negligible, the
EIT is equivalent to the sum of control delays, as shown in Equation 23-12. For
example, for the OLT intersection shown in Exhibit 23-12, the EIT for O-O
movement 04-D¡ is egual to the westbound left-turn average control delay at
Point 1, plus the southbound through average control delay at Point 2, applied to
the flow rate traveling from 04 to D¡. Movement O~-D¡can also be described as
the wcstbound left-turn movement for the DLT interscction as a whole. Thus,
ETT for the movement from origin 04 to destination D¡ is as follows:

0, 0, Exhibit 23-12
IIIustration of the ETT
Concept at a Displaced Left-
Turn intersection

0, ~--------","-"¡':::j:=:t:.=:..:=-=:..:=-==-= 0.
0l "- D.

1 DI 01
Note: SOutt1bound tlHvugh óelay at Point 2 could be zero jf deslgned for a protected-only pitase at Pojnt 1.

£TT41 = dWBL1 + dSBT2 Equation 23-12

Exhibit 23-13 summarizes the LOS criteria for each O-O movcment within
altemative intersections. The values presented in Exhibit 23-13 reflect control
delay thresholds identical to those for conventional signalized intersections and
33% lower than those for interchanges (Le., in Exhibit 23-10).

Condjtion Exhibit 23-13

LOS Criteria for Each 0-0
v/e:S 1andRQS 1 v/e> 1 RQ> 1
s Within A1ternative
ETT veh for Eve lane Grou for An LaneGrou for An lane Grou
lntersections
:S10 A F F
>10-20 B F F
>20-3S e F F
>3S-SS o F F
>55-ll<l E F F
>80 F F F

Olapter 23/Ramp Terminals and A1ternative intersections Coocep"

Version 6.0 Page 23-17
 Highway Capacity Manual: A Guide for MultimOdal MObility Analysis

Interchanges with Roundabouts

Similar to signalized interchanges, the LOS designation for interchanges with
roundabouts is based on the operational performance of 0.0 demands through
the interchange. The LOS for each O-O movement is based on the total average
control delay experienced by that demand as it travels through the interchange.
Exhibit 23-14 summarizes the LOS criteria for each 0-0 of an interchange with
one or two roundabouts. The values presented in Exhibit 23-14 are greater than
those for non-interchange roundabouts to reflect the fact that sorne of the O-O
movements rnight travel through two roundabouts, while others might travel
through only one. The values are also generally lower than the respective values
for signalized interchanges, since drivers would Iikely expect lower delays at
roundabouts.

Exhibit 23-14
LOS Criteria for Each 0-0 of Condition
an Interchange with v/eS 1 andRqS 1 v/e> 1 RQ> 1
Roundabouts ETT s veh for Al! A roaches for An A roach for An A roach
m A F F
>15-25 B F F
>25-35 e F F
>35-50 D F F
>50-75 E F F
>75 F F F

Other Interchange Types

Interchange types and control not explidtly ¡nelude<! in this chapter (e.g.,
two-way STOP-controlIed diamond interchanges) do not have LOS criteria
defined on an O-D basis.ln the absence of such LOS criteria, analyses of these
interchange types and comparisons with other interchange types can be made by
using control deIay for each O-O rnovement, along with other applicable
performance mcasures. These performance measures can be deterrnined with
procedures in this and other HCM chapters, altemative tools, or both,
aggregated as appropriate into O-O performance measures by using the
techniques in this chapter.

Concepts Chapter 23/Ramp Terminals and A1temative Intersections

Page 23-18 Version 6.0
 Highway Capacity Manual: A Guide for Mu/tjmodal Mobility Analysis

Part B: Interchange Ramp Terminal Evaluation

1. INTRODUCTION

OVERVIEW
Interchange ramp terminals are critical components of the highway network.
They provide the connection between various highway facilities (freeway-
arterial, arterial-arterial, ete.), and thus their efficient operation is essential.
lnterchanges must be designed to work in harmony wilh the freeway, the ramps,
and the arterials. In addition, they need to provide adequate capacity to avoid
affecting the connecting facilities.
This section presents the methodology for the analysis of interchanges
involving freeways and surface streets (Le., service interchanges). It was
developed primarily on the basis of research conducted through the National
Cooperative Highway Research Program (1-3), Texas Department of
Transportation research (4), and FHWA research on DDIs (5).

PART ORGANlZATION
This part of Chapter 23 presents methodologies for the evaluation of
interchanges, induding diamond interchanges, DDIs, SPUIs, interchanges with
roundabouts, and pardo interchanges. Section 2 presents additional concepts
specific to interchanges not covercd in Part A. Section 3 presents the core
methodology foe evaluating the operational performance of interchanges for
diamond interchanges, DDIs, SPUIs, and pardos. Section 4 provides extensions
to the methodology, induding evaluation of interchanges with roundabouts and
other unsignalized intersections, and a discussion of pedestrian and bicyde
analysis at interchanges. Section 5 provides appIications of the methodology,
induding example results, a discussion of types of analysis, and considerations
for the use of alternative tools at interchanges.

O1apter 23/Ramp Terminals and Altemative Intersections Introduction

Version 6.0 Page 23-19
 Highway Capacity Manual: A Guide for Multimodal Mobi/ity Analysis

2. CONCEPTS

TYPES OF INTERCHANGES

A number of types of interchanges are recognized in the literature. A PoliCy

011 Geometric Desigll o/ Highways and Streets (9) provides extensive information on
interchange designs and their characteristics. Part A discussed intersections and
interchanges in broad terms; this section more specifically ilIustrates and
discusses the interehange designs considered in this chapter: diamond
interehanges, DDIs, SPUIs, pardos, and interchanges with roundabouts.

Diamond Interchanges
Most forms of diamond interehanges result in two or more dosely spaced
surface intersections, as illustrated in Exhibit 23-15. On a diamond interchange,
only one conneetion is made foc eaeh freeway entry and exit, with one
conneetion per quadrant. Left. and right-turning movements are used for entry
to or exit from the two directions of the surfaee facility. When demands are low
(generally in rural areas), the junetion of diamond interchange ramps with the
surface facility is typieally controlled by STOPor YIELDsigns.lf traffie demands
are suffidentiy high, signalization becomes necessary.
The interchange methodoIogy There are many variations of the diamond interchange. The typical diamond
ÍS only applicable when both
ramp terminal 5 are signalized configuration has three subcategories defined by the spadng of the intersections
or both are roundabouts. formed by the ramp-street connections. Conventional diamond interchanges
provide a separation of 800 ft or more between the two intersections.
Compressed diamond interchanges have intersections spaced between 400 and
SOO ft, and tight urban diamond interchanges feature spadng of less than 400 ft.
Secause of right-of-way constraints, compressed and tight diamonds are more
likely to be used in urban areas, while conventional diamond interchanges are
more Iikely to be used in rural or suburban settings.
Split diamond interchanges have freeway entry and exit ramps separated at
the strect level, creating four intersections. Diamond configurations also can be
combined with continuous one-way frontage roads. The frontage roads become
one-way arterials, and turning movements at the intersections created by the
diamond interchange become even more complex. Separated U-tum roadways
may be added and U-turos removed from the signal scheme, if there is a signa!.
A partial diamond interchange has fewer than four ramps, and not aH freeway-
street or street-freeway movements are served. A three-Ievel diamond
interchange features two divided levels, so that ramps are necessary on both
facilities to allow continuous through movemenls.
AH these forms oi diamond interchanges are depicted in Exhibit 23-15. The
methodology in this chapter is applicable to aU diamond interchange forms
except the split diamond and the three-Ievel diamond. The methodology
addresses interchanges where both terminals are signalized or both terminals are
roundabouts.

Concepts Chapter 23/Ramp Terminals and Altemative Intersections

Page 23-20 Version6.0
 Highway Capacity Manual: A Guide for Muttimodal Mobility Anolysjs

Exhibit 23-15
Types of Diamond
D Interdlanges

í 1\

Con'lentional, D > 800 ft

Compressed, 0= 400-800 ft Diamond with continuous
Tight Urban, 0< 400 ft frontage roads

/ '\
.' . .. .. . '

'. ...
'.

/
"-

Diamond with continuous frontage Split diamond interchange

roads and U-turn lanes (crossroads, one- or two-way)

u •
Partial (half) diamond
interchange
T ...
",

Three.le'lel diamond
interchange
..' .
..

Note: - _. - •• Possib4e altematiYe conf\guratlon cA signa' bypasses operating as unsignali.red rnovements;

ttIese are rnovements th3t are not using the ramp tenninals.

Chapter 23/Ramp Terminals and A1ternative Interseetions Concepts

Version 6.0 Page 23-21
 Highway Capocity Manual: A GU;de for Mulfimodal Mobility Analys;s

Diverging Diamond Interchange

The DOI, sometimes also referred to as a double crossover diamond
interchange, is a variation of the traditional diamond interchange at which the
cross-street movements cross directions twice, once at each ramp terminal.
Except for poten tia! pedestrian conflicts, the DDI aIlows leH tums onto the
freeway to be a free-flowing movement at the internal crossover, which results in
large efficiency gains for interchanges with heavy lcft-turn demando Left turns
from the freeway are similarly able to merge onto the arterial street in their desired
direction of travel without crossing over the opposing through movement. A
schcmatic of a DDI in east-west orientation is shown in Exhibit 23-16.

Exhibit 23-16
IIIustration of a DDI

Research on the operational performance of the DDI (6--8) showed that

despite its potential for enhancing the operational efficieney of a standard
diamond interchange (especially for locations with heavy left-tuming
movements), several unique considerations apply in an operationa! evaluation of
the DOI. Among them are lane utilization at the external crossover, saturation
fIow rate adjustments for the crossover movement, lost-time considerations for
through and turning movements, and capacity of YIELD-controlled tums. The
methodology described in this chapter accounts for these operational
characteristics of DDIs.

Parclo Interchanges
Parclo interchanges are shown in Exhibit 23-17. A variety of parclo
interchanges can be created with ane oc two loop ramps. In such cases, one oc
two of the outer ramps intersect the crossroad in a manner similar to a diamond
ramp, allowing a movernent to take place by means of a right turn. In sorne
parclo configurations, left tums also may be made onto or off of a loop rampo The
methodology in this chapter is applicable to parclo interchangcs where both
terminals are signalized oc both terminals are roundabouts.

Con<epts Chapter 23/Ramp Terminals and AJtemative Intersectioos

Page 23-22 Version 6.0
 Hjghway Capadfy Manual: A Guide for Mulfimodal Mobmty Analysis

Exhibit 23-17
Types of Parclo Interchanges

Parclo A, 2 Quadrants Parclo A, 4 Quadrants

In Pardo A !orms, loop ramps
on the mainfine occur in
advance of the crossroad. In
Pardo B forms, loop ramps on
the mainllne 0CClJf beyond the
aossroad. In Pardo AB forms,
loop ramps on the mainline are
Iocated en the same side of
the crossroad, ene in advance
of the crossroad for its
dlfediOn of travef and the
ParclO B, 2 Quadrants Parclo B, 4 Quadrants other beyond.

Parclo AB, 2 Quadrants ParclO AB, 4 Quadrants

Note: o fl\:Js5iblealtematlve configUfiltion r:I sigroal bypasses; operatlog as unsignallrelll'(M!ments;

these iII1!moYeITlefItStnat are not I,l5ing lhe ramp terminals.
.~_'~ Possible altemattve ooofiguration r:I Oght-wm-only ramp passing lhrough lhe ramp terminal.

Single-Point Urban lnterchanges

A SPUI combines aH the ramp movements into a single signalized
intersection and has the advantage of operating as such. The design eliminates
the critical issue of coordinating the operation of two dosely spaced
intersections. The SPUI is depided in Exhibit 23-18.

Exhibit 23-18
Single-Point Urnan
Interchange

Chapter 23/Ramp Terminals and A1temative Intersections Cooéepl>

Ve~1Qn 6.0 Page 23-23
 Highway Capacity Manual: A Guide for Mulfjmodal Mobility Analysis

Interchanges with Roundabouts

Roundabouts can replace signalized or STOP-controlled interscctions as
interchange ramp terminals. Three types of roundabout ramp terminal designs
are typicalIy used in the United States and are iIIustrated in Exhibit 23-19. The
first dcsign consists of two traditional roundabouts at the two nodes of the
interchange. The second design is caBed the raindrop roundabout interchange,
and it restricts certain movements within cach roundabout by creating raindrop-
shaped central islands. These two designs are essentially the same, except that
the former should be provided when U-tums are allowed or when there is an
additional approach to the roundabout. The third design consists oi a single
roundabout spanning both sides oi the freeway via over- or underpasses.
These three designs are applicable to both diamond and parclo interchanges.
Their major advantage is that they can reduce the number of lanes needed
between terminals, which significantly reduces structure-related costs. They also
eliminate the need for coordinating signal operations at the two c!osely spaced
intersections. A potential disadvantage oi using roundabouts is that spillback
from a downstream fadlity into the roundabout may result in gridlock foc aH
movements at the roundabout, since all movcments must use the circulating
roadway.

Exhibit 23-19
DialllOllCl Interchanges wittl
Orcular Ramp Terminals

Freeway

A
•
(a) Roundabout Ramp Terminals
•
(b) Ralndrop-Shaped Ramp Terminals

VA
11
(e) Single-Point Roundabout Interchange

Con<epts Olapter 23/Ramp Terminals aOOAltemative Intersections

Page 23-24 Vef3ioo 6.0
 Highway Capacity Manual: A Guide for Mu/timodal Mobility Analysis

O-D AND TURNING MOVEMENTS FOR CONVENTlONAL INTERCHANGES

ExhibH 23.20 illustrates the 0.0 movement letters for different types of
interchanges considered in this methodology_ In Chapter 34, Interchangc Ramp
Terminals: Supplemental, Exhibit 34.163 through Exhibit 34.178 provide the
corresponding calculations to obtain tuming movements from 0-0 movements
and vice versa.

Exhibit 23-20
o-D F10wsfor Each

e
I Interchange Configuratioo

Q) @
, ,~ 1 , • 1
8 ""HIM

(a) ~mond Interehal\ge (b) hrdo A-2Q

I, I
~
, G
G

Q) @
"" ••••• 1

(d) Pardo 8-4Q

I
e

Q) Q) @
1
, lJ "" •••••

(e) Pardo AB-4Q (t) PIIrdo A-4Q

(11) Pardo A8-2Q

Olapter 23/Ramp Terminals and Altemative Intersections ro""",,,

Version 6.0 Page 23-25
 H;ghway Capacity Manual: A Gu;de for Multimodal Mobiljty Analys;s

3. CORE METHODOLOGY

SCOPE OF THE METHODOLOGY

Spatial and Temporal Limits
The methodology addresses interchanges with signalized intersections,
interchanges with roundabouts, and the impact and operations oí adjacent
elosely spaced intersections. The methodology also addresses DDIs with both
signalized and unsignalized tuming movements. Interchanges with two-way
STOP-controlled intersections oc interchanges consisting of a signalized
intersection and a roundabout cannot be evaluated with the procedures of this
chapter. Traffic circles (e.g., intersections with a circular island in the middle and
signals at the approaches) are not considered in this chapter.
The scope of this chapter ineludes the operational analysis for a full range of
service interchange types, including diamonds, DDIs, SPUIs, and parclos. The
methodology addresses at-grade intersections, not including the freeway proper,
and focuses on surface streets; it does not analyze freeway-to-freeway
interchanges.
Chapter 34, Interchange Ramp Terminals: Supplemental, ineludes a
methodology for assessing the operational performance of various types of
interchanges for purposes of interchange type selection. The method can be used
to obtain guidance for assessing various interchange types with respect to their
operational performance; it does not provide guidance for selecting an
appropriate interchange type with respect to economic, environmental, land use,
and other non-operational conceros.

Performance Measures
The operational analysis methodology for interchanges provides the
performance measure experienced travel time (ETT). For each movement, the ETI
ineludes the control deJay experienced at each junction encountered, plus the
time experienced in traveling any extra distances required by the designo lt may
be expressed as follows:

Equation 23-13 ETT = L +L di EDTT

where d¡ is the control delay at junction i encountered on the path through the
interchange (seconds) and £DIT is the extra distance travel time (seconds).
The methodology computes control delays at each individual junction
making up the interchange, so useful related measures such as capacity and vlc
ratio are available for each of those junctions. Use of the ETI performance
measure allows comparison of interchanges of different forms on the same basis.
Intersections (conventional and altemative) may be compared with interchanges
having multiple junctions or with rerouted movcments driving longer distances.
Standard diamond and pardo interchanges have non-zero EDTT values
because their travel paths deviate from the freeway centerline. In these cases,
EDTI is calculated at the ramp design speed.

Core Methodology Chapter 23/Ramp Terminals and Altemative Intersections

Page 23-26 Version 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Strengths of the HCM Procedure

This chapter offers a eomprehensive proeedure for analyzing the
performance of several types oí interchanges. Simulation-based tools offer a
more detailed treatment of the arrival and departure of individual vehicles and
of features of the signal control system, but for most purposes, the HCM
proeedure produces an acceptable approximation. The HCM procedure offers
some advantages over the simulation approach:
• The HCM provides saturation £low rate adjustment faetors based on
extensive field studies.
• The HCM produces direct estima tes of capaaty and vJc ratio. These
measures are mueh more elusive in simulation.
• The HCM provides LOS by O-D, whieh facilitates the comparison of
operational performance for different interchange configurations.
• The HCM provides deterministic estima tes of the measures of
effectiveness, which is important for sorne purposes sueh as development
impact review.

limitations of the Methodology

The identified Iimitations of the HCM procedure for this chapter cover a
number oí conditions that are not evaluated explicitly, including the following:
• Oversaturated conditions, particularly when the downstream queue spills
back into the upstream intersection for long periods of time;
• The impact of spillover into adjacent travellanes due to inadequate tum-
pocket length;
• The impaet of spillback on freeway operations (however, the method does
estimate the expected queue storage ratio for the ramp approaehes);
• Ramp metering and its resulting spillbaek of vehicles into the interchange;
• Impaets of the interchange operations on arterial operations and the
extended surfaee street network;
• Interchanges with two-way STOP-controlledintersections or interehanges
consisting of a signalized interseetion and a roundabout;
• Lane utilizations for interchanges with additional approaches that are not
part of the prescribed interehange configuration (however, guidance is
provided for addressing those cases);
• Laek of provision of link travel times and speeds (the methodology does
provide delay estimates); and
• Full cloverleaf interchanges (freeway.to-freeway or system interchanges),
sinee the scope of the ehapter is limited to serviee interchanges (e.g.,
freeway-to-arterial interchanges).
If the user is interested in the analysis of conditions that fall within the aboye
methodologicallimitations or in the investigation of dynamic traffic operations
(Le., those that evolve in time and space), the use of another analysis too!, such as

Chapter 23jRamp Terminals and Altemative Intersections Core Methodology

VetSlOn 6.0 Page 23-27
Hjghway Capacity Manual: A Guide far Multimadal Mability Analysis

simulation modeling. is advised. Section 5 ineludes information on the use of

altemative tools for the analysis of interchange ramp terminals.
~e operat~onal analysis is only one factor to be considered in the design or
redeslgn of.an ~~terchange ramp terminal. Other important factors indude right-
of-way a.vaJlablhty and economic and environmental constraints. The scope of
this chapter does not inelude such considerations; the chapter focuses only on the
traffic operational performance of interchanges.

Alternative Tool Considerations

General guidance on the use of alternative traffic analysis tools for capacity
and LOS analysis is provided in Chapter 6, HCM and Alternative Analysis Tools,
and Chapter 7, Interpreting HCM and AItemative Tool ResuIts. Section 5 of this
part contains more specific guidance for the application of altemative tools to the
analysis of interchange ramp terminals. Chapter 34, rnterchange Ramp
Terminals: Supplemental, contains supplemental examples illustrating the use of
altemative tools for interchange analysis. Additional information on this topic
may be found in the Technical Reference Library in Volume 4.
As indicated in Chapter 6, traffic models may be elassified in several ways
(e.g., deterministic versus stochastic, macroscopic versus microscopic). The
altemative tools used for interchange analysis are generaIly based on models that
are microscopic and stochastic in nature. Therefare, the discussion in Section 5
will be limited to microsimulation tooIs.
This chapter provides a methodology far estimating the capacity, control
delay, queue storage, and LOS for a given set of traffic, control. and design
conditions at an interchange. As with most procedural chapters in this manuaL
simulation outputs, espedally graphics-based presentations, can provide details
on problerns at specific elements of the interchange that might otherwise go
unnoticed with a macroscopic analysis. For exarnple, problems associated with
tom bay overflow or blockage of access to tum bays can be beUer observed by
using microscopic simulation tools. Aitemative tools offer performance measures
such as number of stops, fuel consumption, and pollution. They are also useful
for the evaluation of other modes, incIuding pedestrians, cyelists, and transit,
and their interaction with vehicles at interchanges. The animated graphics
offered by many simulation tools are espedally useful for observing network
operations and identifying problems at spedfic elements.
Simulation tools use definitions of delay (and therefore LOS) different frorn
those of the HCM, especially for movements that are oversaturated at sorne point
during the analysis. Creat care must therefore be taken in producing LOS
estima tes directly from simulation. Chapter 7, Interpreting HCM and Altemative
Tool Results, discusscs simulation-based performance measures in more detail.

REQUIRED DATA ANO SOURCES

The analysis begins with the assembly of aH pertinent input data, such as
geometric eharacteristics, traffie demands, and signalization information. Exhibit
23-21 provides a summary of aH input data required in eonducting an
operational analysis for interchange ramp terminals.

Core MethodoIogy O1apter 23/Ramp Terminals and A1ternative Intersections

Page 23-28 Vef5ion 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Suggested Default Exhibit 23-21

Potential Data
Parameter 1 Source I Value Summary of Required Input
Data fer Final Design and
Geometric Data
Operational Analysis of
Field data aeria\ hoto Must be rovided
A"" Field data aerial hoto Must be revided
Signalized lnterchanges
Number of lanes
Avera e lane width W. ft Field data aerial hoto 12ft
Grade (G, OJo) Field data Flat approach: 0%
Moderate grade: 3%
Stee rade: 6%
Existence of exclusive left- or right-tum lanes Field data aerial hoto Must be rovided
Len th of stora e for each lane rou ft Fietd data aerial hoto Must be rovided
Distance between the two intersections in the Field data, aerial photo Must be provided
interchan e O ft
Distances corresponding to the ¡ntema! Field data, aerial photo Must be provided
storage between interchange intersections
and nearb ad"acent intersections ft
Turnin radii for all turnin movements tt Fierddata aerial hoto Must be rovided
Extra travel distance relative te centerline (ft) Field data, aerial photo Must be provided
Demand and Traffie Data
Demand volume by o-o or turning movement Field data, past counts Must be provided
_. (V, _v~l~)
Right-tum~on-red f10w rates Fi~d _d-ªta,~st~()u!lts 0!l.'!'~/h ~
i.eft-turñ.Oñ~rEidf10wrates Field data, past.counts 0.0 v~h/h____ __
Base sahiration flow rate (s'" pcjhg/ln) Field data, judgment Metro popo~ 250,000:
1,900 pc/h/ln
Oth_erwi.~:1.?50 pc'-~/ln_
Peak hour factor (PHF) Field data, judgment Total entering vol.
O!: 1,000 veh/h: 0.92
Total entering vol.
~1,OOOveh/~0.90
r~rcent heavyy-ehj9es (~\':. %L _ Field_da~, ~~t _,,º-uQts J%--- ----
.APp~d!pe<lestria'"!.Jlow!ates_ (vpoO, p/h) Field data, past_counts Must be provided _
.AP~~c!! bi_<:yclE01~.!ates(VDt bicycles,lh) Field data, past counts t1ust.Pe provided_~~_
Local bus stopping rate (Ns, buses/h) Field data, judgment Central businessdistrict
(CBO) bus stop: 12
buses/h
N9_,"!j;:BD.:J_b.\l.se@
__
Pa~ng _activity(N ••._!J1<Ineuvers/~) _Field data, judgment_ _Musl:_
be provided~ _
A~rival_type(AD Field data, judgll)ent J _
UPS_~Ill~I!..e.!ing adjusb!le!'lt faC!9~ Fietd data, judgl!lent 1.0 ._._
Approach speed (S.••mi/h) Field data, judgment Speed IilTlit.:t..S mifh .__
Free-f!ow speed (S" mi(h) ------- Fierddata, judgment Speed limit + 5 mifh
Signal Data
T of si nal control Field data Must be rovided
Phase uence Field data Must be rovided
Cycle length (if ap ro riate s Field data Must be rovided
Green times {if ap ro riate G s Field data Must be rovided
Yellow-plus-al1-redchange-and-clearance Field data Must be provided
interval inter reen Y. s
Offset if a ro riate Field data Must be rovided
Maximum, minimum green, passagetimes, Field data Must be provided
hase recall for actuated control
Presenceof estrian ush button Field data Must be rovided
Minimum estrian reen G s Field data Must be rovided
Phase plan Field data Must be provided

Agencies that use the methodologics of this chapter are encouragcd to dcvclop
a set oi local default values (where deiaults are applicable) bascd on measuremcnts
at interchanges in their jurisdiction. Local default values providc the bC'st means
oi ensuring reasonable accuracy in the analysis results. In the abscnce of local
deiault values, the values provided in Exhibit 23-21 can be used if appropriate.

Chapter 23(Ramp Terminals and A1ternativeIntersections Core Methodology

Version 6.0 Page 23-29
 Highwoy Copocity Manual: A Guide tor Multimodal Mobility Analysis

OVERVIEW
ExhibH 23-22 summarizes the basic nine-step methodology for the design
and operational analysis of signalized interchange ramp terminals. The
methodology is similar to that of Chapter 19, Signalized Intersections, with
further consideration for imbalanced Iane utilizations, additionallost times due
to downstream queues, demand starvation, and additionallost times due to
interactions with dosely spaced intersections.

Exhibit 23.22
Interchange Ramp Terminals
""" Oiamond/Pardo/OOI
Methodology: Final Design 1 1
and Operational AnarysiS for
Interchanges I Step 1: Determine 0-0 Demllnds llnd Movement Demands I
1
I step 2: Determine Lllne Groups I

I -
Step 3; Determine AdJusted Scturmon F10wRates

j
Step 4; Determine Effective
I
y"",

Green AdJustment due to

Interchllnge Operlltions

No adjacent
¡,terSKtion

""","'
¡'~"xisl$
Step6;
Step 5; Determine Effective Determine
Green Adjustment due to Pelform1lnce of
Oosely Spaced Ad]1lCent Yla.o-Controlled
Intersections Tums

I Step 7: Determinev/c Ratio llnd Queue Storage RatiO I

I • 1
Step 8: Determine Control DelllY llnd ETT tor Ellch 0-0 I
1 1
I Step 9; Determine LOS I
Note: Step 6 also i!"dudes óeterminaborl of performance of free-flow tum movements.

The analysis of SPUIs is outlined on the Idt part of the flowchart. The
flowchart highlights only the components added to thc signalized intersection
methodology for analyzing SPUIs. The right part of the flowchart highlights the
components added to the signalized intersection mcthodology for analyzing
diamond, diverging diamond, and parclo interchanges. Each of the steps
outlined in Exhibit 23.22 is explained and discussed below.

Core Methodology Chapter 23/Ramp Terminals aOO A1temative Intersections

Page 23.30 Version 6.0
 Highway Capacity Manual: A Guide for Mulfimodal Mobility Analysis

COMPUTATIONAL STEPS
Step 1: Determine O~DDemands and Movement Demands
Either o-o demands or intersectíon turning movemcnts for the study
interchange may be available to the analyst. Since both are needed in the
analysis, the first step in the methodology consists of calculating cither the
turning movements by using the O-O demands or the o-o demands by using the
turning movements. If the interchange is a SPUI (Le., has only one intersection),
the O.D demands and the turning movement demands are the same, and the
analysis proceeds similarly to the methodology of Chapter 19, Signalizcd
Intersections, to estímate capacity, v/c, delay, and queue storage ratios. The
Applications section of Part B provides guidance on converting O-O movements
to turning movements and vice versa for each type of interchange configuration
addressed in this methodolog)'.

Step 2: Determine Lane Groups

As in the case of signalized interchanges, thc methodology for interchange
ramp terminals is disaggregate; that is, it is designed to consider individual
intersection approaches and individuallane groups within approaches. Thc
scgmentation of thc interchange into lane groups generall)' follows the same
guidelines that apply for the analysis of signalized intersections as described in
Chapter 19, Signalized Intersections.

Step 3: Determine Adjusted Saturation Flow Rates

The saturation flow rate for each tane group can be measured in the field or
estímated with the following equatíon:
s = soN fwfHVgfpfbbfafRTfLTfLpbfRPbfllfwfDD/ Equation 23-14

where
s = adjusted saturatíon flow rate (veh/h),
So base saturation flow rate per lane (from Exhibit 23-21),
N number of lanes in the lane group,
fu. adjustment factor for lane width (from Chapter 19),
fuvg adjustment factor for heav)' vehicles and grade (from Chapter 19),
Ir adjustment factor for existence of a parking lane and parking activity
adjacent to the lane group (from Chaptee 19),
l/Ib adjustment factor foe local bus blockage (from Chapter 19),
J. adjustment factor foe the area type (from Chapter 19),
/Rr adjustment foe right-turning vehicle peesence in the lane group (fram
Chapter 19, incorporatingf~ interchange saturation flov.•.adjustment
No. 4 fram Equation 23.19),
/Lr adjustment for leH-turning vehicle presence in the lane graup (fram
Chapter 19, incorporatingf~ interchange saturation flow adjustment
No. 4 from Equation 23-19),

Chapter 23/Ramp Terminals and A1ternative Intersections Core Methodology

VersiOn 6.0 Page 23-31
 Highway Capacity Manual: A Guide far Multimodal Mobility Analysis

fl.pb pedestrian adjustment factor for Jeft turos (from Chapter 19),
fRpio pedestrian-bieycle adjustment factor for right tums (from Chapter 19),
fv adjustment for traffic prcssure (interehange saturation flow rate
adjustment No. 1 from Eguation 23-15),
fw adjustment factor for lane utilization (interchange satura non flow rate
adjustment No. 2 from Eguation 23-16 and Equation 23-17), and
fDDJ e:o adjustment for DD! erossovcr [interehange saturation flow rate
adjustmcnt No. 3 [e:o 0.913 according to research (5)].
Most of the factors in Eguation 23-14 are obtained from Chapter 19,
Signalized Intersections. The last three, which are not obtained from Chaptcr 19,
are described in greater detail below. A fourth adjustment factor, fR' which
quantifies the effect of tuen radius on saturation flow rate of a lcft- or righHuen
movement, is used to modify protected tum movement adjustment faetors
(provided in Chapter 19) for interchanges. The term fR is not shown explieitly in
the cquation, sinee the radius adjustment modifies the existingfLTandfRT
adjustments already in the eguation. A different radius adjustment is nreded for
interehanges, since right- and lefHuen radii are often much larger than at a
standard intersection, resulting in higher tuming speeds and thus a lower impact
on saturation flow rates. These four adjustment factors are discussed below.

lnterchange Saturation F/ow Adjustment No. 1: Traffie Pressure, fv

The s.aturation ffowadjustment
factor for trafflC pressure is
Saturation flow rates have general1y been found to be higher during peak
unique to the interchange traffic demand periods than during off-peak periods (9). Traffic pressure refleets
methodology as found in
research but has not been
the display of aggressive driving bchavior for a large numbcr of drivers during
documented fO( a standard high-demand traffic conditions. Under sueh conditions, many drivers aecept
Signalized intersection ana/ysis
in Chapter 19.
shorter headways during queue discharge than they would under different
circumstanccs. The effect of traffic prcssurc varies by traffic movcment. Left-turn
movements tend to be more affectcd by traffic pressure than through or right
movements. To aeeount for this phenomenon, the saturation flow rates at
interchange approaches are adjusted by using the traffic pressure factor. This
factor is computed with the following eguatian:

(Ieft turn)
1.07 - 0.0067~ x min(v¡.30)
fquation 23-15 fv = 1
{
(through or right turn)
1.07 - 0.00486 x min(v¡,30)
where fo is the adjustment factor for traffie pressure and V; is the demand flow
rate per cyclc per lane (veh/cyele/ln).
For values oE v; higher than 30 veh/cycle/ln, 30 veh/cycle/ln should be used,
since the cffects oE demands higher than that value are oot known. Exhibit 23-23
tabulatcs the results oE Equation 23-15 Eoc various demands and for each turning
movement type.

Core Metllodology Chapter 23/Ramp Terminals and A1temative Intersections

Page 23-32 VeISion 6.0
 Highway Capadty Manual: A Guide for Multimodal Mobility Analysis

Movement Type Exhibit 23-23

Demand Flow Rate vD
Throu h and Ri ht Tum Adjustment Factor for Traffie
veh de In LeftTum
Pressure (fv)
3.0 0.953 0.947
6.0 0.971 0.961
9.0 0.991 0.974
12.0 1.011 0.988
15.0 1.032 1.003
18.0 1.054 1.018
21.0 1.077 1.033
24.0 1.100 1.049
30.0 or reater 1.152 1.082

When the ¡ane group is shared by several movements, the adjustment f<lctor
for traffic pressurc is cstim<ltcd as the average (weighted on the basis of flows) of
the respective movements.

Interchange Saturation Flow Adjustment No. 2: Lane Uti!ization, fw

Vehides at interchanges do not distribute cvenly among lanes in a lane
group, and their lane seleetion is highly affected by their ultimate destination.
For example, for two-intersection interchanges, when therc is a high-volume leH
turn at the downstream intersection, traffie at the upstream interscetion will
gravitate toward the left lanes, while through and right-turning vehides will
tend toward the right. While sueh movements may occur at convcntional
intersections as wen, the short links and typically high volumes at interchanges
generally result in greater variation in tane distrihution. Consideration of ramp
configuration makes calculation of lane utilization for O-O pairs more accurate.
Segregation at the upstream interscction may üccur by driver seleetion or by
designated signing <lndpavement marking.
1'0 account for these phenomena, lane utilization rnodels have been
devclopcd specifically for the external through approaches (surface streets) of
two-interseetion diamond interchangcs, as well as OOls. The lane utilization
factors for .111 other interchange appco<lchcs(freeway ramps, internal appro<lchcs,
<lndSPUI appco<lches) <lreestimated by using the proccdures oí Chapter 19.
These laoe utilization factors are then used to adjust the saturation flow rates for
c<lchlane group.

Lane Utl1ization Adjustment for Oiamond Interchanges

The lane utilization factor accounts for the uncqu<ll distribution of traffie
among the lanes in a lane group with more than onc lane. The factor provides an
adjustmcnt to the base saturation flow r<lte.Ttis bascd on the flow in the tane
with the highest volume <lndis ealculated by Equation 23-16:
1
Equation 23-16
fW=OAVo Lmax
xN
where
fLu =o adjustment factor for lane utilization;
'YoVL.....,- =o percent of thc total approach £low in the laoe with thc highest volume,
expressed as a decimal; and
N number oí lanes in l<lnegroup.

Chapter 23/Ramp Terminals and A1ternative Intersections eore Methodology

Version 6.0 Page 23-33
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

A series of models have been developed to predict %V LmOl' for the externa]
arterial approaches of two.intersection interchanges as a function of the
downstream turning movements. The remaining approaches should use lane
utilization factors based either on field data or on values obtained from Exhibit
19-15. Equation 23.17 provides a model to estimate the percent of total volume
per lane. Exhibit 23-24 provides parameters for each type of interchange
configuration and for two-, three-, and faur-Iane arterials.
Equation 23-17
%Ve = -1 + a (----- VR ) + a ( ----- v, ) +a (D--- x V,)
I n 1 VL + UR + Vr 2 VL + VR + Vr 3 106
where
%VU percent of traffie present in lane L¡, with L1 representing the leftmost
lane, L2 representing the seeond lane from the left, and so forth;
n = number of lanes in the lane group;
Q¡ coeffident for j= 1 through ¡= 3 (see Exhibit 23-24);
D distanee (ft) between the two interseetions of the interchange
(Equation 23.17 is valjd for values of D below 800 ft);
Vil O-O demand flow rate traveling through the first intersection and
turning right at the seeond (Vil = O if there is an exclusive right-tum
lane on the external approach);
vL 0-0 demand flow rate traveling through the first interseetion and
turning left at the second; and
VT = O-O demand flow rate traveling through the first intersection and
through the second.
These models estimate the percent of traffic expected to use each through
lane as a function of O-O demands on the subject approach. They focus on the
external arterial approaches and predict the pcrcent of traffie expected to use a
particular lane as a function of downstream turning movements. The turning
movements are expressed in terms of their respective 0-0 flows. 0-0 Flows A
through N are shown in Chapter 34 for each configuration typc. When the
eastbound and westbound approaeh patterns are symmetrical, parameters far
the eastbound and the westbound directions are identical, and only the 0-0
flows differ. Interchange approaehes with identical tuming movement patterns
in the subject direction (eastbound or westbound) are grouped together, and the
models developcd apply to aH configurations in the group. For example, the
ParcIo B-2Q B-4Q and AB-4Q westbound approach are grouped togcther.

(ore MethodoIogy Chapter 23/Ramp Terminals aOOAltemative Intersections

Page 23.34 Version 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobi¡ity Analysis

Exhibit 23.24

Interchange
Tv~
Number
of Lanes in
Lane Groun
2
a,
-0.154
., ., ., ., .,
Leftm0st bane (L))

0.187 -0.181
RightmQst Lane CLJ

- - -
Parameters for lane
Utilization Models for the
External Arterial Approaches
of Diamond and Parclo
Diamond 3 -0.245 0.465 O 0.609 -0.326 O Interchanges
4 -0.328 0.684 O 0.640 -0.233 O
2 O -0.527 O - - -
Parclo -0.363 O O 0.605 O
3 O
A-2Q -0.257 O O 0.747 O
4 O
Parclo 2 0.387 -0.344 O - - -
3 0.559 -0.218 O -0.429 0.695 O
B-2:¡;B;~,
A8-4 WB 4 0.643 -0.103 O -0.359 0.794 O
Parclo 2 -0.306 -0.484 O - - -
A-4Q, AB.2Q 3 -0.333 -0.289 O 0.579 0.428 O
IEB' AB-4n-IEB' 4 -0.233 -0.237 O 0.703 0.641 O

Parclo
2 0.468 O O - - -
3 0.735 O O -0.308 O O
AB-2Q (WB) -0.202 O O
4 0.768 O O
Notes: If there is an exclusive right-tum lane on the external approach, the 0-0 demand (vf or vG from Exhib4t
23-20) should be lefO in !he respective eqU<ltion.

Lane utilízation of tt1e middle!<lne (if present) is estimated by subtraction.

In applying these parameters, the highest value is used in the eqU<ltlon,
Refer to Exhibit 23.17 for types of pardo interchanges.

When an external approach has an exclusive right-turning lane, thc O-O for
that movement (VF or ve from Exhibit 23.20) should be assumed to be zero in the
respective equation. When there is an additional approach in the upstream
intcrscction, the analyst should use the lane utilization factors of Chapter 19.
Equation 23-17 is valid for values of O less than 800 ft. The empirical models
underlying Equatian 23.17 did not consider configurations with longcr distances;
for thesc longer distances between the two intersections, vehicles tend not to pre-
position themselves in anticipation of a downstream turno In those cases, and in
the absence of field data, use of the default values of Exhibit 19.15 is
recommended. If the internallink contains dualleft turns extending to the
upstream approach, the volume in the most heavily traveled left-tuming lane can
be approxirnated as follows:
1. Use the model with number of lanes N - 1, where N is the number of
lanes of the subject external approach;
2. Estimate the leftmost lane volume; and
3. Multiply by 0.515.
Research has shown that as
Research (5) has shown that as operations approach congested conditions, operations approach congested
the lane utilization factor tends to approach 1 (Le., traffic becomes more conditions, the /ane utllization
factor tends to approach 1
uniformly distributed). Lane volume distributions observed in the field should (i.e., trafflC becomes more
be used if available because they are highly dependent on existing land uses and uniformly dlstributed).
on acccss points in the vicinity of the interchange. A lane utilization factor of 1.0
can be used when uniform traffic distribution can be assumed across alllanes in
the lane group or when a lane group comprises a single ¡ane. The Iane utilization
factors are used in the next step of the methodology to adjust the saturation flow
rates for each lane group of the interchangc.

O1apter 23/Ramp Terminats aOOAltemative Intersections Core Methcx1ology

Vers'on 6.0 Page 23-35
 Highway Capacify Manual: A Guide tor Multimodal MobiJity Analysis

Lane Utilization Adjustment for DDIs

Research at OOIs (10) has indicated imbalances in Iane utilization similar to

those described for other interchanges aboye. This section presents Iane
utilization adjustments for the external DDl crossover as a funetion of left-tum
demand ratio at the intemal crossover. In addition, the models distinguish
bctween exclusive and shared lane eonfigurations, for a total of five DOI lane
configurations shown in Exhibit 23-25. Thc lane utilization at DDls is estimated
by using Equation 23-18, with parameters for the equation shown in Exhibit 23-
26. Only the highest-volume lane is needed to ealculate the lane utilization factor.
Depending on the availability of data, the model can then be categorized into
regimcs. The lane with the highest lane utilization ratios is chosen as the
represcntative modeI for the intersection and is indicated in thc table. For design
purposes, knowledge of queue Iengths per lane is critical in ensuring that
adequate storage is provided at the ODl approach.

Exhibit 23.25 3-lane with

Five categOries fer DDI Lane exclusive and
Utilizatian 2-lane shared 3-lane shared 3-lane exclusive shared Jane 4-lane exclusive

,,
,,
,,,
--J:
~\

Exhibit 23-26
Lane Utilizatien Model
Lane Configuratlon Regime
1 (lTDR :5 0.35)
La"e
left
••
0.2129
a,
0.5250
Coefficients for DDIs 2-lane shared
11(lTDR > 0.35) left 0.5386 0.4110
H (lTDR:5 0.13) Middle -0.1831 0.3863
3-lane shared I-2 (0.13 < lTDR::;; 0.43) leftm"" 0.2245 0.3336
11(l TDR > 0.43) leflmo>t 0.6460 0.1523
1 (l TDR ::;;0.33) Middle -0.5983 0.5237
3.lane exclusive
IJ (l TDR > 0.33) leflmo>t 0.9695 0.00%
3-1ane exclusive with 1 (lTDR :5 0.50) Míddle -0.2884 0.5626
middle shared lane II (LTDR > 0.50) leflmo>t 0.4903 0.1761
1 (LTDR :;; 0.35) Center-Ieft -0.5432 0.5095
4-lane exclusive
II (l TDR > 0.35) l.eftmo>t 0.9286 -0.0071
Note: LTDR ~ left'mrn demand ratio.

Core Methodology Chapter 23jRamp Terminals and Altemative Intersectiens

Page 23-36 Version 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

%VUDD1 = al x LTDR + a2 Equation 23.18

where
%VLi.DD! percent of traffic present in fane Li for a DDI, with L1 representing
the leftmost fane, L2 representing the second fane from the left, and
so forth;
coefficient for i '"1 and i'" 2 (St.'C Exhibit 23-26);and
"
LTDR '" left-turn demand ratio (decimal), calrulated as left-turn dcmand at
external crossover divided by total approach vofume.
The maximum %Vti.DDI value is used as %Vtm.u in Equation 23-16 to cakulate
Jw for the DDL

lnterchange Saturation Flow Adjustment No. 3: DDI Factor, fDDI

Research suggests that the
Research on DDls (5) suggests that the conventional diamond interchange
DO! saturation flow rate is, on
model overestimates the saturation £low rate at DDIs. Data collected at average, 8.1% Iower than
approaches to 11 DDls showed that thc average field-measured saturation flows what is e5ómated for
convenlionaf inten:hanges, but
were lower by a factor of 0.913, with a standard deviation of 0.55. The DDI with considerable variation Ji]
adjustment was estimated aftec all rernaining tecms in the saturation flow cate that estimate.

equation wece controllcd foroThis factor suggests that the DDI saturation £low
rate ¡s, on average, 8.7% lower than what is estimated foc conventional
interchanges, but with considerable variation in that estima te.

Interchange Saturation Flow Adjustment No. 4: Turn Radius Effects on Left- or

Right. Turning Movements, fa and ~T
Traffic movements that discharge along a curved travef path do so at rates
lower than those of through movements (3). lhe turning radius has been found
to attect saturation flows tor turning movements at interchanges (3). The
adjustment factor to account for the eHects ot travel path radius JR is calculated
with thc following equation:
1
Equation 23.19
fR = 5.61
I+Tr
where R is the radius oi curvature of the lett- or right-turning path (at the ccntcr
of the path), in feet.
For protected, exclusive left-turn lanes,
ILT = IR Equation 23-20

For protected, shared Icft-turn lanes: Factors for protccted turn movements
are providcd in Chapter 19. The revised left- and right-turn adjustment factors
are calculated as fol1ows,as a function of the adjustment factor to account for the
effects of travel path radius JI" the proportion of left-turning traffic l\r, and the
proportion ot right-turning traffic P"r.
1
fLT=
1 + PLT
(1) fR - 1
Equation 23-21

Chapter 23/Ramp Terminals and A1ternative Intersections (ore Methodology

V~6.0 Page 23-37
 Highway Capacity Manual: A Guide tor Multimodal Mobility Anolysis

where
lu =:- saturation flow adjustment factor for ¡eft turns;
PLT percentage of Icft tums in lanc group; and

IR '" interchange saturation flow adjustment No. 4: tum radius from

Equation 23-19.
For protected, exclusive right-tum lanes,
Equation 23-22 1", = IR
For protected, shared right-tum lanes,
1
Equation 23-23
IRT
1 + PRT
(1)
T;-1
where
fRT saturation f10w adjustmcnt factor far right tums,

PRT '" percentage of right tums in lane group, and

IR tum radius from Eguation 23-19.

Exhibit 23-27 tabulatcs the adjustment factor for tum radius for several radiL
When the lane group is shared by several movements, the adjustment factor for
tom radii is estimated as the average (weighted on the basis of flows) of the
respective movements. Thc adjustment factors for permissive phasing are
estimated by using the procedures of Chapter 19.

Exhibit 23-27 Movement !YO'

Adjustment Factor for Turn Radius ofthe Travel Path ft Left and Ri ht Turn Throu h
Radius (fN)
25 0.817 1.00
50 0.899 1.00
100 0.947 1.00
150 0.964 1.00
200 0.973 1.00
250 0.978 1.00
300 0.982 1.00
350 0.984 1.00

5tep 4: Determine Effective Green Adjustment due to Interchange

Operations
The effective green adjustment involves three components: (a) adjustment in
the effcctive green of the upstream (external) approaches due to the presence of a
downstream queue, (b) adjustment in the effedive green of the downstream
(intemaI) approaches due to demand starvation, and (e) adjustment in the
effective green for signalized DDI ramp movemcnts due to overlap phasing.
The adjusted lost time t{ for extemal arterial approaches and ramp
approaches is estimatcd as shown below. Input variables for the equations are
derived mathematically later:
Equation 23-24 tÍ. = tI+ LO_A + y - e (arterial)
Equation 23-25 tÍ. = tI + LO.R + LoL-oDl + y - e (ramp)

Core Methodology Chapter 23/Ramp Tenninals and A1temative intersections

Page 23-38 V~ion6.0
 Highway Capacity Manual: A Guide for Multimodal Mobi/ity Analysis

where
t{ adjusted lost time (i.c., time whcn the signalized intersection is not
used effectively by any movement) (s),
1, start-up lost time (s),
lost time on cxtcmal arterial approach due to prcscncc of a
downstream queue (Iost time adjustmcnt No. 1) (s),
external ramp lost time duc to prescnce of downstrcam queue (Iost
time adjustment No. 1) (s),
LOL_DD1 lost time on signalized external ramp approach at a DOI due to
overlap phasing (lost time adjustment No. 2) (s),
y yellow-plus-all-red change-and-dearancc interval (s), and
e = extension of cffective green time into the c1earance interval (s).
The adjustcd lost time tí' for the intemal approaches is estimatcd as follo\\'s:
t~'= II + LDS + y - e Equation 23-26

where LDs (lost time adjustment No. 3) is the additionallost time due to dcmand
starvation (s).
The effective green time adjusted due to the prcsencc of a downstream qucue
is then calculated for the external approaches by using thc following equation:
g' = G +Y - t~ Equation 23-27

where g' is the effcctive green time adjusted by prcscnce of a downstream queuc
(s), G is the grecn time (s), and tí is adjusted lost time for external approaches (s).
Similarly, the effectiye green lime adjusted due to demand staryation is
calculated for the internal approaches as follows:
g' = G +Y - t~' Equation 23-28

where g' is the effective grcen time adjusted duc to demand starvation (s), G is
the grecn time (s), and tí' is the adjusted lost time for thc internal approachcs (s).
Estimation methods for the additionallost time duc to the presence of a
downstrcam link queuc and due to demand starvation are given in the following
section.

Lost Time Adjustment No. 1for Diamond Interchanges: Presence of a

Downstream Internal Link Queue on Arterial and Ramp, LD-k LD-R
Thc presence of a downstream queue may reduce or block the discharge of 7ñe presence of a downstream
queue mar reduce or block the
the upstream moyements, increasing the amount of lost time for the upstream discharge of (he upslream
phases. In the analysis of interchange ramp terminal s, the effects of thc prescncc movements, increasing the
amount of fast time for the
of a queue at the downstream link (through movement) are considered by lJpstream phiJses.
cstimating the amount of additionallost time cxpcrienced at the upstream
intersection. The methodology takes into considcration the duration oEcommon
green times between yarious phases at the two interscctions. In this chapter
common green time between Phase A in Intersection 1and Phase B in
Intersection 1Iis defined as the amount of time (in seconds) during which both
phases haye a green indication. Exhibit 23-28 ilIustrates an example of common

Olapter 23jRamp Terminals and A1temative Intersections Core Methodology

Ve~1On6.0 Page 23-39
 Highway Capacity Manual: A Guide tor MultimodaJ MobiJity Analysis

green times between the upstream and downstream through phases (CCuo) and
between the upstream ramp and the downstream through phases (CGRO).

Exhibit 23.28 Intersection 1 Intersection 11

IIIustration of Common Green Com~on Greend~mes
Phasi~-;'-Scheme Phasin" SCheme Westbound
TImes
.;¡\. J.
-------._. .... ---------
- - - - - - - - --- ...__ .-- -- - - - - - --
•
• •
- - - - - - - - - --- -.- ._.,----------- CGw

.... h.h.ji""
• ---------
CG.o

--------~l(-----.--.
......-) ll--------- -J<
•
The additionallost time due to the presence of a downstream queue in the
interna! through movement is cakulated for each of the upstream approaches
with the following equations:
Additionallost time on the externa! arterial approach:
CGuo
Equation 23-29 LD_A = GA - 0.106DQA - 5.39- -
C
Additionallost time on the ramp approach:

Equation 23-30 CGRO

LO_R = GR - 0.106DQR - 5.39- -
C
wherc
LD-A lost time on external arterial approach due to presence of downstream
queue (s) (min '" O),
lost time on external ramp approach due to presence of downstream
qucue (s) (min '"O),
G, green interval for externa! arterial approach (s), \
Gil grecn interval for left-turning ramp movement (s),
DQ.... distance to downstream queue at beginning of upstream arterial green
(ft),
DQR distance to downstream queue at beginning of upstream ramp green
(ft),

Core Methodology Chapter 23/Ramp Terminals and A1temative Intersections

Page 23-40 VernJn 6.0
 Highway Capacity Manual: A Guide far Multimadal Mability Analysis

CCUD common green time bctween upstream and downstream arterial

through green (s),
CCRD common grecn time between upstream ramp grecn and downstream
arterial through green (5), and

C = c)'c1e length (s).

If Eguation 23-29 or Eguation 23-30 results in negative values, the respective
lost times LD_~or L()'R are zero. Furthermorc, if DQ~or DQR exceeds 200 ft, the lost
time will be zero.
DQ" and DQR are cakulated as follows:

DQA = D - QA Equation 23-31

DQR = D - QR Equation 23-32

where
D distance corresponding to storage space bctwccn the two intersections
of thc interchange (ft),
Q" estimated average per lane queue length for through movement in
downstrcam (internal) link at beginning of upstrcam arterial Phasc A
(ft), and
QR estimated average per lane queue length for through movement in
downstrcam (internal) link at beginning of upstream ramp rhasc R
(it).
Thc downstream qucue Icngth (averaged across all through lanes) at the
heginning of cach upstream phasc is estimated with the following cquations:
Queue at the beginning of the upstrcam arterial rhase A:

VR CD GR)
QA = ( 0.0107 N - 7.96(; - 0.082CGUD + 7.96(; Lh Equation 23-33
R

Qucuc at the bcginning of the upstream ramp Phasc R:

VA GD CA) Equation 23-34
QR = ( 0.0107NA - 7.96(; - 0.082CCRD + 7.96C Lh

where
Q~ quclle at thc beginning of upstrcam arterial Phasc A (ft) (min = O);

QIl qUCllCat the beginning of upstrcam ramp Phasc R (ft) (min = O);

Vil ramp £low fecding subjcct gueuc (veh/h);

VA artcrial f10w feeding subject queue (veh/h);

NR number of ramp lanes feeding subject queue;

NA number of arteriallanes feeding subject queuc;

GR green intervai fOf upstream left ramp movement (s);

G~ green intcrvai fOf upstream arterial through movcment (s);

GD green intcrval for downstrcam arterial through movemcnt (s);

Chapter 23/Ramp Terminals and Altemative Intersections Core Methodology

Version 6.0 Page 23-41
Highway Capacify Manual: A Guide for Multimodal Mobility Analysis

CGUD = common green time between upstream arterial green and downstream
thraugh green (s);
CG/lO common green time between upstream ramp grecn and downstream
thraugh grecn (s);
Lh = average queue spacing in a stationary queue, mcasured from front
bumper to front bumper between successivc vehicles (ft/veh); and
e = cyele length (s).
The variables VRJ NR, VA' and NA refer to the movement flows that fecd thc
subjeet queue. For example, for a diamond interchange, vR is the lefHuming fIow
from the ramp, and the variable bccomes VRam¡>-L' For aetuated signals, the analyst
should first determine the equivalent pretimed signal timing plan on the basis of
the average duration of eaeh phase during the study hour and estimate the
parameters described aboye on the basis of that plan.
If QA or QR is ealculated to be less than zera, thc expected gueue is zero, and
no additionallost time due to the presence of a downstream gueue will be
experienced. Similarly, if the 10st time LD-A or L[)'R is estimated to be negative, the
expeeted lost time will be zero for the respective approach. Conversely, if QA or
QR exeeeds the available storage, its value should be set equal to that storage, and
the respective distance to the downstream gueue, DQA or DQRJ should be set to
zero.

Lost Time Adjustment No. 1for DDIs: Presence of a Downstream Internal Link
Queue, LD-ArLD-R
In eoordinating the two elosely spaced DDl signals, a signal designer is ab1e
to progress arterial through traffie or favor off~ramp left-turns. Beeause oí the
difference in signal phasing, the lost time adjustment íor downstream link
gueues at standard diamond interchanges cannot be readily applicd to a DO!.
The designer chooses an option on the basis of predominant fIows into and out
of the corridor during diffcrent times oí day.ln other words, progression oí
predominant traffie írom upstream .through movemcnts allows vehieles to pass
thc downstream signal without stopping, whereas, in the case of predominant
traffie fram the off.ramp left-turn movement, most off-ramp vehicles are able to
pass the downstream signa1 without stopping. Thus, additionallost time at the
upstream interseetion due to the downstream internallink qucue prcsence for
DDls is O1ostlikely to be eonsidcred for either the arterial fhrough or the off-rarnp
left-turn movements based on the progression pattem, but not both.
Input data necessary for estimating the additionallost time at DDls due to an
internal qucuc are presented in Exhibit 23-29.

Core Methodology O1apter 23/Ramp Terminals and A1temative Intersections

Page 23-42 verwn6.0
 Highway Capacity Manual: A Guide for Multimodol Mobi/jty Analysis

Potential Data SOurce Exhibit 23-29

Input Data and Units
Traffteand GeometrieData for
upstream intersectionsignal plan {s} Trafftesignal plan
Additionallost TImeat DDIs
Downstreamintersectionsignal plan (s) ,Traffiesignal plan
Favoring2!gnalplan (through or off-ramp) Traffiesignal plan
Free-f1owspeed, s,.(mi/h) Aeld measurement or default
Averagequeue spaóng in a stationary queue, L, Aeld measurement or default of 25 ft/veh
(ft/veh)
Jam density, 5am(veh/mi) Derivedfrom 4
Internallink density for arterial through Derivedfrom linklength and volume
movements, K¡ (veh/mi)
Downstream left-turn demand ratio (decimal) Fieldmeasurement
Upstream through number of lanes (In) Fieldmeasurement
Upstream off-ramp lefl;-tumnumber of lanes (In) Aeld measurement
Downstreamthrough number of lanes (In) Fieldmeasurement
Upstream through volume in the previouscycle
that queued in the downstream intersection Fieldmeasurement
(veh/ln)
Upstream off-ramp left-turn volume (veh/ln) Aeld measurement
Distance between upstream and downstream
intersections, D(ft) Fieldmeasurement
Stopping shock wave speed for arterial through
movements due te the downstream queue, VW,""I' Greenshields'mocIel
(ft/s)
Starting shock wave speed for arterial through
movements due to the downstream queue, V~~ Greenshie1ds'model
(ft/s)

Additionallost time at the upstream interseetion due to downstream internal

queuc presence at a DO! can be estimated by using the seven stcps provided
below:
(a) Obtain basic traffic and geometrie data presented in Exhibit 23-29.
(b) Estimate the internal queue length (by using methods in Chapter 19).
ldentify the position of the last queucd vehicle in the internallink from
thc upstrcam off-ramp left turn when the upstream arterial through
movements reeeive a grccn indication.
(e) Identify the first vchicle arTival of the arterial through movement,
keeping thc average queue spaeing obtaincd as part of 5tep (a).
(d) Estimate jam density Kj.>rn as the inverse of average queue headway
(default is 25 ft per passenger car) and internallink density K¡ as a
fundion of link length and vehide volume.
(e) Estimate the stopping shock wave spl.-'cd by using Greenshields' model
(Equation 23-35) with values determincd during Step (a).
-Sr x K¡ 5,280 ftfmi
~ - -~-- x ---~- Equation 23-35
W.stop - K¡am 3,600 sfh
(f) Estimate the starting shock wave speed by using Greenshiclds' model
(Equation 23-36) with values determined during 5tep (a).
5,280 ['/mi Equation 23.36
VW.start = -Sr x
3.600 s/h

Chapter 23/Ramp Terminalsand A1temativeintersections eore Methodology

Version 6.0 Page 23-43
Hjghway Capacity Manual: A Guide for Multimodal Mobiljty Analysis

(g) Estimate the additionallost time at the upstream intersection (LO-A or LO-R
as appropriate) by subtracting the stopping shock wave intersection time
from the starting sho'ck wave intersection time.
There are sorne caveats in this analytical approach:
• The sarne free-flow speed is assumed for all vehiclcs, when in actuality
there is sorne fluctuation.
• Start-up lost time and acceleration and deceleration rates are not
considered in the anaIytical estimation.
• Shock wave speeds are obtained on the basis of Grecnshields' model, and
the intemal density was cakulated on the assumption that arterial
through vehicles approach the intemal queue at 50% of frce-flow spced.
• Cycle lengths at the DDI and adjacent intersections are assumed to be the
same in a coordinated system.
Even with the caveats, this analytical approach provides a fairly intuitive
estimation of how much green time would be lost at the upstream intersection
because of downstream intemal queue presence with a two-phase signa!.
Internal queuing patterns are also like1yto be sensitive to signal progression
patterns between the two 001 crossovers, which can be configured to progress
arterial through movements, left tums from thc freeway, or a combination.

Lost Time Adjustment No. 2: Overlap Phasing on Signalized Externa! Ramp

Approach at a DD~ fOl.-DOI
Most DDIs operate by using a signal with two critical rnovements at each
ramp terminal. When one or more off-rarnps are signalized, each crossover
through movement is concurrent with either the Icft- or right-turn off-rampo
However, the off-ramp right- and left-tum movements are often a significant
distance from the crossover, which makes intersection c1carancc times (Le., all-
red) longer. Overlap phasing can allow the opposing crossover through movement
to start before the concurrent right- or lejt-tunJ maneuver. This permits efficient
crossover operation while vehicles c1earthe ramp terminals. Exhibit 23-30
illustrates how phase overlaps can be used when left- and right-turn ramp
movements from the freeway are signalized. Phase overlaps are represented as A,
B, e, and D.
The resulting reduction in effective green time for the DDI ramp movements
is accounted for through the LOL.DDllost time adjustment factor. The lost times
applied to DDI off-ramp movements are calculated on the basis of free-flow
speeds Si' c1ear-zone widths W, vehide lengths L, and the space between ramp
stop bar and conflict zone D. Distance D can be significant, since most exit-ramp
stop bars are set to accommodate appropriate sight distances at DDIs. A
graphical representation of these distances is provided in Exhibit 23.31.

Core Methodology Chapter 23jRamp Termínals and A1temative intersectíons

Page 23-44 Vef5ion 6.0
 Highway Capacify Manual: A Guide for Mu/timodal Mobility Analysis

Exhibit 23-30
Standard Phasing &heme at a
DOI with Signalized Ramp
Movements

_. ~ _. .
0_.• ~~
0_'
••••••
_. -, _.$
•••••• Ir

.- .-
Ring 1
••••••
0_.
Rlng 2 M" ••••••
o..~1)
"'- "'-
o.._~"Pc

Exhibit 23-31
Graphical Depiction of the
Distances Needed to calculate
Off-Ramp Lost TIme at DDIs

W - width of dear zone (ft)

L = length of vel1ide (ft)
o = distante 10 nearest conflict
point (ft)

The overlap phase, required to allow through movements at the crossover to

begin before off-ramp left and right tums, can be computed by using the all-red
dearance interval calculation. This overIap phase delays the start of a signalized
left- or right-tum off-ramp movement (as compared with the concurrent through
movement). In essence, this shortens the effective green for signalized ramp
terminal movements because of the additionallost time that must be considered.
For the purposes of this method, the overlap phase time is induded in the
analysis as additionallost time for signalized off-ramp movements from the
freeway. On the basis of the Exhibit 23.31 variables, lost time for a signalized off-
ramp movement can be calculated as follows:

Chapter 23jRamp Terminals and A1ternative Intersections Core Methodology

VefWn6.0 Page 23-45
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

W+L-D
LOL-DDI = -¡-.4-G-7-X-S- -
f
where
W width of the dear zone for the Iongest vehide path (ft), measured
along the centerline of the outside lane, which is the dosest conflicting
vehide path to the ramp;
L design vehide length (ft), typically 20 ft;
D distance from the ramp movement stop bar to the conflict point (ft)
measured along the centerline of the off-ramp approach; and
51 = free-flow speed of the vehide (mi/h)_

Exrubit 23-32 provides a graprucal representation of Equation 23-37. Speeds

ranging from 20 to 40 mi/h are used.

Exhibit 23.32
Lost Time for OOI Off-Ramp
Based on O¡stance Terms
'",
B

,
,
o~- _J
••• 60 80
'"o 120 140 160 180 200 220 ,...
Distance W+ L-D(ft)
I -20 mlfh -25 mI/h----30 mlfh _.- 35mlfh _.- 40mi/ti I

The speed used should provide safe passage of the conflict zone for the vast
majority of drivers. If a DDI is newly constructed, the design speed V of the
curve should be utilized to provide an estímate of speed; however, to provide
safe crossing foc the slowest vehides, it should be decreased by 5 mi/h. Although
start-up 10st time could be considered for the off-ramp movcments, the
recommendation is made that it be ignored to allow additional time for
conflicting vehides from the through crossover movement to dear safely as
drivers react to the green signal indication at the off-rampo

I
1I
Core Methodology Chapter 23/Ramp Terminals and A1temative Intersections
Page 23-46 v~6.0
11

i
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Lost Time Adjustment No. 3: Demand Stafllat;on for the Downstream (Internal)
Approaches (Los)
This methodology aceounts for the effects of demand starvation in diamond Lost time due to demand
stalVation (Los) {s defined as
interehange operations by computing the lost time experienccd at the
!he amount of green bme
downstream interseetion that rcsults from demand starvation. Lost time due to during which there i5 no queue
demand starvation (Los)is defincd as the amount of green time during which present to be di5charged from
(he interna' link and there are
there is no queue present to be diseharged from the internallink mld there are no no arrivafs from ether of the
upstreiJm approache5 due lo
arrivals from either of the upstream approaches due to signalization. The signaliZation.
common green time between t\\'o phases that may lead to demand starvation is
called commotl greell time with demalld starvatioll poteJ1tial (CCos)' Exhibit 23-33
providcs an iIIustrative example of an interval with demand starvation potentia1.
In that example, there is potential for demand starvation for the wcstbound
internal through movement of the intcrchangc.
For DDIs, the demand starvation is assumed to be zero. Because of the
geometric configuration of DDls with two directional crossovcrs and thc two-
phase signal timing scheme applicd at most DDls, there is generally no
opportunity in thc cycle for demand starvation to occur. In other words,
whenever the outbound movement at the intemal erossover has a grcen
indication, one of the upstream movements from the external crossover is being
served (either through traffie or left turns from the freeways). For diamond
interehanges, the following equation is used to estimate lost time due to demand
starvation:
LDS = CGDS - Qlnitial X h¡ Equation 23.38

where
Los additional lost time due to demand starvation (s);
CCos common green time with demand starvation potential (s), as shown in
Exhibit 23-33;
1I¡ saturation headway for internal through approach (= 3,6CHJ/saturation
flow per lane) (s); and
Q[nili.>.1 length of queue stored at internal approach at beginning of intcrval
during which this approach has demand starvation potential,
ealculated from Eguatian 23-39.

Chapter 23jRamp Terminals and Alternative Intersections Core Methodology

VesíOn 6. a Page 23-47
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Exhibit 23-33 Demand Starvatlon

IIIustration of Interval with Intersection I Inter.section 11 Potential
Demand 5tarvation Potential Phasina SCheme Phasina Scheme (Westbound Intemall

.;¡'. - - - - - _ ....
---lo
• -- - - - -. - -. a;~
------ .•.•• i1-" ---------------

• __________________ •
••••••. n_n_

----------if"
• ---------

..........
......... ~1(
)ll........ ........
Jo
•
Equation 23.39 Qlnitlal
VRamp_L X
= [ NRamp-L X
C
3,600
(CGRD - td] + [ VArt~r1al X C _ (CGUD - tL)]
h/ NArterial X 3,600 h,
where
VRamp-L upstream ramp left.turning flow (v/h),

VArt<'ri.o1 upstream arterial through flow (v/h),

C =o eycle length (s),
NRamp-L =o number of lanes for upstream ramp left-turning movement,
NArteri3l1 =o number of lanes for upstream arterial through movement,
CG/ID =o eommon green time between upstream ramp and downstream
through green phase (s),
CGUD eommon green time between upstream through and downstream
through green phase (s),
h¡ saturation headway for internal through approach (=o 3,600/saturation
flow per lane) (s), and
tL =o lost time per phase (s) from Equation 23-24or Equation 23-25.
Equation 23-38 calculates the amount of time that would not be used because
the internallink queue has eompletely discharged and the upstream demand is
bloeked and eannot arrive to the internallink stop lineoThe initial queue at the
beginning of the demand starvation interval is estimated as a funetion of the
demands of the upstream approaches and of the respective eommon intervals
between the upstream and downstream green.

Core Methodology
Page 23-48
Chapter 23/Ramp Tennlnals and A1temative Intersections
VetSion 6.0
I
I
 Highway Capacity Manual: A Guide tor Multimodal Mobi/ity Analysis

Equation 23-39 is valid for values of CGRo and CGuo <:: tL' If CGRO or CGuo< tL,
the analyst should assume that CGRO or CGuo '" tL' Also, in applying Equation 23-
39, no vehicles are assumed to have to wait for more than one eycle (i.e., none of
the approaehes is oversaturated). If the time required to discharge the queue is
equal to or larger than the CGoS'the lost time due to demand starvation will be
zero. The model for estimating lost time due to demand starvation assumes
uniform arrivals and departures and that operations at the interchange are not
oversaturated.

Step 5: Determine Effective Green Adjustment due to Closely Spaced

Adjacent Intersections
The presence of closely spaced signalized intersections in the vicinity of an The presence of cJosefy spaced
Signallted intersections in the
interehange may affect operations of the entire interchange system and ean vidnity of an interr:hange may
present unique operational challenges. Pirst, lane utilizations of arterial affed operations of the entire
interchange system and can
approaches would be affected by the presence of the interchange as vehicles present unique operational
position themselves to make a tum downstream or as they enter the arterial from chaflenges.

the interchange. Queuing from adjacent intersections could affect the discharge
rate of the upstream (intemal) link of the interchange. Furthermore, demand
starvation in the intemallink can coexist with queues upstrearn, in the extemaI
approaches of the interchange. If this extemallink is short, queue spillbaek may
affeet adjacent interseetions and have a long-lasting impact throughout the
interchange area. GeneralIy, closely spaced signalized intersections whose
signals are poorly timed can cause flow blockages on the next upstream link due
to queue spillback, even during nominally undersaturated conditions.
In the anaIysis of interchange ramp terminaIs, effeets of the presence of
closely spaced intersections are considered by adjusting lane utilizations of the
intersections' arterial approaches, by estimating the additionallost time
experienced at the upstream interscetion due to the presence of the downstream
queue, and by estimating additional1ost time due to demand starvation.
Lane utilization factors for through approaches of closely spaced
intersections should be estimated by subtracting 0.05 from the lane utilization
factors obtained from Exhibit 19-15.Research (1) has shown that those utilization
factors are generalIy lower than those at a typical intersection approach.
The additionallost times experienced at the approaehes to closely spaced
intersections are estimated as discussed in the previous section. A brief overview
is provided here for eonvenience. Additionallost time may be experienced at any
of the upstream approaches to the dosely spaced intersections. The additional
lost time due to the presence of the downstream queue is ealculated for each of
the upstream approaches i by using the following equation:
CCU/D Equation 23-40
Lo•ul = CUI - 0.106DQ¡ - 5.39- -
C
where
LD_u, lost time on upstream approach i due to presence of a downstream
queue (s),
Gu; green interval for upstream approach ¡(s),

Chapter 23/Ramp Terminals and Altemative Intersections Core Methodology

VetWn6.0 Page 23-49
Highway Capacity Manual: A Guide for MultimOdal Mobility Analysis

DQ, distance to downstream queue at beginning of upstream green for

approach i(ft), and
CCU¡D common green time between upstream approach i and dO\\'Ilstream
through green (s).
The distance to the downstream queue at the beginning of the upstream green is
calculated on the basis of the estimated average per lane queue length (in feet)
for the through movement in the downstream link at the beginning of the
respective upstream phase with Equation 23-31 through Equation 23.34. When a
significant portion of the traffic demand is held at the upstream adjacent
interscction, demand starvation can occur on the external approaches to the
interchange. The lost time caused by demand starvation on the external
approaches to the interchange is estimated in Step 4 by Equation 23-38.
When the operations of adjacent c10sely spaced intersections affect and are
affected by operations at the interchange, the external and internal approaches of
the interchange could experience both lost time due to a downstream queue and
demand starvation. For example, the internal approach of a diamond interchange
may experience lost time due to a downstream queue created at the downstream
intersection, and at the same time it may experience demand starvation. In those
cases, the procedures of this chapter should not be applied; simulation or other
alternative tools should be used instead.

5tep 6: Determine Performance of YIELD-Controlled Turns

OOIs and other interchanges may feature YIELD-controlled right and left
turns from the freeway that need to be considered in the analysis. The procedure
presented in this step estimates the delay incorred by YIELD-controlled tums at
0015, thereby allowing the operations of these movements to be compared with
signalized (right- or left-tum) or free-f1owing (right-turn) altemativcs.

Operations of Free-Flow Turning Movements

Free-flowing movements are treated the same way as free-flow bypass lanes
at roundabouts and signals, with a zero delay for those volumes. Most
important, the left-tum movement onto the freeway at the intemal signaJ of a
DDI is general1y free-flowing (other than pedestrian- and bicycle-induccd delay).
Sorne DDIs and olher interchange forms also fealure free-flowing right tums.
These free-flowing volumes are considered to have zero deJay, which is included
in the interchange's weighted average delay aggregation by approach or
movement. The zero delay is also included in the delay and experienced travel
time estimate for each O-O movement al the interchange.

Operations of YIElO-Control/ed Movements

Sorne OOIs feature 'r1ELD-controlled left or right turns that need to be
evaluated and cornpared with an alternative of signalizing the turos or providing
free-flow lanes (right turos only). YIELD-controlled right turns are also a potential
feature of standard diarnond interchanges and should be treated consistently.
Given the signals upstream of the YIELD-Controlledmovement, YIELD-Controlled
leH-tum (and right-tum) capacity is evaiuated in three f10w regimes:

(ore Methodology Olapter 23/Ramp Terminals and A!temative Intersections

Page 23-50 Version 6.0
 Highway Capacity Manual: A Guide tor Multimodal Mobility Analysis

• Regime 1: blocked by conflieting platoon when the conflicting signal has

just tumed green, resulting in zero capaeity for the tuming movement;
• Regime 2: gap aeceptanee in eonflicting traffie after the initial platoon
has cleared, with gap acceptance controlled by critical headway, follow-
up time, and conflieting f10wrate; and
• Regime 3: no conflicting flow when the conflicting signal is red, resulting
in full eapadty, controlled by follow-up time of the YIELD-controlled
approaeh.

Exhibit 23-34
COl Crn!i!iolrer Signal Contlicting Flow Regimes
IlIustrated fer OOl

\ \

Yl"fw-Controllf'd Left.Tum
• Reqime 1 - B10ded from Pliltoon
• Ró'!gime 2 - Gdp A(Ceptdnce
O Regime3 - No OPPOSi'l9 Flow

The concept of Regimes 1 and 2 is similar to the proeedure for adjacent signal
platooning eHeets on two-way STOP-controlled(TWSC) intersection operations.
However, the methodology can be greatly simplified for OOls, since eaeh turning
movement has only one souree of conflicting traffie. For a nvsc intersection on
an arterial street, platoons ocrur from four separate movements (the through
movement and the left-tum platooo from adjacent signals in two directions).
Regime 3 is a new cúncept, which requires estimation of saturation flow rates for
the YIElD-controlled movement without eonflicting movements. The three
regimes are iIIustrated in Exhibit 23-34 for a YIELD-controlledleft tumo

YIELo-Control Regime 1: B/ocked by Confliding P/atoon

The capacity of the YIElD-controlled movement during this regime is zero,
sinee tuming vehides have to yield tú the conflicting platoon discharging from
the crossover signal (inbound for left tums, outbound for right tums). The
method then reHes on the estimation of thc proportion of time blocked for
movement x, which is denotcd by Pb.•. The proportion of time blocked is equalto
the amount of time that the conflicting llow rate is high enough to resull in
headways that are too short tú be entered by the YIElD-controlled movement. The
critical platooo flow rate qe is equalto the inverse of the critical headway te for
that movement (Le., qe = 3,600/fcl, consistent with guidanee given in Chapter 30.
The appropriate critical headway for DO! tuming movements is discussed
further below. The resulling bloeked period duration t; is illustrated in Exhibit
23-35, whieh is adapted from Exhibit 30-5.

Chapter 23/Ramp Terminals alld Aliemative Intersections Core Methodelogy

Version 6.0 Page 23-51
 Highway Capacity Manual: A Guide for Mulfjmodal Mobility Analysis

Exhibit 23-35
Estimation of Blocked Period Combined Arrjval E!ow f'roljle rO( IhrOl.!:lh-lane G"oup
Duration

-"-
-~
oC
~
~
q', -
~
o
ii:

O
O e
Time (steps)

As a result, the proportion of time blocked for movement x can be estimated

explidtly as follows:

Equation 23-41

where
Pb,z proportion of time blocked for movement x (decimal),
t; blocked period duration (steps),
dI '" time step duration (s/step), and
e cycle length (s).
One challenge with the fonnulation aboye is that it requires an iterative
computation of the 001 as part of the time step-based urban street procedure. In
a stand-alone DDI evaluation, this factor may be approximated by the time
needed to clear the conflicting queue length at the upstream signal, plus the time
needed for the last vehicle in the queue to dear the travel distance to the crossover.
This method is illustrated in the Extensions to the Methodology section.

YIELD-Contro/ Regime 2: Gap Acceptance in Conflictíng Traffie

For the second regime, a gap acceptance--based capacity model is used that
mirrors the roundabout capacity procedure. Roundabout approaches and YIELD-
controlled tums at DDIs are both YIELDcontrolled, but gap acceptance behavior
needs to be calibrated separately for DDIs because of differences in geometry.
The capacity, or maximum entering fIow rate, for a roundabout q"mu can be
estimated as a fundion of critical headway tO' follow-up headway tf and the
confliding flow rate by using Siegloch's equation shown below. In this example,
qz.mJU is expressed as the capadty during the gap acceptance regime cGA•

3,600 tI te -"2 )
Equation 23-42 CGA =--exp ----xq,
tI ( 3,600

Core Methodology Chapter 23/Ramp Terminals and A1temative Intersections

Page 23-52 Ve1Sk:¥7 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

where
cGA '" capacity during the gap acceptance regime (veh/h),
qe conflicting flow rate (veh/h),

te critica! headway (s), and

tI = follow-up headway (s).
Default values for critica! headway and follow-up headway were obtained
from field data at YIELD-controlled DDIs and are shown in Exhibit 23-36. The gap
acceptanee-based capacity (Regime 2) for movement x, PGA.x' is applied for the
durabon of the DDI crossover signal green phase that is not blocked by the
conflicting pIatoon, as shown in Equation 23-43.
9 - (tCQ + tdear)
PGA.x = e Equation 23-43

where
PGIo.' •• proportion of time of gap acceptance regime (decimal),
teQ time to clear conflicting queue (s),
fdel:r time for last queued vehicle to clear distance from stop bar to yield
point (s),
g effective green time of the DDI crossover movement (s), and
e eycle !ength of the DDI crossover signal (s).

Para meter and Units LeftTums RlghtTums Exhibit 23-36

Defauft DDI Tum Calibration
Critical headway, tes) 3.' l.' Parameters
Foliow-up headway, t(5) 2.6 2.4
Capacity intercept (veh/h) 1,399 1,481
Capadty slope (veh/h) -0.00073 -0.00016

The rccommended gap acceptance capacity models are represented

graphically in Exhibit 23-37. While this discussion is focused on the parameters
of Regime 2 (gap acceptance), the intercept of the curves a1so corresponds to the
capacity under the non-conflicting flow condition in Rcgime 3.

1,600 Exhibit 23-37

PIots of Gap Acceptaoce
:c 1,400 Capacity Models
i
__ 1,200
~
~ 1,000

i!' 800

~ 600
I
E, «JO
E 1
;¡ 200
1
% o I
O 100 200 300 400 500 600 700 800 900 1,000
Conftlcting F10w (pc/h)
I .~-Ri9tt-TI,I1l Capacity -letHum capacityI

Chapter 23/Ramp Terminals and Altemative Intersections Core Methodology

Version 6.0 Page 23-53
 Hjghway Capacity Manual: A Guide for Mulfjmodal Mobilify Analysis

YIELD-Control Regime 3: No Conflicting Flow

In Regime 3, the capacity oí the YIELD-controlled movement is no longer
subject to gap acceptance, since no vehicles can arrive at the yield entry point.
The capacity dllring Regime 3 is controlIed by the saturation flow rate oí the
movement. Given the curved geometry oí the approach, this saturanon flow rate
is expected to be less than that at a standard approach to an intersection.
The satura non flow rate is deíined as the inverse oí the saturation headway,
converted to hOllrs. The capacity oí Regime 3 with no conflicting flow, C.'KF' is
estimated with Equation 23-44.
3,600
Equation 23-44 CNCF =--
tI
where
CNeF capacity oí Regime 3 with no conflicting flow rate (veh/h) and
1/ foIlow-llP headway (s).

The no-conflicting-flow capacity (Regime 3) for movement x, PNCF,z' is applied

for the duration not used by Regimes 1 and 2. Conceptually, it is equal to the
duration oí the DDI effective red phase (which is equal to the cycIe length minus
the effective green time) divided by the cycIe length, as shown in Equation 23-45.

Equation 23-45
r e-g 9
PNCF,x = C = -C- = 1-
C
where
PNCF.z proportion oí time oí no conflicting flow (decimal),
r "" effective red time oí the DDI crossover movement (s),
g effective green time oí the DDI crossover movement (s), and
e ""eycle length of the DDI crossover signal (s).

Capacity Estimation
The combined capacity oí the YIELD-controlled tum, Cm-, can be estimated by
the sum oí the individual component regime capacities, weighted by the
proportion oí time each regime is active as shown in Equation 23-46, which can
be simplified to Equation 23-47.
Equation 23-46 cYCT = cb x Pb,x + CCA x PCA;C + CNCF x PNCF;C
1
Equation 23-47 CYCT = CX [CCA X (g - tCQ - tclear) + CNCF(C - g)J
where
Cyo "" combined capacity oí the YIELD-controlled tum (veh/h);
c~ capacity during the blocked regime (veh/h), which is zero;
Pb~ '" proportion oí time blocked for isolated DDI analysis (decimal);
capacity during the gap acceptance regime (veh/h);
proportion oí time oí gap acceptance regime (decimal);

CDreMcthodology O1apter 23/Ramp Terminals and Altemative Intersections

Page 23-54 VetSiM 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

C~n capacity of Regime 3 with no conflicting flow flow rate (veh/h); and

PNCf.x proportion of time of no conflicting flow (decimal).

The unsignalized movement may be controlled by a STOPsignoThe method is
not calibrated for such cases. A STop-controlled approach is expected to result in
a reduced capadty relative to a YlElD-controlledapproach.

Step 7: Determine v/e Ratio and Queue Storage Ratio

The determination of LOS for each 0-0 invoJves the calculation of three
performance rneasures: the queue storage ratios RO' the v/c ratios, and the
average control delays. The queue storage ratios and vle ratios for each Jane
group are estimated first. If for any given lane group one or both of these
variables exceed 1.0, the LOS for every 0-0 that travels through that lane group
will be F. Next, the average control deJay for each lane group is estimated.
Finally, the average control delay for each O-O is estimated as the sum of the
control delays for each lane group through which the O-O travels.

Queue Storage Ratio Estimation

The procedure for estimating the queue storage ratio RQ is described in detail
in Chapter 31, Signalized Intersections; Supplemental.

v/e Ratio Estimation

For a given lane group í, Xi is computed with the following equation:

Xi = (~) = _V_,_ = v¡C Equation 23-48

e ¡ Sí (~) S¡9i

where
Xi vle ratio for lane group i,
Vi actual or projected demand flow rate for Jane group i (veh/h),
s, saturation flow rate for lane group í (veh/h),
gi effective green time for lane group ¡(s), and
e cyele length (s).
Note that the effective green time g should be replaced by the adjusted green
time g' if there is additionalJost time due to a downstream queue and by the
adjusted green time g" if there is lost time due to demand starvation.

Step 8: Determine Control Delay and Experienced Travel Time tor

Each O-O
This step estima tes the average control delay and ETI for each 0-0
movement. The average control delay for each lane group and movement is
estimated by using the procedures provided in Chapter 19, Signalized
lntersections. Chapters 20 and 22 provide the delay calculations for YIElD-
controlled movements and roundabouts. The average control delay for each O-D
is estimated as the total deJay experienced by that 0.0. If the O-O trave1s only
through one intersection, its average control deJay is equal to the average control

Chaptef 23jRamp Terminals and A1temative Intersections Core Methodology

Version 6.0 Page 23-55
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

delay of the respective lane group.lf the O-O travels through both intersections,
its average control delay is the sum of the delays experienced at each of the lane
groups along its path. Operations at the dosely spaced intersections are generaHy
assessed by using the procedures of Chapter 19. The additionallost time
estimation, which is computed with the procedures of this chapter, is used to
determine the adjusted effective green time for aH affected approaches. EIT is
estimated as the sum of intersection control delays di and anyextra distance travel
time (EOIT) due to diverted paths. Ttis estimated as shown in Equation 23-49:
Equation 23-49
ETT= ¿d;+ ¿EDTT
The intersection control delay for each junction is estimated by using the
control delay procedure in this chapter. For pardo interchanges with loop ramps,
EOIT may be estimated from Equation 23-SO.For diamond interchanges and
OOTs, Dt should reflect the extra distance traveled away from the center of the
interchange.
Equation 23-50 D,
EDTT=¡ 7
.4 x VD
+a
where
EDIT extra distance travel time (s);
DI distance traveled along the loop ramp or diverted movement (ft);
VD design speed of the loop ramp or diverted movement (mi/h); and
a = delay due to deceleration into a tum and acceleration after the next
tum (s), assumed to be 5 s for a loop ramp movement.

Step 9: Determine LOS

With ETI now determined in Step 8, the LOS for each 0-D movement can be
estimated by using the thresholds given in Exhibit 23.10. Note that LOS is
automaticaHy F, regardless of the ETI value, if the queue storage ratio Ra for any
movement component exceeds LOor if volume exceeds capacity.

Core Methodology O1apter 23jRamp Terminals and A1temative Intersections

Page 23.56 VetSiOn 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

4. EXTENSIONS TO THE METHODOlOGY

FINAL DESIGN AND OPERATIONAL ANALYSIS FDR INTERCHANGES

WITH ROUNDABOUTS
Roundabouts are generally analyzed with the procedures in Chapter 22.
Chapter 34 provides guidance for translating 0.0 demands into movement
demands at a roundabout, in preparation for applying the Chapter 22
procedures.

INTERCHANGES WITH UNSIGNALIZED INTERSECTIONS

Interchanges with unsignalized intersections cannot be evaluated with the
procedures of this chapter, since research has not yet been performed 00 the
operation of two dosely spaced unsignalized intersections. In the absence of such
research, the intersections of sueh interchanges can be analyzed individuany
with the procedures of Chapter 20, Two-Way STOP-ControlIedIntersections, or
Chapter 21, All-Way STOp-ControlIed Interscctions.

ESTIMATING PROPORTION OF TIME BLOCKED FOR AN ISOLATED DDI

Step 6 in the eore interchange procedure for YiElD-controlledturos requires
an iterative computation as part of the time step-bascd urban strcet procedure to
estimate the proportion of time blocked for the YIElD-controIledturning
movement. This factor may be approximated in a stand-alone OOI evaluation by
the time nceded to dear the conflicting queue length at the upstream signal plus
the time nceded for the Iast vehide in the queue to dear the travel distance to the
crossover. This is shown in Equation 23-51.
, tCQ + tdear Equation 23.51
Pb,x = e
where
pí", proportion of time blocked for iso/afed ODI analysis (decimal),
tCQ time to dear conflicting queue (s),
le••••, time for last queued vehicle lOdear distance from stop bar to yield
point (s), and
e = cyc1elength of the ODI crossover signaI (s).
For an isolated interchange with assumed random arrivals, the time to dear
the conflicting queue tCQ.I •.••• can be approximated from the queuing diagram
iIlustrated in Exhibit 23-38 and calculated from Equation 23-52.This
approximation may be overly conservative for DOI crossover signals that do not
have random arrivals, since shorter queue lengths often oceur in a coordinated
signal system. This approximation furthcr assumes that there is no residual
queue at the beginning of the red phase resulting from oversaturation in the
preceding green interva1.

Chapter 23/Ramp Terminals and A1temative intersections Extensions to the Methodology

VetWn6.0 Page 23.57
 Highway Capadty Manual: A Guide for Multimodal Mobility Analysis

-
Exhibtt 23-38
.•...•.. ,:
-
A
Queuing Representation as an
Approximation ot Time te
Oear Conflicting Queue for -
-- - , ,,,
Random Arrivals

¿:f Va¡.p - - ,,4

-• -
.•.•.•.• 001
(

u
r
r x Vapp
Equation 23-52 tCQ,free = -----
SDDl - Vapp

where
time to dear conflicting queue for an isolated interchange with random
arrivals (s),
approach flow rate (veMl),

saturation flow rate for the DDI approach (veMl), and

r '" duration of the effective red interval for the conflicting movement (s).
For a coordinated interchange (coordinated with external intersections, or
coordination between the two DDI crossovers), the assumption of a random
arrival distribution is likely not valid. For an interchange in a coordinated
system, the time to dear the conflicting queue tm<roTd can be estimated by
considering hvo distinct arrival types.
Specifically, the method distinguishes between the arrival flow rate during
red v""",r and the arrival flow rate during green V""",g' as shown in Exhibit 23-39
and calculated in Equation 23.53. This method assumes no residual queue at the
beginning of the red phase resulting from oversaturation during the preceding
green interva1. A comparison between the random and platooned arrivals shows
that the conflicting queue length and the time to dear the conflicting queue are
much reduced for coordinated arrivals, as illustrated in Exhibit 23-39. Effectively,
a reduced time to dear the queue will increase the available time for the gap
acceptance regime and will thereby increase the overall capacity of the YIELD-
controlled movement.

Exhibit 23-39
Queuing Representation as an
Approximation of Time te
Oear Conflicting Queue for 11 : •••••

Coordinated Arrivals vapp,g ~ ••~ •••••

V_,(
------.,..... "",,>si~"'~
--
,: ¡,SDO¡ v_
•..•... ,.~--

..
.:,. :
- !
• .:
r t~axxd

Exteosions te ttle Methodology Chapter 23/Ramp Terminals and Attemative Intersections

Page 23-58 Version 6.0

11
 Highway Capacity Manual: A Guide for Multimoda/ Mobility Analysjs

Cx(l-P)
t -------_-
CQ.,,",d- Som _ [p x (il)-I] Equation 23-53
Vapp e
where
time to clear conflicting queue for a coordinated interchange (s),
p proportion oi arrivals during green (decimal),
approach flow rate (vehlh),

saturation flow rate for the DDI approach (veh/h),

g duration oi the effective green interval for the conflicting movement

(s), and
e = cyde length of the crossover signal (s).
Note that in the spccial case when P = glc (proportion oi arrivals equal to
glq, Equation 23-53 simplifies to Equation 23.52. The following three conditions
must be met for either Equation 23-52 or Equatian 23-53 to apply, to avoid
violating the procedure's assumptions:
1. The green flow rate must be less than the saturation flow rate:
P x Vapp x (C/g) < 50D1 or P < (SDDI X g)/(vapp x C)
2. The approach must be undersaturated:
Vapp x C :5 500/ X 9 ar (Vapp/SODl) :5 g/C
3. The time tú clear the conflicting queue must be less than or equal to the
effective green time:

tcQ,coord :5 9
If any of these assumptions is not met, the interchange should be evaluated
as part of an urban street facility analysis by using a computational engine or
should be evaluated with alternative (simulation-based) tools.
The time for the last queued vehide to dear the distance between the stop
bar and the yield conflict point tel"" is estimated from Equation 23-54, on the
assumption that the vehicle was able to reach free-flow specd by the time it
reached thc stop bar.
Xdear
tdear= 147 xS [.DOI Equation 23-54
.

where
tck-ar time for thc last queued vehicle to clear the distance between the stop
bar and the yield conflict point (s),
xck-.r distance bctween the 001 crossovcr stop bar and the yield conflict
point (ft), and
Sf.DDI free-flow speed between the 001 crossover stop bar and the yield
conflict point (mi/h).

Chapter 23JRamp Terminals and A1temative Intersections Extensions to the Methodology

Version 6.0 Page 23-59
Highway Capacity Manual: A Guide for Multimodal Mobility Ana/ysis

60th Pb", and P~", were used in Step 6 in the procedure to calculate the overaIl
capacity of the YIELD-controlled movement. The capacity for the blocked period
(Regime 1) is assumed to be zero (Cb = O).
This method for approximating the blocked period duration is only needed
in a stand.alone DDI evaluation or in the absence of a computational engine. If
the DDI is evaluated with a computational engine and integrated into an urban
streel facility analysis, the calculations are automated by using the concept given
in the core procedure and associated discussion.

PEOESTRIAN ANO BICYCLE ANALYSIS

Most interchanges provide for pedestrian and bieyde rnovements.
Pedestrians can typically travel through the interchange parallel to the arterial to
cross the grade-separated freeway facility. At sorne interchanges, pedestrian
crossings are provided across the arterial; at others, the arterial can only be
crossed at adjacent intersections. Bicyclists may choose to use the roadway or
(where perrnitted) the sidewalk network.
No spedfic rnethodologies for pedestrian and bicycle LOS for interchanges
have becn developed to date. However, pedestrian control delay at signalized
crossings can be evaluated with methods given in Chapter 19, Signalized
Intersections, while pedestrian crossing delay al unsignalized locations may be
approximated with the procedure in Chapter 20, Two-Way STOP-Controlled
Intersections.
The quality oí service oí pedestrians and bicyclists traveling along an arterial
through an interchange may be approximated with the procedure in Chapter 16,
Urban Street Facilities. Because no specific research on pedestrian and bicyclist
perceptions and quality of service has becn conducted to date at interchanges,
this analysis should be performed with careo

Extensions to the MethodoIogy O1apter 23/Ramp Terminals and Altemative Intersections

Page 23-60 Vef5ioo 6.0
 Highway Capacity Manual: A Guide for Mu/timodol Mobility Analysis

5. A~~UCAnONS

EXAMPLE PRDBLEMS
Seetion 2 of Chapter 34, Interehange Ramp Terminals: Supplemental,
provides 11 example problems that demonstrate the computational steps for the
interchange methodology. A listing ol these example problems is shown in
Exhibit 23-40.

fxamp~ Exhibit 23-40

Problem Descriptlon AppHc!tion Usting of Inten:hange
1 Diamond ¡nterchange Operatiooal Example ProbIemsContained
2 Pardo A-2Q interchange ()p@rational in Chapter 34
3 Diamond ¡nterchange with four-phase s1gnallzationilOOQU@Ue spIl1bad: Operationat
4 Diamond interchange with demand starvatiOn Operatiooal
5 Diverging diamond interchange with s1gnalízedcontrol Operational
6 Divergillg diamond interchange with YIELD-controlled tums Operational
7 Single-point urban interchange Operatlonal
8 Diamond interchange wittl dosely spaced Intersections Operatiooal
9 Diamond interchange with roundabouts Operational
Interchange type
la Compare eight types of signalized interchanges
selection
11 Diamond inten:hange analysis using simulatioo Altematlve tools

fiNAL OESIGN ANO DPERATIDNAL ANALYSIS

Final design and operational analysis for signalized interchanges is to be
eonducted when the type of interchange is known. The objective is either to
provide final design details for LOS or to assess the interchange and pro vide LOS
and other performance measures. Two subcategories are distinguished: (a)
design analysis (where the input is the desired LOS and the outputs are design
elements) and (b) operational analysis (where the input is complete design and
the output is LOS).
Design analyses inelude highway design and signal design and are
eoncemed with the physical, geometric, and signal control characteristics of the
facility so that it opera tes at a desired LOS. For those types of analysis, the
evaluation is conducted iteratively. The input data typically required for design
analysis are fairIy detailed and based substantialIy on design attributes that are
being proposed. The objective of the interchange design analysis is to
reeommend geometric elements, sueh as the number of lanes and storage bay
length, or a signal control scheme, to maintain a given LOS. The principal inputs
for design anaIysis are the design hourly volumes and the desired LOS for a
given interchange configuration.
The objective of operational analysis is to obtain the LOS of a facility under
given traffie, design, and signal control conditions. As in design analysis, the
operational analysis is conducted for a given interehange configuration. The
input data inelude the tuming movement demands, number of lanes and their
respective lengths and channelization, and traffic control information.

Chapter 23/Ramp Terminals aOOAltelTliltive Int~ Applkations

Verskln 6.0 Page 23-61
Highway Capacity Manual: A Guide lor Multimodal MobiJity Analysis

The operational analysis for OPERATIONAL ANALYSIS FOR INTERCHANGE TYPE SELECTION
interchange type seIedion is
found in Chapter 34. lt can be This type of analysis should be used when the type of interchange is not
lJSeÓto evaluare the
operationdl performance of known yet and the analyst is interested in assessing the traffic operations of
variOuSinterchange types. lt various altematives. Oiscussion of this approach can be found in Chapter 34,
allows the user to rompa"!
eight fundamental types of
lnterchange Ramp Terminals: SupplementaL For this type of analysis, detailed
interchanges fer a given set of information is not known (e.g., signalization information, design details). The
demand fTows.
principal inputs for an interchange type selection analysis are O-O demands and
a list of feasible configurations that can be tested according to site physical and
right-of-way conditions. This type of analysis considers signalized interchanges
but does not consider unsignalized interchanges or interchanges with
roundabouts.

0.0 ANO TURNING MOVEMENT ESTIMATION

Worksheets in Chapter 34 iIIustrate how 0-0 movements can be obtained
from tuming movements and how turning movements can be obtained from
O-O movements, for different interchange typcs.

USE OF ALTERNATIVE TOOLS

Development of HCM-Compatible Performance Measures Using
Alternative Tools
Simulation tools provide a wealth of information with regard to performance
measures, including queue length, travel time, emissions, and so forth. However,
simulation tools often have different definitions for each of these performance
measures. General guidance on developing compatible performance measures on
the basis of analysis of individual vehicle trajectories is provided in Chapter 7,
lnterpreting HCM and Aitemative TooI Results, with supplemental examples
provided in Chapter 36, Concepts: Supplemental. Chapter 19, Signalized
Intersections, provides specific guidance on performance measures for signalized
approaches that also applies to this chapter. To obtain LOS for a specific O-O, the
analyst will need to obtain the performance measures for the specific approaches
using that particular O-O and aggregate them as indicated in the methodology
section of this chapter.

Conceptual Differences Between the HCM and Simulation Modeling

That Preclude Direct Comparison of Results
For interehanges, the definitions of delay and queuing are the most
significant conceptual differences between the HCM and simulation modeling.
80th are measures of effectiveness used to obtain LOS for each 0-0, and
simulated estimates of them would produce results inherently different from
those obtained by the analytical method described in this chapter.
Lane utilization is also treated differently. Simulation tools derive lane
distributions and utilization implicitly from driver behavior modeling. while the
deterministic model used in this chapter develops lane distributions froro
empirical models. Differences in the treatment of random arrivals are also an
issue in the comparison of performance measures. This topie was discussed in
detail in Chapter 19, and the same phenomena apply to this chapter.

Applications Chapler 23/Ramp Terminals and Altemative Intersections

Page 23-62 Versiorl6. O
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

In sorne cases, and when saturation flow rate is not an input, simulation tools
do not explidUy account ior differences between left-turning, through, and right.
turning movements, and all three have similar saturation headway values. Thus,
the left- and right-turn lane capacity would likely be overestirnated in those
types of tools.

Adjustment of Simulation Parameters to Match the HCM Parameters

Sorne adjustrnents will generally be required befare an alternative tool can be
used effectively to supplement or replace the procedures described in Part B of
this chapter. For example, thc parameters that determine the capacity oi a
signalized approach (e.g., steady state headway and start-up lost time) should be
adjusted to ensure that the sirnulated approach capacities match thc HCM
values.
One parameter specific to this chapter is the lane utilization on the
approaches within the interchange. Driver behavior model parameters that affect
lane choice should be examined closely and modificd ii necessary to produce
better agreernent with thc lane distributions estimated by the procedures in this
chapter.
Simulation tools do not produce explidt capacity estimates. The accepted
method oi determining the capadty oi a signalized approach by simulation is to
pcrform the sirnulation run(s) with a demand in excess oi the cornputed capadty
and use the throughput as an indication oi capacity. Chapter 7 provides
additional guidance on the determination oi capadty in this manner. The
Chapter 7 discussion points out thc complexities that can arise when self.
aggravating phenomena occur as the operation approaches capacity. Because oi
the interaction oi traffic movements within an interchange, the potential ior sclf-
aggravating situations is especially high.
In complex situations, conceptual differences between the deterministic
procedures in this charter and those oi simulation tools may make the
production oi compatible capacity estima tes impossible. In such cases, the
capadty differences should be noted.

Step-by-Step Recommendations tor Applying Alternative Tools

General guidance on selecting and applying alternative tools is provided in
Chapters 6 and 7. Chapter 19 provides recommendations specifically ior
signalized intcrscctions that also apply to interchange rarnr terminals.
One step that is speciiic to this chapter is the emulation oi the traffic control
hardware. Generally, simulation tools provide great flexibility in emulating
actuated control, particularly in the type and location oi detectors. In most cases,
simulation tools attach a controller to each intersection (or nade) in the network.
This creates problems for sorne interchange opcrations in which a controller at
one node must be connected to an approach to another nade. A diamond
interchange operating with one controller is an example oi thc complexities that
can arise in the emulation oi the traffic control schemc.

Chapter 23/Ramp Terminals and Altemative Intersectiofls Applications

Version 6.0 Page 23.63
Highway Capadty Manual: A Guide for Multimodal Mobilify Analysis

Some tools are able to accommodate complex schemes more f1exibly than
others. The ability ~oemulate the desired traffic control scheme is an important
consideration in the selection of a tool for interchange analysis.

Sample Calculations IlIustrating Alternative Tool Applications

Supplemental probIems Example Problem 1 in Chapter 34 involves a diamond interchange that offers
invoMng the use of altemiJtive
tooIs for Sígnafized inrersection the potential for ilIustrating the use of altemative tools. There are no limitations
anafysis are presented in in this example that suggest the need for altemative tooIs. However, situations in
07apter 34.
which altemative tools might be needed for a proper assessment of performance
can be introduced.
Chapter 34 indudes supplementaI examples that apply aItemative tools to
deal with two conditions that are beyond the scope oi the procedures presented
in this chapter:
1. A two-way srOP-controlled intersection in cIose proximity to the
diamond interchange, and
2. Ramp metering on one oi the freeway entrance ramps connected to the
interchange.
In both cases, the demand volumes are varied to examine the seIf.
aggravating effects on the operation of the facility.

Applications Olapter 23/Ramp Terminals and A1temative Intersections

Page 23--64 Vers'on 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Part C: Alternative Intersection Evaluation

1. INTRODUCTION

OVERVIEW
Alternative intersections are created by rerouting one or more movements
from their usual places to secondary junctions. Often, the rerouted movements
are leH tums. Alternative intersection forms such as the jughandle and the
median U-tom have been used in New Jersey and Michigan, respectively, for
decades. Other altemative intersection designs are newer. Altemative
intersections may be used to enhance safety or reduce delay. Previous editions of
the HCM provided methods by which the individual pieces of an altemative
intersection could be evaluated. However, this part provides a methodology for
evaluating an altemative intersection as a whole, including the primary junction,
any secondary junctions, and any extra rerouted travcl that motorists experience.
This methodology can be applied to three of the most common altemative
intersection types used in the United States: the restricted crossing U-tom
(RCUT), the median U-tum (MUT), and the displaced left tum (DLT).
The RCUT maintains aH mainline left, through, and right moves with no RCUTs are afso known as the j-
tum, the redlJCffkonfJid
rerouting. However, an RCUT reroutes the minor-street leH tum and through intersedkJn, the superstreet.
movements to one-way U-tum crossovers on the major street. These crossovers and the synchroni2"eÓ street.

are typically located 450 ft or more from the central junction. Because RCUTs
reroute minor.street through movements, they are typically used where minor-
street demands are below 25,000 veh/day. However, they couJd accommodate
much higher minor-strcet demand if the right-turn proportion is particularly
The Mlff iS afso known as the
high. An MUT maintains aH mainline and side street through and right moves Michigan Ieft tum or the t/lru
with no rerouting. However, it reroutes allleft turos to one-way U-tum <Uro.
crossovers typically located on the major strcct 500 to 800 ft from the central
junction. Because MUTs reroute allleft toms, they are typically used where left-
tom demands are relaHvely low. The DLT reroutes left torns to crossovers The Di T iS atso known as the
continlJOUS-t1ow intersedion.
upstream of the central junction; the left-tom traffic streams then approach the
central junction to the leH of the opposing through movement. DLTs can move
left-tum and through vehic1es during the same signal phase without conflict, so
they are typically used whcre maximum vehic1ecapacity is desired.

PART ORGANIZATION
This part is organized into five sections, induding this introductory section.
The second secHon provides more detailed descriptions of RCUTs, MUTs, and
DLTs and presents the unique operational characteristics of these three designs
that must be considered by analysts. The third presents the core analysis
methodology. The fourth describes extensions to the core methodology for
altemative intersection types not covered by the core methodology and provides
guidance for analyzing pedestrian and bicyc1eoperations. The fifth describes
potential applications of the methodology, presents example results, and
provides guidance on the use of alternative tools.

Chapter 23/Ramp Terminals and A1temative Intersections Introduction

Version 6.0 Page 23-65
Highway Capacify Manual: A Guide for Multimodal Mobilify Analysis

2. CONCEPTS

RESTRICTED CRDSSING U- TURN AND MEDIAN U- TURN

INTERSECTIONS
RCUTs and MUTs can be controlled by traffic signals, STOPcontrol on the
minor-street approach, or merges and diverges. The core computational
methodology in Scction 3 can evaluate aH three types of control at RCUTs and
MUTs with three or four approaehes. RCUTs and MUTs with signals are
typically built in urban or suburban areas with higher traffie demands, while
those with STOPsigns or merges are typieally built in rural areas on high-speed
roadways with lower minor-street traffie demands. RCUTs with signals or STOP
signs and MUTs typieally have 450 to 800 ft from the main junction to a U-tum
crossover. RCUTs with merges typically have more than 800 ft from the main
junetion to a U-tum crossover, to make the weaving maneuvers easier.
Exhibit 23-41 through Exhibit 2343 illustrate three types of RCUT designs
covered by the methodology, while Exhibit 23-44 illustratcs one of the MUT
designs covered by the methodology. These iIIustrations are not to scale, and the
number of lanes shown is ilIustrative. RCUT and MUT intersections can have
one- or two-lanc erossovers, may or may not have exclusive right-tum lanes, and
may have one to four through lanes per approaeh.
YiEI.J)control is not considered Sorne agencies use YIELOsigns to control minor-street, U-tum crossover, or
to be signiticontJy different
from STOP sign operation for left-turn erossover operations instead of STOPsigns. For purposes of the
the purpose of thíS method. operational analysis procedure, YIELD-signoperation is not considered to be
significantly differcnt from SlDr-sign operation. The remainder of this part will
simply refer to SlDP control.

Exhibit 23-41 I
Four-legged RCUT wittl I
Signals I
I
I
I L Cross-s1:red. right-tum lrafflctums right
~nalsononesideofarterial I
arelndependentofslgnalsOll 1 _ Cross-s1:red.left-tum tr<lfflcmoves tllrough
ott--"Ide l!.i!

Arterial trnffk: no different

lbn convetltlonallntef"5edl~
.. ~
Cross-stred. traffk:
~
Cross-stred: left-tum and
must tum right tllrough trnffk mak~ a U-tum
In the wide median

Soorce: Humrner et al. (6).

1 1

""",,,pis Chapter 23/Ramp Terminals and A1temative Intersections 11

Page 23-66 Version 6.0

\
 Highway Capacify Manual: A Guide tor Mu/timodal Mobility Analysis

Exhibit 23-42
L Cross-street: right-tum traffictums right four-le9ged RQJT with
_ Cross-street: left-hlm trnffic moves through Merges and Oiverges

----------------------~--------------------
-------_/
If: _ )')
, -------- .
~-----------------:------------------------
,,..,~~I"""''','eB<d'ooI
Arteriaitrafflcnodifferetlt ~

'rr-
Cross-street: trnfflc
must tum right
Cross-stred ieft-tum
aOOthrough trnfficmakesa
U-tum in tite wide median

SOOrce: HlJmmer el al. (6).

I Exhibit 23-43
I Three-legged RQJT with
I Signals
•• •
-------------- I
•
1--------------------------------
SOOrce: HlJmmer et <11.(6).

Exhibit 23-44
I Four-legged MUT with Signals
I
I
•• •
:---------- - --
I
.•
••I
I
I

Source: HlJmmer el al. (6).

Other types of RCUTs covered in tros chapter inelude three-legged with

merges and diverges, three-Iegged with STOPsigns, and four-legged with STOP
signs. Two types of four-legged MUTs are covered by this part's methodology.
Both have a signalized main junction. One has signalized U-tum crossovers,
while the other has STOP-controlled U-tum crossovers. A three-legged MUT with
a signal-controlled U-tum crossover is the same design as the three-legged RCUT
in Exhibit 23-43. This part's methodology also covers three-Iegged MUTs with a
signaI-controlled main junction and a STOP-controlled U-tum crossover.

Chapter 23/Ramp Terminals and A1ternative Intersections Concepts

Ver.Wn 6.0 Page 23-67
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysjs

DISPLACEDLEfT-TUIlN INTEltRtnONS
DLT intersections provide one or more left-tum crossover locations several
hundred feet upstream of the main intersedion. The crossover locations are
typically signalized. They may be referred to as "supplemental" intersections,
because their purpose is to supplement (improve) the effidency of the main
intersection. At "full" DLT intersections, supplemental intersections are present
on aH approaches to the main intersection. At "partial" DLT interseetions,
suppIemental interseetions exist on some, but not aH, approaches to the main
intersection. Exhibit 23-45 contrasts the full and partial DLT designs.

Exhibit 23-45
Roadway Geometry for Full
and Partíal DLT lntersections

(a) Full OlT (b) Partíal DLT

Tñese supplemental These supplemental interseetions, which are typically signal-controlled, are
intersections,
whidJ are
typically signal-controlled, are installed to eliminate left-tum conflicts and left-tum phases at the main
installed to eliminate felNlJm intersection. Left-tuming vehicles cross the opposing through traffie lanes at the
conflicts and leIt-áJm phases
at the trlJin intersection. supplemental intersection and then approach the cross street on a separate,
channelized set of tum lanes on the outside of the opposing travellanes. On
reaching the cross street, left-tuming movements are served by the same green
phase as opposing through traific, without the eonflicts that exist at a
conventional intersedion. Exhibit 23-46(a) illustrates a dualleft-tum lane
crossover at the upstream supplemental intersection that approaches the cross
street on a separate road.
Exhibit 23-46(b) shows a right-tum lane from the cross street that is
channelized to the outside of the left-tum lanes. The channelized right-tum lane
provides three possible benefits: (a) right-tuming vehicles bypass the main
intersection without stopping; (b) right-tuming vehicles bypass the downstream
supplemental intersection without stopping; and (e) an advaneed design exists in
which opposing left-tum vehides merge into the channelized right-tum lane, to
bypass the downstream supplemental interseetion without stopping. However,
some sta tes have chosen not to build channelized right-tum lanes when right-of-
way costs are predicted to exceed the operational benefits.

(O"",,", Qlapter 23/lQmp Tenninals and A1temative Intersections

Page 23-68 Version 6.0
 Highway Capacify Manual: A Guide lar Mulfimodal Mobility Analysis

Exhibit 23.46
lane Geometry fOf DLT
Interseetions

(a) DiSplaced Left-Tum Roadway (b) Rlght-Tum Olannelization Options

Sour<:e: Hughes et al. (1).

In many cases, where left-tum and through demand volumes are both
sufficiently high, DLT intersections are expected to operate more efficiently than
conventionaI intersections. However, this efficieney may depend on many
factors, including reIative lefHum and through demands on the approaches,
demands on the opposing approaches, geometric design elements such as
crossover angle, and overall intersection demand volume.1his part's
methodology may help in performing a more careful analysis of this intersection
type's operational efficiency. Further information about DLT operational
characteristics and benefits of the DLT intersection are provided in an FHWA
publication (8).

Olapter 23jRamp Tenninals and A!temative Intersections Con<ept>

V~6.0 Page 23-69
Highway Capacity Manual: A Guide for Multimodal Mobilify Analysis

3. eORE METHODOLOGY

SCOPE OF THE METHODOLOGY

Spatial and Temporal Limits

Part A of this chapter discusses spatial and temporallimits for altemative
intersection analysis.

Peñormance Measures
The operational analysis methodology for altemative intersections provides
the performance measure experienced travel time (ETI). For each O-O movement,
the ETT ineludes the control delay experienced at each junction encountered,
plus the time experienced to travel any extra distances required by the designo
The methodology computes control delays at each individual junction that
makes up the alternative intersection, and capacity and vlc ratios are available for
each of those junctions. The method also computes LOS for each 0-0 movement.

Strengths of the Methodology

The strengths of the operational analysis methodology outlined in this
scction He in its ability to estimate the performance of alternative intersection
forms and to compare their performance with that of conventional intersections
and interchanges. The following are other strengths of the methodology: (a) it
establishes a framework for the analysis of alternative intersections, for which
there is a growing necd; (b) it establishes LOS criteria for evaluation of
performance of alternative intersections and comparison with other forms of
control and configuration; (c) it provides a recommendcd method for integrating
current HCM principies, procedures, and computations into the analysis of
altemative intersections; and (d) it establishes a basis for further research into the
operation of alternative intersections (e.g., arterial weaving).

Limitations of the Methodology

The overall methodology is relatively new and has thus received a limUed
amount of validation. Analysts should also keep in mind that the two sources of
EIT -control delay and extra distance travel time (EOIT)-are weighted equally
by the LOS methodology. A different weighting is possible-for example,
motorists could weight time spcnt traveling as being less onerous than time
spent in a queue-but there is no definitive research at this point to support
providing different weights to different ETI components.

Alternative Tool Considerations

Altemative tools like microsimulation have been applied to alterna tive
intersections for many years and have been validated against field data. Results
from simulation may be translated to the LOS framework presented in this
chapter with confidence as long as analysts keep certain caveats in mind. First,
analysts should be aware that simulation models will have to be calibrated to
match sorne of the recommendcd default values provided in this procedure, such
as the saturation flow at a signalized U-tum crossover or the critical headway at

Cofe Methodology Chapter 23/Ramp Terminals and A1temative Intersections

Page 23-70 VetSoo6.0
 Highway Capacity Manual: A Guide for Multimodol Mobility Ano!ysis

a U.tum crossover with a STOPsign. Sccond, analysts must ensure that

performance measures from the simulation model match the ETT produced by
this procedure. If the simulation model provides travel times by link and delay
estimates by node, this matching could be done several ways. For example, the
analyst could use simulation results to create an estimate of ETT as the sums of
corred component dclays and travel times. An altemative procedure would be to
make simulalion runs on networks with no control delay and no extra distance
traveled. Travel time results from those runs could then be subtraded from
travel time estimates made during runs with the altemative intersection in place.

REQUIRED DATA AND SDURCES

Generally, the same data are required for performing an operational analysis
on an altemative intersection as are required for analyzing each individual
junction with the methods from Chapter 19, Signalized lntersedions, and
Chapter 20, Two-Way STOP-Controlled lntersections. A few other inputs are
needed to account for the extra travel distances covered by rerouted movements.
At the individual junctions, the required data generally inelude traffic demand
data, traffic control device data, and intersection geometry data. For individual
junctions with signal control, Exhibit 19.11 and Exhibit 19-12show the list of
required input data; signal coordination is an important feature at DLTs, MUTs
with signals at the U.tum crossovers, and signalized RCUTs, so the signal data in
Exhibit 19.12 should nol be neglected. For individual junctions with two-way
STOPcontrol, Exhibit 20-5 shows the list of required input data.
Collecting required input data for the individual junctions is not significantly
different from the process at a conventional intersection, with the important
exception of the collection of demand flow rates for each movement at an
existing RCUT or MUT. At an RCUT or MUT, the distance between main
junction and U-tum crossovers is typically too great for both to be observed by a
single person or camera. Consequently, the recording of tuming movement
demands in the field requires more labor or equipment than usual. In addition,
manual or camera observations al lhe main junction of an MUT may require
linkage to those at the U-tum crossover to determine, for example, which
vehicles tuming right from the minor street then used the crossover and were
actually left-turning vehieles. Section 2.1.4.2 of the Manual ojTrallsportafioll
Ellgilleerillg $fudies, 2nd edition, discusses how to make tuming movement
counts at RCUTs and MUTs (11).
The following are the additional required data to analyze an RCUT, MUT, or
DLT,
• Volume of U-tums on red at a signalized crossover,
• Median width at a signalized U-tum crossover,
• Distance from the U-tum crossover (at an RCUT or MUT) or leH-tum
crossover (at a DLl) to the main jundion, and
• Free-fIow speed along the major street between crossovers.
lf the free-fIow speed along the major strect belween the crossovers is
unavailable, analysts can use the method outlined in Chapter 18, Urban Street

Chapter 23/Ramp Terminals and A1temative Intersections Core MethodoIogy

Versrn 6.0 Page 23-71
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Segments, to estimate that variable. Tuming radius data al crossovers are not
required by the procedure, because of the availability of default values.
However, use of field.measured tuming radii is recommended to improve the
accuracy of the saturation flow rate estimation.

OVERVIEW OF THE METHODOLOGY

Exhibit 23.47 shows the basic steps for evaluating alternative intersections.

Exhibit 23-47
OperationalAnalysis Input Parameters
FrameworkforAltemative Geometry
Intersections Traffi::control
Demand by movement

1. Detennlne O-O Demaoos and

Movement Demaoos
Convert to demand at each junction
AdJust di5placed Ieft-tum demands

2. Detennine Lane Groops

3. Detennine Lane Utilizatlon

Mnor-street tumng movements

4. Slgnal Progre:ssion Adjustrnents

5. Additional Control-Based Adjustrnents

Un~naized elements Sígnaized elements

Saturation Flow
RTOR, UTOR

6. Estimate Junetioo-Specific Perfonnance Measures

for each junction usng appropriate HCMmapter

7. Calwlate Extra Oistance Travel Time

8. Estimate Additional Weaving Delay

9. Calwlate Experlenced Travel TIme

10. Calwlate LOS

Note: RTOR•• right tums on red, UTOR = U-tums on red.

CoreMethodorogy Chapter23{RampTerminalsandA1temative
Intersections
Page23-72 VerskJr¡ 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobi/ity Analysis

The same basic analysis steps are used Coralternative intersections as for
conventional intersections. Only Steps 1 and 10 are performed Corthe
interscction as a whole. Steps 2 through 6 are performed for each individual
junction in the alternative intersection. Steps 7 and 8 are performed for each
relevant link between a main junction and a U-tum crossover at an RCUT or
MUT. Step 9 is performed foc each movement through the altemative
intersection. The subsections bclow describe the application oC each step to each
movement at RCUTs, MUTs, and DLTs. Sincc the procedures for RCUTs and
MUTs are similar, they are discusscd together despite RCUTs and MUTs being
different types of intersections intended to serve different demand pattems.

COMPUTATIONAL STEPS
Restricted Crossing U-Turn and Median U-Turn Intersections
5tep 1: Determine O-D Demands and Movement Demands
The standard tuming movement demand pattern of left tum, through, and
right tum from each approach is converted into left tums, through vehieles, and
right tums at each camponent junction.

Step 2: Determine Lane Groups

Step 2 is performed Coreach signalized junction within the RCUT or MUT
intersection. The determination of lane and movement groups is not relevant for
RCUT junctions with STOPsigns oc merges or for an MUT with STOPsigns.
During this step, analysts determine appropriate movement and lane groups
and convert forecast or counted volumes into flow rates. The peak hour factor is
used primarily for a planning analysis when a forecast hourly volume is
providcd and an analysis of the peak 15-min period is sought. Normally, the
demand flow rate is computed as the count of vehieles arriving during thc period
divided by the length of the period, expressed as an hourly flow rate, and
without the use of a peak haur factor.
VI Equation 23-55
v---
1 - PHF
where Vi is the demand flow rate for movement j (veh/h), Vi is the demand
volume for movement j (veh/h), and PHF is the peak hour factor.
Analysts should apply the rules describcd in Steps 2 and 3 of the Chapter 19
methodology to find flaw rates per lane and movement group.

Step 3: Determine Lane UtJlization

Step 3 estimates the appropriate lane utilization factor for lane groups with
multiple lanes. This step is only needed at RCUT and MUT junctions with signals
and for lane groups with multiple lanes and at least one shared lane. Approaches
to RCUT junctions with STOPsigns or merges or MUT junctions with STOPsigns
typically have just one lane.

Chapter 23/Ramp Termínals and A1temative Intersections Core Methooology

v~6.0 Page 23-73
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

RCUT with Mu/tip/e-Lane Left- Tum Crossover

Analysts should obtain field data if available. If no field data are available,
analysts should use the lane utilization factor default values in Exhibit 19-15.

RCUT or MUT Signa/ized Dual-Lane U- Tum Crossover

Analysts should use field data if possible, since the factors involved could be
complex. If a field study is nol possible, an initial estimate of the lane utilization
factor should be based on the eventual destination of the traffic stream. Vehides
heading for the major street should be placed in the left lane, and vehides
headed for the minar street should be placed in the right lane. Analysts should
apply common sense to this initiallane assignment to ensure that the placements
are not too unbalanced.

RCUT or MUT Signalized Multiple~Lane Right Tum

Analysts should use field data if possible, since the factors involved could be
complex. If a field study is not possible, the ¡ane utilization factor should be
based on the eventual destination of the traffic stream.
For a two-Iane minor-street approach at an RCUT, left-tuming and through
vehieles heading for the U-tum crossover should be placed in the left lane and
right-tuming vehides should be placed in the cight lane. For a three.lane minor-
street approach at an RCUT, left-tuming vehides heading for the left ¡ane of the
U-tum crossover should be placed in the left lane, through vehieles heading for
the right lane of the U-tum crossover should be placed in the middle lane, and
righHuming vehicles should be placed in the right lane.
For a two-Iane exclusive right.tum lane on the minor-street approach at an
MUT, left-tuming vehieles heading foc the U-tuen crossover should be placed in
the left lane and right-tuening vehicles should be placed in the right ¡ane. lf the
minor-street approach to an Mur ineludes a shared through and right lane plus
an exclusive righHum lane, the lane distribution can be estimated with Step 4 of
the core methodology foc roundabouts provided in Chaptec 22. In aH of these
cases, analysts should consider modifying these initial distributions if they
produce large demand imbalances that do not appear logical.

Step 4: Signa/ Progression Adjustments

lhis step assembles the data to account for progression through multiple
signals properly. Exhibit 23-48 shows which signals are encountered for each
movement through each type oí four.legged RCUT analyzed by the
methodology. Exhibit 23-49 and Exhibit 23-50 show similar information for each
type of three-Iegged RCUT and each type of MUT, respectively, analyzed by the
methodology. There may be non-random acrivals at the initial RCUT signa!
encountered. Arriva!s at the second or third signals encountered willlikely have
non-random arriva!s as welI.

Core Methodology O1apter 23/Ramp Termina!s and Altemative Intersections

Page 23-74 Version 6.0
 Highway Capacity Manual; A Guide for Multimodal Mobility Analysis

lraffic Control Device Exhibit 23.48

Int~ion
Movement First seco"d !hit<! 'DTI? Junctions and Extra Travel
Time 5egments at a Four-
With signals Left from Through at U-tum left tum at main
.uncbon sional
Non, No legged ROJI
maior street aossover sianal
Major-street Ihrough at U-tum Through at main
Non, No
throuah crossover sienal iuncbon Sional
Right from Through at U-tum Right tum at main
.uncbon tinnal
Nene No
malor street crossover sinnal
Left from Right tum at main U-tum at crossover ~roU9h a:in~ain
minor street iullClion sienal sianal undíon si nal
V"
Minor-street Right tum at main U-tum at crossover Right turn at main
throuah fundíon Sianal sional iunction si"nal V"
Right from Right tum at maln
iunction sinna1 No"" None No
minor street
Wittl STOP left from left tum at main
.unction steo No"" None No
signs or maior street
merges and Major-street
diverges
None None None No
throuqh
Right from
No"" Non, Nene No
malor street
l£ftfrom Right tum at main U-tum at crossover
None
minor street .unction stoofmerae stoD/merae V"
Minor-street Right tum at main U-tum at crossover
throunh .unction ston/meme ston/meme No"" V"
Right from Right turn at main
No"" None No
miner street .uncboo stODjmeroe
Note: EOTI = extra distarocelravel time to <lOOfrom U-turn crossover.

Traffic Control Device Exhibit 23-49

Intersection
Type Movement First seco"d Ihird .DTI? Junctions and Extra Iravel
Time 5egments at a Three-
With signa1s left from Through at U-turn left tum at maln
None No legged RCUT
maior street crossover siDnal iunctien sianal
Major-street
Ihrough at U-turn
No"" Non, No
(:o~g~) crossover signal
t efT
Major-street
Through at main
through None No"" No
junction signal
(stem sld.e)
Right from Right tum at main
None None No
malor street iunction siDnal
left from Right tum at maln U-tum at
fundían SiQnal crossover si"nal
Nene y"
minor street
Right from Right tum at main
Non, Non, No
minor street iundían sienal
With STOP Left from left tum at main
Non, None No
signs or maior street iunctian stoo
merges and Major-street
diverges Nene No"" Non, No
1~,"9~\
ofT
Major-street
through No"" Non, Nene No
(stem side)
Right from
None None None No
major street
left from Right tu~o~~~ain U-tumo~~~=ver
.unction sto merae sto mer e
None y"
miner street
Right from Right tum at main
Nono Nene No
minor street iunction stoojmerne
Note: EOTI = e~tra distarocelravel time lO <lOOfrom U-turn crossover.

Chapter 23jRamp Iennlnals and A1temative lntersections Core Methodology

Version 6.0 Page 23-75
 Highway Capacity Manual: A Guide for Multimoda/ Mobility Ana/ysis

Exhibit 23-50 Intersection Tri""" Contro! Deytce

Junctionsand ExtraTravel Tv_
TIme5@grTl@flts
at an MUT
MOY~ent "m second Thl'" 'ou,," 'DTT7
Four-Iegged """ from Throughat U- Throughat U-tumat Righttum at
with Signals maiorstreet tum crossover maln luñction Ve;
maln lunctloo
at U-tum Majof-street Throughat U- Th~~h at '""""'None
" None No
cro;",.,,~ throuoh maln unction
R1ghtfrom tum '""""'"
Throughat U- Righttum at
No"" None No
malar street tum crossover maln iunction
"'" from Righttum at U.tum at Throughat
mlnorstreet maln iunction None Ve;
malnlunction
Mlnor-street Throughat
None
'""""'" None None No
th~~h maln iuOCoon
Rlghtfrom Rlghttum at
None None No"" No
mlnorstreet maln lunctiOn
Four-legged Throughat U-tumat R1ghttum at
wittlSTOP ma I~~rfrom
r street maln iu'nction crossoverstop No"" Ve;
maln iunction
signs at Major-street Th~9h at
U-tum None None No"" NO
th~h maln unction
=e~ RIghtfrom Righttum at
Nene None None No
maio<_ malnlunction
"'" from Righttum at U-tumat Throughat
Nene re;
minar street maíniunction crossoverston malnjunction
Minor-5treet Throughat
maln.lunction Nene None None No
th~h
RIghtfrom RIghttum at
None Nene Nene No
mlnorstreet maln iunction
Three-Iegged "'" from left tum at None None None No
withsigna!at malar street maln iunctiOn
malnjunctiOn Major-street
and STOP Nene None Nene None No
,:.,rough
slgns at oin
U-tum Majof-street
=e~ through
Throughat
None None Nene No
{stem Sid~' malnjunction
RIghtfrom Rlghttum at
None None Nene No
malar street malnlunction
"'" from RIghttum at U-tumat
minar street maln lunctlon None None Ve;
'""""'" >too
Rightfrom R1ghttum at
None Nene Nene No
mlnorstreet maln .unctioo
Note: EOTT = elrtril distance tlaVeI time lO aOOfrom lI-tum crossovt'r.

For mo~ts witf¡ sigools at

whicil fIOt'I-random arrlvals are
For movements with signals at which non-random arrivals are expected, the
expected, ~ best way lo best way to account for progression is to assemble a complete set of signal data,
aa:ount for progression iS lo
a~bIe a compfetese( of
including offsets, and apply the f10w profile procedure from Chapter 18. If
Sigool data, incIuding offset$, detailed signa! data are unavailable, analysts should coUect arrival type data in
and apply tñe fbw' proñle
the field. Arriva! types and field collection of arriva! types are discussed in
procedure from OIapter 18.
Chapter 18. If neither detailed signa! data nor field arrival type data are
available, analysts can estimate the arrival type of the second or third signal
encountered by using Exhibit 23-51. Defau!t values in Exhibit 23-51 assume that
signals at RCUTs and MUTs are timed for optimum progression in both majar-
street directions and that signals on each side of the RCUT majar street are timed
independently.

Core Methodology Olapter 23/RampTermlnalsand A1tematlveIntersections

Page 23-76 Vet5ioo 6.0
 Highway Capacify Manual: A Guide for Multimodal Mobilify Anolysis

Intersection pefault Auival Tyoe Exhibit 23-51

T Movement 2nd Si nal 3rd Si nal 4th Si nal Default Arrival Types for
Four-Iegged RCUT Lett trom ma" street 3 None None RCUTand Mur Movements
Ma'or-street throu h 5 None None Encountering More Than One
Ri ht from ma'or street Signal
5 None None
lefl: from minor street 3 5 None
Minor-street throu h 3 5 None
Three-Iegged RCUT Left from ma"or street 3 None None
Lefl:trom minor street 3 None None
Four-legged MUT Left from ma"or street 5 2 5
with signals at Ma'or-street throu h 5 None None
U-tum crossovers Ri ht from ma'or street 5 None None
Left from minor street 3 5 None
Four-Iegged MUT Left from ma'or street 3 None None
with stop signs at
U-tum crossovers
Lett from minor street 3 None Non,
Note: Use only if it is not possible to apply the Chapter 18 fIow prome procedure and no fleld data are available.

Step 5: Additional Control-Based Adjustments

This step estimates additional control-hased adjustments needed to
approximate control delay in 5tep 6. This step is applied for each RCUT or MUT
junction.

RCUT Jundions with Merges

The procedure assumes that there is no control delay associated with
merging onto the major street or, for vehicles that need to weave, weaving from
one side of the major street to the other. Analysts should consider whether this
assumption holds for any particular RCUT with merges, since these maneuvers
could add travel time.

RCUT and MUr Jundions with SToPSigns

The procedure from Chapter 20, Two-Way STOP-ControlledIntersections, is
applied to estimate control delay at junctions with STOPsigns. The two additional
parametcrs that must be applied at an RCUT or MUT U-tum crossover with a
STOPsign are (a) the base critical headway and (b) the base follow-up time.
Because of the U-tum crossover geometry, it is reasonable to believe that the
critical headway and follow-up time from that crossover are different from the
left-tum cases presented in Chapter 20. lf field data or representative local data
for critical headway and follow-up time are unavailable for the U-tum crossover,
default values can be applied on the basis of a site with three through lanes and a
55-mijh speed limit (12), where a critical headway of 4.4 s and a follow-up time
of 2.6 s were observed. It is reasonable that these values are less than the default
values for these parameters given in Chapter 20, since the maneuver is relatively
simple, with only one conflicting traffic stream for motorists to observe.

RCUT and MUT Jundions with Traffie Signals

The signalized intersections procedure in Chapter 19 is applied at junctions
with signals to estimate saturation flow rates, capacity, and v/e ratio. Two
modifications are made to this procedure to (a) adjust the saturation flow rate at
U-tum croswvers and (b) estimate the effects of right tums and U-tums on red.

Chapter 23/Ramp Terminals and Altemative lntersections Core Methodology

VffliOn 6.0 Page 23-17
Highway Capacity Manual: A Guide for Mulfimodal Mobilify Analysis

Chapter 19 does not provide a U-tum saturation f10w rate adjustment factor.
Exhibít 23-52 provides default values for this factor for three categories of
median width: Icss than 35 ft, 35 to 80 ft, and greater than 80 ft. The saturation
flow rate is lower with narrow medians and higher with very wide medians,
because with narrower medians, drivers have to slow to make a sharper U-tu m;
with wider medians, drivers can maintain higher speeds during a more
sweeping U-turno

Exhibit 23-52 Median Width saturation Flow Rate Adjustment Factor

MlJTand RCUTDefault Narrow (<35 ft) 0.80
saturationAdjustmentFactors Typical(35-80 ft) 0.85
forU-TumCrossovers Very wide (>80 ft) 0.95
SOurce: Hummer (1.1).

5tep 6: Estimate Jundion-Specific Performance Measures

This step estimates the control delay, v/e ratio, and queue lengths at each
junction in the RCUT or MUT.

Control De/ay andv/c Ratio

The method used to estimate control delay and vlc ratio depends on the type
of traffic control used for the movement:
• Merges: Control delay is assumed to be zero for RCUT junctions with
merges, unless Step 5 showed otherwise. Control delay and v/e ratio for
major-street left turos with YIELD signs must be computed with the
Chapter 20 methodology.
• Stop control: The Chapter 20 methodology is applied to find control
delay and v/c ratio.
• Traffic signals: Thc incremental queue acrumulation procedure from
Chapter 19 is applied to find control delay and vlc ratio.
lf desired, control delay and v/e ratio results for each junction with a STOP
sign or signal can be converted into LOS by using the appropriate exhibit from
Chapter 20 or Chapter 19, respectively.

Queue Lengths
Queue lengths must be checked by using the procedure from Chapter 19 (for
a signal) ar Chapter 20 (for a STOPsign) to ensure that queues do not spill back
into adjacent through lanes ar to another junction:
• MUTs: Queue lengths should be checked for the U-tum crossover and for
the majar street from the main junction back toward the U-tum crossover.
• RCUTs: Queue lengths should be checked for the U.tum crossover, for
the majar street from the maio junction back toward the U-tum crossover,
and for the leH-tum crossover.
If the 95th percentile queue length at any of the above lacations exceeds the
available storage space, queue spillback is likely to be an issuc for a significant
portian of the time, and the travel times produced by tros method willlikely be
significantly underestimated.

CoreMethodology Olapter23/RampTerminalsandA1temative
Intersections
Page 23-78 Vet500 6. O
 Highway Capacity Manual: A Guide for Multimodal Mobility Ana/ysis

Step 7: calcu!ate Extra Distance Travel Time

5tep 7 is conducted foe each 0-0 movement that experienccs EDIT. Exhibit
23-48 through Exhibit 23-50 showed which movements expeeience EDIT, for
each type of RCUT and MUT addresscd by this methodology. For both RCUTs
and MUTs, EDIT is alwa)'s experienced as a "round-trip" from the main
junction to the U-turn crossover and back.

RCUTs with Merges

For RCUTs with merges, EDIT is estimated as follows:
Dt + DI Equation 23.56
EDTT= 7 S +a
1.4 x I
where
EDIT extra distancc travel time (s),
D, distance from the main junction to the U-tum ceossover (ft),

Di = distance from the U.tum crossovee to the main junction (ft),

1.47 = conversion factor from mi/h to ft/s,
Si = major-street free-fIow speed (mi/h), and
a = deJay associated with deceleration into a tum and acceleration from
the tum (s).
For minoe-street left-tum movements, a is assumed lo be 10 s; for a minor-
street through movement, it is assumed to be 15 s.

RCUTSand MUTs with SToPSigns and Signa/s

For RCUTs and MUTs with STOPsigns or signals, EDIT is given by

EDTT = Dt + DI Equation 23-57

1.47 x S,
where the variables are as defined aboye. There is no term for acceleration and
deceleration in this case, because it is airead)' accounted for in the formula for
control delay at junctions with STOPsigns or signals. When multiple signals exist
along the majar street, analysts may measure free-flow speed in the field or use
the procedure given in Chaptee 18, Ueban 5treet Segments.

Step 8: Estimate Addltiona/ Weaving De/ay

5tep 8 is onl)' used for RCUTs with merges and only when, in the anaJ)'st's
judgment, there willlikel)' be significant weaving deJay.ln this case, the analyst
must develop an estimate of weaving delay from field measurements oc an
altemative tool and add it to the EDIT estimate calculated in 5tep 7.

Step 9: ca/cu/ate Experienced Trave! Time

ETI is calculated for each 0-0 movement and is the sum of the control
delays expericnced, aoy geometric delay, aod EDTI:

ETT = ¿ +¿
di EDTT
Equation 23-58

Chapter 23/Ramp Terminals and A1temative Intersections Core Methodology

Version 6.0 Page 23-79
Highway Capocity Manual: A Guide for Multimodal Mobility Anolysis

where d¡ is the control delay at each junction i encountered on the path through
the facility and EDITis the extra distance travel time experienced, with aH
values in seconds.
Exhibit 23-48 through Exhibit 23-50 showed the junctions traversed by eaeh
0-D movement for the various RCUT and MUT intersection types analyzed by
this method, and they can be applied when the control delays experieneed are
eompiled. EDTf is obtained from the results of Step 7 and (i~applieable) Step 8.

5tep 10: calculate Level of Service

In this step, the ETI for each movement is converted into a LOS by using the
eriteria given in Exhibit 23.13. If the v/e ratio for any lane group at any junction is
greater than LOor if the queue-to-storage ratio exceeds LO, the LOS is
automatically F, regardless of the ETI value.
Computation of an average ETI for the intersection approach is often
desirable. This aggregated ETT is a weighted average E1T, with each movement
ETI being weighted by its demand flow rate. This helps establish a context for an
RCUT or MUT, where movements that are rerouted to U-tom crossovers often
have relatively poor levels of serviee and are balaneed by other movements with
relatively good levels of service. The approach E1T is eomputed with Equation
23.59, with the summations being for aH movements on an approach.
¿(ET7j x vi)
Equation 23-59 ETTA = --",---
L.. vi
where
EITA approaeh experieneed travel time (s/veh),
EIT¡ experieneed travel time for movement j (s/veh),
v¡ demand flow rate for movement j (veh/h), and
j E set of aH movements on the approach of interest.
Similarly, an interseetion ETf can be eomputed with Equation 2J.6O, with
the summations being for aH movements at the intersection.
¿¡ETTk x Vk)
Equation 23.60 ETT¡ = --",---
"" Vk
where
EIT¡ = intersection experienced travel time (s/veh),
EITi experieneed travel time for movement k (s/veh),
Vl demand flow rate for movement k (veh/h), and
k E set of aH movements at the intersection.

Core MethodoIogy Chapter 23/Ramp Terminals and A1temative Intersections

Page 23-80 Ver!iion 6. O
 Highway Capacity Manual: A Guide for Multimodol Mobility Analysis

Partial Displaced left.Turn Intersections

Similar to RCUT and MUT intersections, OLT interscctions can be analyzed DLr intersedion operations are
similar ro conventi0rk3f urban
with the lO-step analysis framework presented in Exhibit 23-47. DLT intersection street and signalized
operations are similar to conventional urban street and signalized intersection intersection operations, and
therefore the DL r pnxedure
operations, and therefore the OLT procedure can be viewed as an extension of the can be vfewed as an extension
urban streets and signalized intersection procedures. Exhibit 23-53 illustrates the of the urban streets and
signalized intersection
urban street layout of a partial DLT having three signalized intersections. The procedures.
middle intersection is caBed the mai'l jntersectioll. The other two intersections,
where left tums cross over the majar street, are the supp/emental intersectiollS.
Major-street left tums are hidden at the main intersection; they are shielded by
through movements and do not need to be modeled. For advanced cases where
major-street left tums experience delay at the main intersection, altemative tool
analysis is recommended.

Exhibit 23-53

_ Jl\k _ Urban 5treet layout for a

Partial Dl T Intersection

¿e e _e..-
~ ,t(
5tep 1: Determine o-D Demands and Movement Demands
In this step, standard 0-0 demands are converted into tuming movement
demands at each component junction. For the most part, this step is performed
the same way for OLTs as previously described for RCUTs and MUTs. However,
one extra adjustment is needed for displaced left-turn approaches at the main
intersection. lhis extra adjustment is only applicable when the approach's
through and left-tuming movements are served by the exact same signal phasing
and timing. Jf this condition is met, zero lcft-tuming vehides are assumed for thc
approach.
This adjustment is made hccause displaced lcft-tum vehides are not expccted
to be delayed at the main intersection when the signals are timed properly. This
adjustment is neccssary because the Chapter 19 signalizcd interscction procedure,
applied in 5tep 6, cannot simultaneously move proh..>cted-phase left turns and
opposing through vehides. In cases where displaced left-turn vehides are
significantly delayed at the main intersection or in cases where the approach's
through and left-turning movemeots are oot served by the exact same signal
phasing and timing, use of an aItemative tooJ (most IikeJy a microsimuJation t001)
is recommended.

5teps 2 and 3: Determine Lane Groups and Lane Utilization

5teps 2 and 3 are performed in accordance with the lane group determination
and lane utiJization procedures described in Chapter 19, 5ignaJized Intersections.

Chapter 23{Ramp Terminals and A1temative Intersections Core Methodology

Version 6.0 Page 23-81
Highway Capacity Manual: A Guide tor Multimodal Mobility Analysis

Step 4; Signa! Progression Adjustments

In Step 4, rather than adjusting the arrival types, a complete flow profile
analysis should be performed as described in Chapter 18, Urban Street Scgments.
The flow profile analysis will estimate the proportion of vehicIes arriving on
green.

Step 5; Additiona! Contro!-Based Adjustments

The adjustments made during 5tep 5 provide for better estimation of control
delay in 5tep 6. At the supplemental intersections, the Chapter 19 right-tum
saturation flow rate adjustment factor (typically IRr e 0.85) should be applied to
left-turning vehicIes crossing over the opposing through lanes.
Two validity checks should be made during this step. The first check is that
grecn times at the main interscction should always be large enough to serve
displaced left-turning vehicIe demands fully. The second check app!ies to a 10w-
volume side street scenario or any scenario with re!atively short green times on
the side street, where the main street green may be started before left turners
arrive and may be terminated before left-tuming platoons are served. Because
the DLT method assumes that the entire left-turning platoon is fuUy served on
every cycIe, side street green durations should exceed the sum of (a) main street
travel time between the supplemental and main intersections and (b) displaced
left-tum queue cIearance time. If these two checks are not satisfied, this method
is less reHable, and altemative tool analysis might be needed.
An important signal-timing adjustment is also made during this step. The
adjustment is only applicable for displaced left-tum approaches at the main
intersection and onIy when through and left-tuming movements are served by
the exact same signal phasing and timing on that approach. If these conditions are
met, offsets should be set so that displaced left-tum vehicles always arrive
during the guarantL'ed grecn window at the main intersection. The offset
computation is based on severa! factors, incIudiÍlg free-flow travel times, phase
durations at each intersection, reference phases at each intersection, reference
points at each intersection, and the background system cycle length. The basic
computation process for offsets is provided below in Equation 23-61 through
Equation 23-66; Example Problems 5 and 6 in Chapter 34 iIIustrate the offset
computation process.
1. Determine the travel distance from upstream stop line to downstream
stop Jine for the displaced Icft-tum roadway TDoLT, in feet. The displaced
left-tum roadway is the roadway llsed by displaced left-tuming vehicles
as they travel from the upstream crossover at the supplemental
interscction to the stop bar at the main intersection.
2. Compute the left-tum travel time TTDI.T' in seconds, from the free-flow
speed of the displaced left-turn roadway SIOLT' in miles per hOllr; this
calculation incIudes a conversion from miles per hour to feet per second:
TDDl,T
Equation 23-61 TTDLT = -----
Sr.DLT x 1.47

Core Methodology Chapter 23/Ramp Termínals and Altemative Intefsections

Page 23-82 VetSiOn 6.0
 Highway Copacity Manual: A Guide for Multjmoda/ Mobility Analysis

3. For the upstream supplemental intersection, obtain the duration betwccn

the reference point and the start of the displaced leH-tum phase LAGDW
in seconds. For the downstream main intersection, obtain the duration
between the reference point and the start of the major-street through
phase LAGm, in seconds. These durations should be based on input
phase splits instead of output phase durations.
4. Obtain the offsets at the upstream supplemental intersection 0supp and
the downstream main intersection DMA/N' both in seconds.
5. Compute the system start time of the displaced left-tum phase STDLT, in
seconds, for the upstream crossover at the supplemental intersection:
STVLT = LAGVLT + °supp Equation 23.62

6. Compute the system start time oí the major-street through phase ST TH at

the main intersection, in seconds:
STTH = LAGTH + OMA/N Equation 23.63

7. Change 0supp so that ST TH is equal to STOLT + 17 Dn :

0supp(s) = OsuPP - STDLT + STTH - TTDLT Equation 23-64

8. lf the offset value is greater than the background eycle length value,
decrement the offset value by the eycle length e to obtain an equivalent
offset within the valid range:
If 0supp > e then 0supp = 0supp - e Equation 23-65

9. If any offset value is lower than zero, increment the offset value by the
eycle length to obtain an equivalent offset within the valid range:
Ir Osupp < o then Osupp = Osupp +e Equation 23-66

5tep 6: Estímate Junetíon-5pecific Performance Measures

Offsets computed in Step 5 will influence the proportion of vchicles arriving
on green (PVG), according to Chapter 18's methods. This PVG will then affect the
computation of control delay and v/e ratio at each component intersection,
according to Chapter 19's methods.

Steps 7 and 8: Ca/cu/ate Extra Dístance Trave/ Tíme and Addítiona/ Weavíng
De/ay
EOIT and weaving delay are assumed to be negligible for OLT intersections.

5tep 9: Calcu/ate Expedenced Trave! Time

EIT is ca1culated for each O-O movement in the same way as described
previously for RCUTs and MUTs.

5tep 10: Ca/cu/ate Leve/ of Serv/ce

LOS for the OLT intersection is based on EIT. A weightcd average EIT is
computed for the intersection by using Equation 23-67.To avoid double-
counting vehicle trips inside the spatial boundaries (see Exhibit 23-5), the
summation of 0-0 demand volumes VOD should be idcntical to what it would be
for a conventional intersection.

Chapter 23/Ramp Termillills and A1ternative Intersections (ore Methodology

Verskln 6.0 Page 23-83
Hjghway Capacity Manual: A Guide tar Multimodal Mability Analysjs

¿(d/xvJ
Equation 23-67 ETTDLT = ~
£.. VOD

where
EITDLT = weighted average experienced travel time for the OLT intersection
(s/veh),
di control deJay for movement j (s/veh),
vi demand volume for movement j (veh/h), and
VOD 0-0 demand volumes (veh/h).

FuI! DLT Intersections

Overview
The analysis of a fuIl DLT intersection involves aggregating the resuIts of
two partial DLT analyses (one in the east-west direction and one in the north-
south direction). Exhibit 23-54 illustrates both partial DLT analyses, which share
one main intersection. Channelized right-tum lanes (separated from the main
intersection) and the demand volumes that use channelized right-tum lanes
should be omitted from the analysis.

Exhibit 23-54
Urban 5treet Layout for a Full
DlT Intersection
: ~ 1
1 ¡ I
0_ -

0
¿
-

•-1
R
1__
.4 .7
_1- ñ - _1- - - --
1 ) l ••••...1
(l-
.1l_I _
1 1 1
1 • 1
1
1
,¡ 1
1
Pseudo-Ríght- Tum Assígnment
The "pseudo--righHum" concept exists mainly to overcome fundamental
limitations in the HCM computational framework, in which protected left tums
cannot move together with opposing through movements. An extra step for the
analysis of full DLT intersections is to assign pseudo--right-tum movements to
one or more approaches to the main intersection. This extra step compensates for
the prior assumption of zero displaced left.tum flows at the main intersection,
which was necessary for accurate modeling of the main intersection. Unless the
pseudo--right-tum technique is used, this assumption willlead to incorrect
modeling oí downstream intersections. Pseudo--right turns allow íor proper flow
balancing and flow profiles at downstream intersections (14). Delays are not

Core Methodology Chapter 23/Ramp Tenninals and Altemative Intersectlons

Page 23-84 ver.stcn 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

tabulated for pseudo-right tums, becausc the displaced left-tuming vehides they
represent are typically not delayed at the main intersection.
Pseudo-right tums were not needed for partial DLTs, beeause there was no
concem over incorreet modeling of downstream intersections. Within the spatial
boundaries of a partial DLT analysis, no downstream intersections exist for
displaced left tumers once they pass through the main intersection. However, at
foil DLTs, displaced left tumers are sometimes stopped at a downstream
supplemental intersection.ln the advanced design under which opposing left-
tum vehides merge into the ehannelized right-tum lane and bypass the
downstream supplemental interseetion without stopping, pseudo-right turns
should be omitted from the analysis.
In the example iIlustrated in Exhibit 23-55, pseudo-right tums are defined on
the southbound approach. Displaced left tums on the northbound approach are
then removed from the analysis. Pseudo-right tums should only be defined at
the main intersection and not at the supplemental intersections. Traffie demand
and number of lanes should match the displaced left tums. For maximum aecuracy
of the downstream f10w profiles, the Chapter 19 saturation flow adjustment
factor for left tums (typically ftT = 0.95) should be applied to the pseudo-right
tums. Start-up lost times should be set to zero for the pseudo-right tums to
refleet the uninterrupted flow of displaced left turns.

Exhibit 23.55
Pseudo-- Example Conversion of
right turns Displaced Left Tums lo
I Pseudo-Right Tums
I ,

11 11111111
11111111 I
1111
Displaced
left turns

Signal Timing Considerations

When signal timing for the two partial DLT analyses is defined, the same
background cyde length should be used at all intersections. Furthermore, the
main intersection should have exactly the same signal timing in both parHal DLT
analyses. Finally, the maio intersection should have fixed.time control (no
actuation), to allow guaranteed grecn windows along both majar and minor
streets. These guaranteed green winrlows allow displaced left-turning vehides
on any approach to move through the main intersection without stopping.
Actuation at the supplemental interseetions is feasible and will not reduce
guaranteed grecn windows at the main interseetion.

Chapter 23/Ramp Tenninals and Altemative Intersections Core MethodoIogy

V~6.0 Page 23-85
Highway Capacity Manual: A Guide for Mu/fimodal Mobility Analysis

Performance Measure Calculation

To avoid double-counting delays and to account for flow profiles properly,
minor-street performance measures should not be tabulated in the partial DLT
analyses. Instead, major-street performance measures from both analyses are
combined and averaged. Example Problems 16 and 17 in Chapter 34 illustrate the
volume-weighted averaging of control delays.

Core Methodology Olapter 23/Ramp Termínals and Altemative Intersed:ions

Page 23-86 Version6.0
 Highway Capacity Manual: A Gujde for MuJtimodal MobUity Analysjs

4. EXTENSIONS TO THE METHODOLOGY

EVALUATION OF OTHER ALTERNATlVE INTERSECTION FORMS

The methodology described in Section 3 can be extended to other types of
RCUT and MUT intersections, including the following:
• An RCUT with direct minor-street left turos and rerouted major-street Idt
turos (15),
• An RCUT with no direct Idt tum,
• An MUT with three or four U-tu m crossovers, and
• An MUT with one or two direct left turos allowed ("partial MUT').
In all o£ these cases, the basic analysis £ramework remains the same. The
control delay can be estimated at each junction for each movement, the EDTI can
be estimated £oreach link £or rerouted vehicles, and the ETI can be estimated as
the sum of those componcnts.
Sorne RCUTs and MUTs providc a two-way driveway or side street at the
end o£ the U-tu m crossover, as illustrated in Exhibit 23-56.Thc side street is in
the upper-Ie£t comer of Exhibit 23-56. When demand from the side street is
substantial, the traffic stream from the crossover behaves like a permissive left
tum £rom a shared le£t and through lane, and the corrcsponding methodology
£rom Chaptcr 19 should be applied to estimate the saturation £low rate o£ the
crossover lane group.

Exhibit 23-56
Side $treet at the EOO01 an
MlJT U-Tum Crossover in
Michigan

Source: @2015GoogIe.

The concepts presented in this chapter could also be used to analyze othcr
types o£ altemative intersections that have been built in the United States. For
example, jughalldle intersections use right-side ramps to reroute le£t turos from thc
major street. The qlladrant roadway ilJtersectiOIl reroutes aH four left tu ros at a four-
legged intersedion to a coonedor betwecn thc two intersecting roadways (7).
Finally, the colltillllOUS greell T-illtersectioll is a three-Iegged design in which the
major-street through movement on top o£ the T is separated from thc rest oí the
intersection and does not travel through the traffic signal (7).

Chapter 23jRamp Terminals aOOA1temative Intersections Extensions to the Methodology

Vers.0n 6.0 Page 23-87
 Highway Capacity Manual: A Gu;de far Multimodal Mobility Analysis

PEOESTRIAN ANO BICYCLE ANALYSIS

Minor-Street Crossings at Signalized RCUTs
At an RCUT with signals, the experience foc pedestrians and bieyclists
crossing the minor street is, from an operational standpoint, not significantly
different from their experience at a conventional intersection. Therefore, the
operational analysis procedures for pedestrians and bieyclists from Chapter 19,
Signalized Intersections, can be applied directly for those movements.

Major-Street Crossings at Signalized RCUTs

Exhibit 23-57 shows the typical pedestrian crossing designs at a signalized
four-legged RCUT. For example, a northbound pedestrian crossing the majoc
street on the cast side of the minor street must make a three-stage crossing, from
D to C, then from C to E, and finally from E to B. The movernent from C to B will
almost always be in two stages, with the pedestrian likcly waiting for the WAlK
signal from E to B, because the signals on each side of the RCUT main street are
typically timed independently of each othcr.
During each of the three crossing stages, the operational analysis procedures
foc pedestrians and bieyclists from Chapter 19 can be applied directly. The
complete three-stage movement will then receive three individual sets of
performance measures and LOS.

Exhibit 23-57
Typical Pedestrian Crossing of
a Four-legged Signalized
RQJT

•• • -~-
• ••

SOuI'ce: Humml!f et al. (6).

Exhibit 23-58 shows the typical pedestrian crossing designs at a signalized

three-legged RCUT. For example, a northbound pcdestrian crossing the major
street on the east side of the minar street must make a thrce-stage crossing, from
D to C, then from C to E, and finally from E to A; a northbound pedestrian
crossing the major street on the west side of the minor street must make a two-
stage crossing, from C to E and then from E to A. Exhibit 23-58 shows that the
crossing from E to A is signalized. Signals such as this on RCUTs and MUTs are
typically easy to coordinate with other signals a!ong the arteria! and introduce
minima! vehicle delay.

Extensions to the Methodology Chapter 23/Ramp Terminals and A1temative Intersections

Page 23-88 Version 6.0
 Highway Capacity Manual: A Guide lar Multimodal Mobility Analysis

Main pedestrian crossing Exhibit 23-58

Typical Pedestrian Crossing ef
a Three-le<]ged Signalized
RUIT

e o

Source: Hummer et al. (6).

At a typical signalized RCUT, a bieydist crossing the major street has a

choice of using the pcdestrian crossing as illustrated in Exhibit 23-57 and Exhibit
23-58 (whether riding or dismounted) or of using the vehicle path to and from
the U.turn crossover. In either case, the methods described in Chapter 19 to
estimate performance measures and LOS for bicydes crossing the intersection do
not apply.

Crossings at MUTs
At an MUT, the operational analysis procedures for pedestrians and bieydes
from Chapter 19 can be applicd directly. The pcdestrian LOS score cquation
(Equation 19-71)and the bieyde LOS scoee equation (Equation 19-79)were not
developed fram user experiences at MUTs, but the pedestrian and bicyele
crossing experiences at MUTs are not that different from conventional
intcrsections, so the results should still be useful.

Crossings at DLTs
DLT pedestrian crossings differ significantly from the pedestrian crossings at
conventional intcrsections, and these differenccs can affect both pedestrian and
vehicular delay (8). As a result, the Chapter 19 anaIysis procedores for
pedestrians and bicydes are not applicable to DLT intersections. Instead, there
are two basic methods for handling pedestrians at a DLT. One method prioritizes
pedestrian safety over displaced Icft-turning vehide operations. The second
method does the opposite, by favoring vehicles (8).
In the first method, pedestrians can pass between the four outer corners of
the intersection, similar to a conventional intersection. However, this method
would often require displaced left-turning vehicles to stop at the main
interscction, which would dcfcat the purpose of continuous flow. The only way
to avoid stopping displaced lcft-turn vehicles would be for pedestrians to receive
the WALKsignal after left-turning platoons had cleared. This might not be

Chapter 23jRamp Terminals and Altemative Intersections Extensions te the MethodoIogy

VersiOn 6.0 Page 23-89
Highway Capacity Manual: A Guide lor Multimodal Mobility Analysis

possible undee all combinations of vehicular and pedestrian demando In

addition, the combination oi long pedestrian crossings (e.g., requiring 40 s of
walk time) and the nt..>edfoe separate pedestrian phases can sometimes (a) require
inefficient longer eycle lengths and (b) cause pedestrians to be the critical
movement affecting signal timing designo
In the second method, pedestrians must move between safe "refuges" on
their way across the intersection. This prevcnts any conflict betwecn pedestrians
and displaced left-turning vehicles. In some cases, pedestrians would endure
high-speed vehicle movements on both sides of the refuge area. The second
method is illustratcd in Exhibit 23-59..

Exhibit 23-59
Two-Stage Pedestrian
Crossing at a DLT

Source: Steyl1 et al. (8).

Extensions to the Methodology O1apter 23/Ramp Tenninals and A1temative Intersections

Page 23-90 Version 6.0
 Highway Capacity Manual: A Guide for Mulfimodal Mobility Analysis

5. APPLICATIONS

EXAMPLE PROBLEMS
Exhibit 23-60lists the example problems presented in Chapter 34, lnterchange
Ramp Terminals: Supplemental.

Example Exhibit 23-60

Problem Description Application Listing of Altemative
12 Four-Iegged ROJT with merges Operational Intersection Example
13 Three-Iegged RCUT with STOPsigns Problems Contained in
Operational
14 Four-legged RCUT with signals Olapter 34
Operational
15 Four-Iegged MUr with STOPsigns at U-tum crossovers Operational
16 Partial Dl T with signals (displaced left tums on two approaches) Operational
17 FuI! Ol T with signals (displaced Ieft tums on four approaches) Operational

EXAMPLE RESULTS
This section presents the results of applying this chapter's method in typical
situations.

DLT Intersections
The relevant input parameters and the sensitivity oi results to those inputs
are much the same for DLTs as for conventional intersections. However,
comparison of iacility type performance at various traffic demand levels can be
informative. Exhibit 23-61 compares average control delay per vehide for a
conventional intersection with that of an intersection having displaced leH turns
on one or more approaches. Original turn movement volumes are representcd by
a demand multiplier of 100%.Thesc demands were then multiplied by 25%, 50%,
7So/o;>, and 125%. Interscction conditions used to genera te Exhibit 23.61 involved
heavy left-tum demands on the major street and low demands on the side street.
Lane geometry for the DLT was designed to match lane geometry for the
conventional intersection, to the extent possible. Turn movemcnt demands were
also identicaL Results suggest that the partial DLT configuration would be
efficient under this specific set of volume demands.

160 , Exhibit 23.61

Example Results Companng
Ol T and Conventional

:c 120 j Intersection Performance

-'"
~

¡; 80
¡¡
g
8:[-"-1~
o 20 40 &O 80 100 120 140
Demand Multiplier (%)

-+-COnventional ...•.. üne DlT __ Tv.o DlTs -Three DLTs -Four DLTs I

Chapter 23/Ramp TeflTlinals and A1temative Intersections Applications

Version 6.0 Page 23-91
Highway Capacity Manual: A Guide for Multimodol MobUity Anolysis

Exhibit 23-64
Sensitivity of lOS te Changes
--iD~.~m~.;;";;d¡j(~yi.h~{¡¡h~):--r""!c50;;"~I;;t9;;1~Di.~I'~Y~("'~IV;;.~h~)-T-:----i:LiiD:SS::--=-:--
Minor-5treet Minor-5treet Minor-5treet U'Tum Minor-Street Minor-Street
in Minor-$treet Demand LeftTum RiohtTum ADDroach Crossover LeftTurn RiohtTum
25 25 11.6 9.2 D B
50 50 12.3 9.4 D 6
100 100 14.8 9.7 D 6
150 160 19.4 10,0 D 6
200 200 27.6 10.5 D e
250 250 51.0 11.0 E o
300 300 104.3 11.6 F F

Right Turns and U.Turns on Red

The performance of MUT and RCUT intersections is particularly aided by
right turns and U-turns on red, because demands for those movements are
higher than at conventional intersections. Exhibit 23-65, based on Example
Problem 15 in Chapter 34, shows the effect on EIT at a MUT when right turns on
red are allowed from the minor-strl..'t?tapproaches (the minor street contains the
casthound and westhound approaches) with exclusive right-tum lanes. If 40% of
the right-turning volume (which includes the traffic that will eventually turn left)
is able to tuen on red with an estirnated zero control deJay, EIT will be reduced
by more than 11 s/veh for sorne of the minor-street rnovernents. This will affect
LOS in sorne cases.

Exhibit 23-65 ETT (s/yeh)

5ensitivity of En to Movement
0% RTOR 10%RTOR 20% RTOR 30%RTOR 40%RTOR
Percentage of Minor.$treet NB left 78.0 77.9 77.7 77.6 77.4
Traffic Turning Right on Red SBIoft 56.1 55.9 55.8 55.6 55.4
NB through 9.3 9.3 9.2 9.2 9.1
5B through 12.3 12.2 12.1 12.0 11.9
NB tight 9.4 9.3 9.3 9.2 9.1
5B right 13.7 13.6 13.5 13.4 13.3
EBleft 67.4 64.3 61.4 58.7 56.1
W61ett 87.5 85.2 82.9 80.7 78.5
EB thr0tJ9h 25.1 25.2 25.3 25.3 25.4
W6 thr0tJ9h 22.2 22.2 22.3 22.3 22.4
EB right 23.7 20.6 17.8 15.1 12.6
WB ht 20.2 18.0 15.8 13.7 11.6
Note: NB " J'lOrtt1bound, S6 " southbound, Ea " eastbound, WB " westbound, RTOR " right turn on red.

PLANNING-LEVEL ANAL YSIS

Intersections with Signal Control

The methodology described in Section 3 is for the operational analysis of an
RCUT, MUT, or DLT. A planning-level analysis for any of the three designs with
signal control is possible and may be useful when detailed data are unavailable.
In this case, analysts should use the simplified method for determining the
critical intersection volume-to-capacity ratio XCI described in Chapter 32,
Signalized lntcrsections: Supplemental, at each signalized junction associated
with the intersection. The junction with the highest volurne-to-capacity ratio will
be "critica!" to overall intersection performance. The highest v/c ratio at any point
at an interscction is a useful predictor of overall intcrsection performance at a
planning level, although sorne factors affeding opcrational performance, such as
sorne nuances of signal control and operational effects of geometrics, are not
captured (6). The sirnplified method for determining the v/c ratio at each junction
also will not capture the EDIT experienced by rerouted vehicles.

Applications Chapter 23/Ramp Terminals and A1ternativeIntersections

Page 23-94 Version 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Intersections with SToPControl

A planning-Ievel analysis of RCUTs and MUTs with sroP-controlled junctions
can be performed by applying the planning-Ievel procedure described in Chapter
20, Two.Way STOp.Controlled lnterseetions, at each of the STOP-controIled
junctions. The Chaptcr 20 planning method uses <111 of the geometric and traffie
data required for an operational analysis, and the eomputations are identieal.
However, input variables for heavy vehicle pereentage and peak hour factor are
typically estimated (or defaults used) whcn planning analyscs are pcrformed.

RCUT Intersections with Merges

A planning analysis of a merge-eontrolled RCUT intersection applies the
Chaptcr 20 planning-Ievel proeedure for STOP-controlledinterseetions at the two
YIELD-eontrol1ederossovers to determine whether either is at or near eapacity.
Such an approach is conservative, since a YIELD-controlledmovement would
have a higher capacity than a STOP-eontrolledmovement. Howevcr, even at an
eady projeet devclopmcnt stage, most analysts are Iikely to conduct an
operational analysis rather than a planning analysis, because the effort to do so is
only inercmentally greater (6).

O1apter 23jRamp TermiJlals aOOA1terJlative Intersections Applications

Version 6.0 Page 23-95
Highway Capacity Manual: A Guide for Multimodal Mobi/jty Analysis

6. REFERENCES

Some of these references can 1. Elefteriadou, L., C Fang. R. P. Roess, E. Prassas, J. Yeon, x. Cui, A. Kondyli,
be foond in the TedmiCal
Reference Library in VoIume 4. H. Wang. and J. M. Mason. Capacity and Qllality o/ Seroice o/lnterchange Ramp
Termina/s. Final Report, National Cooperative Highway Rescarch Program
Project 3-60. Pennsylvania State University, Univcrsity Park, Mareh 2005.
2. Elcitcriadou, L., A. Elias, C Fang. C Lu, L. Xie, and B. Martin. ValidatiOIl and
Enhoncemmt o/ the Highway Capacity Manlla1's Interchange Ramp Termina/
Mdhod%gy. Final Report, National Coopcrative Highway Researeh Program
Project 3.60A. University oi Florida, Gaincsville, 2009.
3. Mcsscr, C L and J. A. Bonneson. Capacify o/ Interchange Ramp Termina/s. Final
Report, National Cooperative Highway Research Program Projeet 3-47.
Texas A&M Researeh Foundation, College Station, Apri11997.
4. Bonneson, J., K. Zimmerman, and M. Jacobson. Revietv ond Eva/uation o/
Illterchange Ramp Desigll Considerations for Facilifies Without Frontage Roads.
Research Report 0-4538-1. Cooperative Research Program, Texas
Transportation Institute, Texas A&M University System, College station,
2004.
5. Federal Highway Administration. EDC-2 Intersection and Interchange
Geometrics websi te. https://www.fhwa.dot.gov/everydaycounts/edctwo
/20]2/. Accesscd Dcc. 3D,2014.
6. Hummcr, L B. Ray, A. Daleiden, P. Jenior, and J. Knudscn. Restricted
Crossillg U- Turn ln/ormatiolla/ Guidc. Report FHWA-sA.]4-070. Federal
Highway Administration, Washington, D.C, 2014.
7. Hughes, W., R. Jagannathan, D. Sengupta, J. Hummer, and M. smith.
A/ternative Infersections/lnterchallges: /Il/ormatiollal Report (AIIR). Report
FHWA-HRT-09-Q60. Federal Highway Administration, Washington, D.C,
2009.
8. Steyn, H., Z. Bugg. B. Ray, A. Daleiden, P. Jenior, and J. Knudsen. Displaced
Left Turn /nterscctiOIl IIl/ormatiollal Cuide. Report FHWA-sA-14-068. Federal
Highway Administration, Washington, D.C, 2014.
9. A Policy 011 Geometric Dcsigll o/ Highways mld Streels. American Association oí
Sta te Highway and Transportation Officials, Washington, D.C, 2011.
10. Yeom, C, B. J. Schroedcr, C Cunningham, C Vaughan, N. M. Rouphail, and
J. E. Hummer. Lane Utilization at Two-Lane Arterial Approaehes to Double
Crossover Diamond Interehanges. In Trallsportatioll Research Record: ¡oumal o/
the Trallsportation. Rescarch Board, No. 2461, Transportation Research Board oí
the National Academies, Washington, D.C, 2014, pp. 103-112.
11. Schroeder, B. J., C M. Cunningham, D. J. Findley, J. E. Hummer, and R. S.
Foyle. Manllal o/Trmlsportatioll Ellgilleerillg Studies, 2nd ed. Institute oí
Transportation Engineers, Washington, D.C, 2010.

References Chapter 23/Ramp Terminals and Altemative Intersections

Page 23-96 Ver.sk1n 6. o
 Hjghway Capacity Manual: A Guide for Mu/tjmodal Mobility Analysis

12. Hummer, J. E. Proposed Critical Headway and Fo/low-Up Time HCM Procedure
1nputs. Draft Technical Memo, Wayne State University, Detroit, Mich., June
24,2014.
13. Hummer, J. E. Pro¡lOsedMUT alld RCUT HCM Procedure IIlPUtS, U-Turll
Crossover Safuratioll Flow Adjustmellt Factor. Draft Technical Memo, Wayne
State University, Detroit, Mich., June 21, 2014.
14. Hale, D., A. Kondyli, F. T. Creasey, and J. Ma. Suggested RTOR
Methodology Improvements for Urban Street Segments in the HCM.
Presented at 94th Annual Meeting of the Transportation Rescarch Board,
Washington, o.c., 2015.
15. Hummer, J. E., V. J. Blue, J. Cate, and R. 5tephenson. Taking Advantage of
the F1exibility Offered by Unconventional Designs. ITE ¡oumaI, VoL 82, No.
9, Sept. 2012.

Chapter 23/Ramp Tenninals and AJtemative Intersections References

Version 6.0 Page 23-97
 Highway Capacity Manual: A Guide for Multimodol Mobilify Anolysis

CHAPTER 24
OFF-STREET PEOESTRIAN ANO BICYCLE FACILITIES

CONTENTS

1. INTRODUCTION 24-1
Overvie,v 24- 1
Cha pter Organizalion 24-1
Related HCM Content 24~2
Computational Engine 24-2

2. CONCEPT5 ................................................................••...•...•...•...••..•...•...••...•...•...•. 24.3

LOS Concepts 24-3
LOS Criteria 24-3

3. CO R E M ETH O DO LOG 1E5 24.6

Scope of thc Mcthodologies 24-6

Requircd Data and Sources 24-8
Exclusive Off-Strcct Pcdestrian Facilities 24-10
Pedestrians on Shared -Use Paths 24~14
Off-Street Bicyele Fadlitics 24-17

4. EXTEN5ION5 TO THE METHODOLOGIE5 24.27

Pedestrian Plazas 24-27
PcdestrianZones 24-28
Walkways with Grades Above 5% 24-28
Paths Segregating Pcdcstrians and Bicyclists 24-29

5. APPLICA TIONS ........................................................................................•......... 24-30

Example Problems 24.30
Example Results 24-30
Typcs of Analysis 24-33
U se of Alterna ti ve T 001 s 24.34

6. REFERENCES ...•...•...•...•......••..•...•...•.......•...•...•...•...•.......•...••..••...•...•...•...•...•...•... 24-35

Chapter 24jOff-Street Pedestrian and Bicycle Facilities Contents

Version 6.0 Page 24-i
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

LIST OF EXHIBITS

Exhibít 24-1 Random-Flow lOS Criteria for Walkways 24-4

Exhibit 24.2 Platoon-Adjusted LOS Criteria for Walkways 24-4
Exhibit 24-3 lOS Criteria for Stairways 24-4
Exhibit 24-4 Pedestrian LOS Criteria for Sharcd-Use Paths 24-5
Exhibit 24-5 lOS Criteria for Bicycles on Sharcd-Use and Exclusive Paths 24-5
Exhibit 24-6 Required Input Data, Potential Data Sources, and Dcfault
Values for Off-Street Facility Analysis 24.9
Exhibit 24-7 Flowehart for Analysis of Exclusive Off-Street Pedestrian
Facilities 24-10
Exhibit 24-8 Width Adjustments for Fixed Objects and Linear Features 24.11
Exhibit 24-9 Typical Fixed-Objcct Effeetive Widths 24.12
Exhibit 24.10 Flowehart for Analysis of Pedestrian lOS on Shared-Usc
Paths 24.15
Exhibit 24-11 Flowehart for Analysis of BlOS on Off-Street Facilities 24.18
Exhibit 24.12 Sehematic of Active Passing Events 24-19
Exhibit 24.13 Sehematic of Meeting Events 24-21
Exhibit 24-14 Effective lanes by Path Width 24-21
Exhibit 24-15 Rcquired Bieycle Passing Distanee 24-22
Exhibit 24-16 Frequency of Blocking of Two lanes 24.23
Exhibit 24.17 Pedestrian Circulation Space in a Pcdcstrian Plaza 24.27
Exhibit 24.18 Effect of Vertical Climb 00 Horizontal Distanee Walked 24-28
Exhibit 24-19ll1ustrative Effect of Pedestrian Volume, Effective Path
Width, and Average Pedestriao Speed 00 Average Pcdestrian Spaee 24.30
Exhibit 24-20 ll1ustrative Effect of Pedestrian Volume and PHF on
Average Pedestrian Space 24.31
Exhibit 24.21 ll1ustrative Effect of Bicycle Volume, Average Pedestrian
Speed, and Average Bicyele Spced 00 Weighted Event Rate 24.31
Exhibit 24.22ll1ustrative Effect of Bieyele Volume, PHF, and Direetional
Distribution of Passing Bieyelists on Weighted Event Rate 24-32
Exhibit 24.23 Il1ustrative Effect of Path Volume, Path Width, and
Centerline Prescnce on BLOS Score 24-32
Exhibit 24.24 II1ustrative Effect of Path Volume and Bicyclist Perceotage
on BLOS Score 24-33

Contents Chapter 24jOff-Street Pedestrian and 8icyde Fadlities

Page 24-íi Vers.0n 6.0
 Highway Capacify Manual: A Guide for Mulfimodal MobiJjfy Analysis

1. INTRODUCTION

OVERVIEW
VOLUME 3: INTERRUPTED FLOW
Off-street pedestrian and bicycle facilities (a) are used only by nonmotorized
16. Urban 5treet Facilities
modes and (b) are not considered part of an urban street or transit facility. Thus, 17. Urban 5treet Reliability and
a shared-use path only 10 ft from a roadway but separated by a sound barrier ATOM
18. Urban 5treet Segments
may be eonsidered an off-street facility, but a sidepath scparated from the 19. Signalized lntersections
roadway by a 10-ft planted buffer would generally be considered an on-street 20. lWSC lntersections
21. AWSC lntersections
facility. Facilities located directly along an urban strcet (e.g., bicyclc (anes, 22. Roundabouts
sidewalks) are not addressed in this chapter. In general, the eharacteristics of 23. Ramp Terminals and Altemative
lntersections
motor vchicle traffic do not playa strong role in determining the quality of 24. Off-5treet Pe<!estrian and
service from the perspcctive of bicyclists and pedestrians on off-street facilities. Bicycle Fadlities

Off-street facilitieS are

Facilities located within approximately 35 ft of an urban street are generally
general/y !ocated more than 35
not considered off-street, allhough the precise definition of off-street varies by ft from a roadway, although
facility as described carlicr. These types of pedestrian and bicycle facilities are the exact distance may va!}' on
the basis of the local context.
covered in Chaptcr 16, Urban Street Facilities, and Chapter 18, Urban Street
Pedestrian and bicycle facilities
Segments. The definition also excludes crosswalks and queuing arcas; these areas along uroan streets are
are addressed in the intersection chapters (Chapters 19-23). Pedestrian addressed in Chapters 16 and
18.
components of transit facilities are addressed in the Trallsit Capacity alld Qua/ity of
Bicycle faafities along multifane
Servia Malll/al (1). The 35-ft threshold is based on studies of pedestrian and and two-fane highways are
addressed in Chapter 15.
bieycle facilities (2-4) in whieh it was found that motor vehicle traffic influenced
1ñeTransit Capadty and
pedestrian and bicycle quality of service on facilities located within at least this Quality of 5ervice Manual
distance of the roadway. COvef5 the ana/ysJ's of
pedestncJn facilities serwng
transit stops and stations.
CHAPTER ORGANlZATION
Chapter 24 provides eapacity and level-of-service (LOS) cstimation
procedures for the following types of facilities:
• Wa/kways: paved paths, ramps, and plazas that are generally located more
than 35 ft from ao urbao street, as weU as streets reserved for pedestrian
traffic 00 a full- or parl-time basis;
• Sfairways: staircases that are part of a longer pedestrian facilily;
• Shared-use pafhs: paths physically scparated from highway traffie for the
use of pedestrians, bicyclists, runners, inline skaters, and other users of
nonmotorized modes; and
• Exclusive ofj-street bicycle pafhs: paths physically separated from highway
traffie for the exclusive use of bicyeles.
Deseriptions and iIIustrations of cach of the aboye facility types are providcd
in Chapter 3, Modal Charaeteristics.
Chapter 24 is divided into five seetions. Section 2 defines LOS erHeria for off-
streel pedestrian and bicycle facilities. Section 3 provides the core methodologies
for evaluating the operation and quality of service of off-street pedestrian
walkways and stairways, bicycle paths, and sharcd-use paths. Section 4 provides
guidane!.' on extending these methods to pedestrian streets and plazas, walkways
with grades exeeeding 5%, and off-street facilities segregating pedestrians and

Chapter 24jOff-5treet Pedestrian and Bicycle Facilities Intr'oduction

VersiOn 6.0 Page 24-1
Highway Capacity Manual: A Guide tar Multimadal Mobilify Analysis

bieyelists. Section 5 presents guidance on using the results of an off-street fadlity

analysis, ineluding example results from the methods, information on the
sensitivity of resuIts to various inputs, and potential applieations.

RELATED HCM CONTENT

Other HCM content related to this chapter ineludes
• Chapter 3, Modal Characteristics, which introduces the pedestrian and
bicyele modes, provides examples of off-strcct pedestrian and bicyele
facility types, and discusses the variability of bieyele demand;
• Chapter 4, Traffie Operations and Capacity Concepts, whieh ineludes
sections on pedestrian and bieyele flow and eapadty;
• Chapter 35, Pedestrians and Bicyeles:Supplemental, which provides
example problems with step-by-step ealculations applying the off-street
pedestrian and bicyele facility methods; and
• Section O, Pedestrians, Bieyeles, and Public Transit, of the Planning alld
Preliminary Engineering Applications Cuide to the HCM, found in online
Volume 4, which provides guidance on incorporating this ehapter's
methods and performance measures into a planning effort.

COMPUTATIONAL ENGINE
The Federal Highway Administration research that developed the shared-
use path method (5) presented in this chapter also developed a spreadsheet-
based computational engine for the method. A modified version of the original
engine is available in the Chapter 24 seetion of the Technical Reference Library in
online Volume 4. The original research applied the peak hour factor (PHF) at a
different point in the calculation process than it is applied in the HCM methods.
The version of the engine posted in Volume 4 applies the PHF as described in
this chapter. Users should note that the engine's fixed segment length of 3 mi
cannot be ehanged by the user without significant modifications to the engine.

Introduction O1apter 24/Off-Street Pedestrian and Bícycle Facilities

Page 24-2 Vetron 6.0
 Highway Capacity Manual: A Guide for Multimoda/ Mobility Anaiysis

2. CONCEPTS

lOS CONCEPTS
Pedestrian and bicycle quality-of-service measurcs for off-street facilities
differ from those for on-street facilities. On-street quality of scrvice, as described
in the previous chapters in Volume 3, strongly reflects the effects of motorizcd
traffic on nonmotorized travelers' perceptions of comlort and safety. However,
by definition, motorized traffie effects are absent on off-street facilities. Instead,
quality of service lor off-street facilities reflects the interactions of facility users
with each other.

lOS CRITERIA
Three service measures are defined in this ehaptcr. The measure(s) to apply
to an analysis depend on the travel mode and type of off-street facility:
• For pedestrians on exclusive pedestrian facilities, the service measure is
pedestriall space, measured in square feet per pedestrian;
• For pedestrians on facilities shared by pedestrians and bicycles, the
service measure is the number of bicycle meeting and passing evcnts per hour;
and
• For bicycles on both shared.use and exclusive paths, the scrvicc measure
is a bicycle LOS (BLOS)score incorporating meetings per minute, active
passings per minute, presence of a centerline, path width, and dclayed
passings.
In the case of the pedestrian space measure, difierent LOS thresholds apply,
depending on the type ol facility under study and, in sorne situations, the nature
of the pedestrian flow along the facility. LOS thresholds for pedestrian facilities
in a transit station context, as given in the Transit Capacity and Quality o/ Service
Mallual (1), allow for higher levels of erowding for a given LOS than do the
thresholds foc off-street pedestrian facilities.
LOS does not refled whether a
LOS thresholds foc off-street pedestrian and bicycle facilities are bascd on
fao7ity complies with the ADA
available user perception research and in other cases on expert judgment. LOS or other stiJndards.
does not reflect whether a facility complies with the Americans with Disabilities
Act (ADA) or other standards. For the purposes of evaluating LOS, a
"pedestrian" is considered to be someone who is walking; therefore, pedestrian
LOS does not nceessarily reflect the quality of service experieneed by joggers,
persons in wheelchairs, or others who could also be considered pedestrians.

Walkways
The waIkway LOS tables apply to paved pedestrian paths, pedestrian zones
(e.g., exclusive pedestrian streets), walkways and ramps with up to a 5% grade,
and pedestrian walking zones through plaza areas. Exhibit 24-1 applies when
pedestrian flow along the facility is random. Exhibit 24-2 applies when platoons
of pedestrians form along the faeility, foc example, when a signalized crosswalk
is loeated at one end of the segment being analyzed, or when the walkway's
operation during special events is being analyzcd.

Chapter 24/Off-5treet PedestJian aOOBicycle Facilities Concept5

Verskvl6.0 Page 24-3
Highway Capacity Manual: A Guide lor Multimodal Mobility Analysis

Cross flows oecur at the interscetion of two approximately perpendicular

pedestrian streams (e.g., where two walkways interscct oe at a building
entranec). Beeause of the ineeeased number of confliets that occur between
pedestrians, walkway eapacity is lowee in a eross-flow situation than along other
parts of the walkway.ln eross-flow loeations, the LOS E-F threshold is 13 square
feet per peeson (fP/p), as indicated in the notes foe Exhibit 24-1 and Exhibit 24-2.

Exhibit 24-1
Random-FlowlOS Criteriafor
••;;,,---1------------
---AA~v;e,~.;¡g¡;el-=:--=~¡;R¡;e~I.;¡le;¡¡djM~e;:.;;,"u;;
Space FlowRate Average
Walkways (ftl/Dl (D/min/ftt S~d (n/s) v/cRatiob Comments
--.!"OS
Ability to move in desiredpath,
A >60 >4.25 :!:0.21
rlOneed to alter movements
OCcasionalneed to adjust path to
B >4()-ó() >5-7 >4.17--4.25 >0.21-0.31 avoid confliets
Frequentneedto adjust path to
e >24-40 >7-10 >4.00-4.17 >0.31-0.44 avoid conflicts

o >15-24 >10-15 >3.75-4.00 >0.44--0.65 =~~sa=::ss slower

>8--15< Speedrestricted,very limited

E >15-23 >2.50-3.75 >0.65-1.00 ability to passslower pedestrians

F Variable :!:2.50 Variable Speedsseverelyrestricted,

frequent contad with other users
Notes: Exhlbit 24-1 does IlOt apply to walkways wlth steep grades (>S%). See 5ection 4 for further dlscussioo. v/e
'" volume to capacity .
• Pedestrians per minute per too! of walkway wk:!th.
b vlcratio '" Ilow rate/23, based on random Ilow. LOS is based un aver(!ge sp.ace per pedestrian.
< In cross-I'\ow situatiolls, the LOS E-f threshok:! is 13 ft2/p.

Exhibit 24-2 Related

P1atoon-Adjusted
lOS Criteria Average Measure
for Walkways Space FlowRate'"
...!.Q.S__ (~z/pL _(p: min n b Comments
A >530 :!:0.5 Abllity to move in desiredpath, no needto alter movements
B >90-530 >0.5-3 OCcasiooa! need to adjust path to avoid conflicts
e >40-90 >3-6 Frequentneedto adjust path to avoid conflicts
D >23-40 >6-11 Speedand ability to passsIowerpedestriansrestricted
E > 11-23< > ll-1B Speedrestricted,very Iimitedability to passslower pedestrians
F :!:11< > lB s severel restricted,f uent contact with other users
Notes: • Rates in the table represent aver~e I'\ow rates over a S.min periodo Flow rate is directly related to space;
Ilowever, LOS is based on average space per pedestrian.
b Pedestrians per minute per foot of walkway width.

<In cross-flow situations, the LOS E--f threshold is 13 tt'{p.

Stairways
Exhibit 24-3 peovides the LOS criteria for stairways.

Exhibit 24-3
LOSCriteriafor Stairways Space
':--r-----------------
--AA~v~e~••~g~el_:JR~e~I~.~ted~~M~e;;.:;:'u;:
•••••
FlowRate
LOS (~p)_1 r;;¡minift)~~ Rati_~~c:~mments,-¡;;;;;;;;;;;;;;;;;;;;- _
A >20 :!:5 ~ 0.33 No need to alter movements
B > 17-20 >5-6 >0.33-0.41 Dccasionalneedto adjust path to avoid cooflicts
e > 12-17 >6-B >0.41-0.53 Frequentneed to adjust path to avoid conflícts
o >8--12 >8--11 >0.53-0.73 Umited ability to passslower pedestrians
E >5-8 > 11-15 >0.73-1.00 Very limited ability to passslower pedestrians
F <5 Varlable Variable Speedsseverelyrestricted,frequent contact with
- other users
Notes: • Pedestrians per minute per foot of wall<.way wIcfth.
b vlcratio '" f\ow rate{lS. LOS Is based on average space per pedestTian.

Concepts Chapter24{Off-StreetPedestrianand BicycleFacilities

Page24-4 V~6.0
 Highway Capacity Manual: A Gujde for Multimodal Mobility Ana/ysjs

Pedestrians on Shared-Use Paths

Exhibit 24-4 shows pedestrian LOS crHeria for paths shared between
pedestrians and bicycles.

Related Measure Exhibit 24-4

Event Bicycle Service Flow Rate Pedestrian lOS Criteria for
LOS Rate/h per Direction (bigocles.f.!1l Cornments Shared-Use Paths
A :S38 :S28 Optimum conditions, conflicts with bi<.ycles rare
B >38-60 >28--44 Good conditions, few conflicts with bicycles
e >60-103 >44-75 Dífficult to walk two abreast
D >103-144 >75-105 Frequentconflictswith <.yclists
E > 144--180 > 105-131 Conflicts with <.yclists frequent and disruptive
F > 180 > 131 5iqnificant user oonflicts, diminíshed experience
Notes: An 'event" is a t>icyclemeeting or pa~ng a pedestrian.
Blcycleservice flow rates (te" flow during the peak 15 min) are shown for refererx:e i1ndare based on iI
SOl50directional split of bk:y<:les;LOSis basedon number of events pe.-hour an<!apphesto any
directional split.

Exclusive and Shared Bicycle Facilities

Exhibit 24-5 providcs LOS criteria for bicycles on both shared-use and
exclusive off-street paths.

LOS BLOSSCore Cornrnents Exhibit 24-5

A >4.0 Optimum conditions, ample ability to absorb more riders LOS Criteria for 8icydes on
B >3.5-4.0 Good conditions, sorne ability to absorb fTIO(e riders Shared-Use and Exclusive
e >3.0-3.5 Meets current demand, marginal ability to absorb more riders Paths
o >2.5-3.0 Many oonflicts, sorne reduction in bicyde travel speed
E >2.0-2.5 Very crowded, with signiflcantly reduced bi<.ycle travel speed
F :S2,0 Sionificant user oonflicts and diminished eXOf>fience

Chapter 24jOff-Street Pedestrian and Bicycle Facilities Concept5

Version 6.0 Page 24-5
Highway Capacify Manual; A Guide tor Mulfimodal Mobility Analysis

3. eORE METHODOLOGIES

SCOPE OF THE METHODOLOGIES

Pedestrian and bicyde facilities Off-street pedestrian and bicycle facilities serve only nonmotorized traffic
along urban streets are
addressed in Chapters 16 aOO and are separated from motor vehicle traffic to the extent that such traffic does
18. not affect their quality of service. Thus, although sidewalks primarily serve only
Bkyde facilities along multifane pedestrians, they are not addressed in this chapter-the quality of service
and two-Iane highways are
addressed in Chapter 15. afforded to pedestrians on sidewalks depends in part on the presence and
characteristics of the adjacent motor vehicle traffic.
Procedures for estimating LOS are divided into three categories: pcdestrians
on exclusive pedestrian facilities, pedestrians on shared-use paths, and bicyclists
on both shared-use paths and exclusive bicycle facilities.
Much of the material in this section is the result of research sponsored by the
Federal Highway Administration (5-7). 60th commuter and recreational
bicyclists werc included in the off-street bicycle-path research (5).

Spatial and Temporal Limits

The analysis ot off-5treet The analysis oí off-street pedestrian and bicycle facilities occurs at the
pedestrian and bicyde faollties
occurs at tfIe segment leve!. segment level. A segment ends and a new segment begins when any of the
following occurs:
• There is a street crossing;
• The width of the facility changes significantly;
• There is ao intersection with another exclusive pedestrian or bicycle
facility where user volumes change significantly or cross f10ws are
created; or
• The type of facility changes (e.g., a walkway becomes a stairway).
In most cases, the minimum segment length will be around 0.25 mi, and the
maximum segment length will be 2 to 3 mi (5). Certain kinds of facilities, such as
The analysis period fength is stairways, cross-f1ow areas, and pedestrian plazas, will have shorter segment
15min.
lengths. The analysis period length is 15 min.

Performance Measures
These methods produce the following performance measures:
• Average pedestrian space on exclusive pedestrian facilities;
• Weighted event rate for pedestrians on shared-use facilities, where an
"event" is a bicycle meeting or passing a pedestrian;
• 6LOS score for bicyclists on off-street facilities that reflects bicyclists'
perceptions of the facilitis operational quality;
• LOS derived from the above three measures;
• Pedestrian volume-to-capacity ratio on exclusive pedestrian facilities; and
• Rates of bicyclists meeting, actively passing, and being delayed in passing
other off-street facility users.

Core Methodologies Chapter 24jOff-Street Pedestrian and Bicyde Fadlities

Page 24-6 Versbn 6.0
 Highway Capacity Manual: A Guide for Multimodal MobiJjty Analysis

lirnitations of the Melhodologies

Each facility type is treated from the point of view of pedestrians or
bicyclists. Procedures for assessing the impact of pedestrians and bicyclists on
other facility users (e.g., inlinc skaters) are not considered. Additional
information on other users may be found elsewhere (8). The methodologies do
not address LOS for pedestrians with disabilities such as vision or mobility
impairments. The reader is encouraged to consult material published by the
United Statcs Access Board to ensure compliance with the ADA.
These methodologies do not consider the continuity of walkways, bikcways,
and shared-use paths when determining LOS. Facilities that are interrupted with The methodoIogy does not
frequent roadway crossings will provide lower capacities and travel speeds than address!he impact of roadway
crossings on !he LOS af off-
facilities with long. unintcrrupted stretches. In addHion, roadway crossings, street paths.
especially crossings of high-volume oc high-speed facilities, may negativcly
affect the pcdcstrian and bieyele environment and user pcrceptions of quality of
scrvicc. However, the methodologies described here only consider discrete,
uninterrupted facilities for nonmotorized travel and do not assess the impact of
intersections with motorized vehicle facilities.
The methodologies are based solely on facility characteristics and do not
consider external factors such as weathcr, landscaping, adjacent land uses, and
lighting conditions, which may also affect users' perceptions of a facility.

Pedestrian Facilities
The capacity of pedestrian facilities is based on research conducted on
constrained facilities (e.g., bridges and underground passageways) wherc thcre
is no opportunity for pedestrians to walk outside thc designated area. Off-street
pedestrian facilities, in contrast, typically have no barricrs kccping pcdestrians to
the designated path. As a result, thcse facilities reach effective failure (Le., Where the opporrunity exists,
pedestrian spillover) at densities lower than capacity. Foc this reason, off-street pedestriilns wfif spifl aYer the edges
of a waikway at densities beIow
walkways are desirably designed to achieve LOS e or better (Le.,to achieve capacity. Off-stre!'t pathways are
uncrowdcd walking conditions within the designated path area), rather than desfrably designed to achfeve LOS e
or better Iv avoid thfs situation.
LOS E (i.c., capacity). The methodologies are generally appropriate regardless of
the type of surface used for the pedestrian facility.

Exclusive Bicycle Facilities

The methodology for exclusive bicycle facilities is based on research !he exclusive bicyde faolity
methodoIogy may not be
conducted only on paved surfaces and may not be applicable to soft surfaces appficable Iv faalities with 50ft
such as gravel, dirt, or wood chips. surfaces.

Shared-Use Paths
The methodology for shared-use paths does not account for the effect on The pedestrian sharecl-use
path methodoIogy does not
pedestrian LOS of path width oc the effects of meeting and passing events. No acr:ount for !he effeds af
credible data were found on fixed objects and their effects on users of these types nonbicydist users of the path
on pedestrian LOS.
of facilities. The methodology also does not account for the effect of nonbicyelist
users of the path (e.g., skateboarders, inline skaters) on pedestrians. However, it
is expected that pedestrians will often encounter thcse users on shared-use paths
and that because of their higher speeds, these users can have a negati ve eHect on
pcdestrian LOS.

Chapter 24/Off-$treet Pedestrian and Bicycle Facilities Core Methodologies

Version 6.0 Page 24-7
Highway Capacity Manual: A Guide for MultimodaJ MobWty Analysis

The methodology for BLOS on shared-use paths incorpora tes the effects of
five user groups: bicyc1ists, pedestrians, runners, inline skaters, and child
bicyclists. However, severa! user groups that may be a part of the mix on sorne
trails are not incorporated, including push scooter users, wheelchair users,
equestrians, cross-country skiers, and users of electric vehicles. The methodology
is based on research conducted only on paved surfaces and may not be
applicab!e to soft surfaces such as graveL dirt, or wood chips. The methodology
is not applicable for paths wider than 20 ft. This methodology was developed
fram data col1ected on two-way paths but may be applied to one-way paths by
setting opposing volumes egual to zero.
Sorne shared-use paths are signed or striped, or both, to segregate pedestrian
and bicycle traffic. The research that developed the shared-use path
methodology did not address those kinds of paths; guidance on such paths may
be found in Section 4.

REQUIRED DATA ANO SOURCES

Exhibit 24-61ists the information necessary to apply the methodologies and
suggests potential sources for obtaining these data. It also suggests default values
for use when specific information is not available (5). The user is cautioned that
every use of a default value instead of a field-measured value may make the
analysis results more approximate and Jess reJated to the specific conditions that
describe the facility. HCM dcfaults should only be used when (a) field data
cannot be collected, and (b) !ocally derived defaults do not existo
In particular, service measure resuIts are moderateIy sensitive to the choice
of PHF and highly sensitive to the choice of directional factor. Use of field-
measured peak 15-min volumes by direction avoids the need to apply these
fadors. In addition, the service measure for exclusive pedestrian facilities is
highly sensitive tú the average pedestrian specd input into the method.

Core Methodologies Chapter 24/Off-Street Pedestrian and Bicyde Facilities

Page 24-8 Versian 6.0
 Highway Capacity Manual: A Guide for Mu/tjmodal Mobility Analysjs

Mode and facility Iyoe Exhibit 24-6

Required Data Pedo Pedo Bike Potential Data Suggested Required Input Data, Potential
and Units Excl.l (Shared) (Aln Source(s) Default Value Oata Sources, and Default
Values for Off.5treet Facility
Facility width (ft) • • Field data, aerial photo Must be provided
Analysis
Effective facility width (ft) • Field data, aerial photo Same as facility width

PedestJian volume (p/h) • Field data Must be provided

Bicycle volume
lbiCvcleS/hl • Reld data Must be provided

Tota! path volume (p/h) • Field data Must be provided

Bicycle mode split 0.55 (i.e., 55% of
(decimal) • Field data
total Dath volume)
Pedestrian mode split
(decimal) • Field data 0.20

Runner mode split

(decimal) • Field data 0.10

~~line s~~ter mode split

decimal • Field data 0.10

Child bicyclist mode split

(decimal) • Reld data 0.05

Peak hour factor

(decimal) ~ • • • Field data 0,85

Oirectional volume
.!elit(decimal) b • • Field data 0.50

Average pedestrian
.!~(ft/min) • Field data 300 ftjmin

Average pedestrian speed

(mi/h) • • Fiefd data 3.4 mí/h

~~~an
mí/h
speed SO
• Field data 0.6 mí/h

~ve~~{e bicycle speed

mífh • • Field data 12.8 mi/h

Bicycle Speed SO (mífh) • Field data 3.4 mífh

Average runner speed
(mífh) • Field data 6.5 mífh

Runner speed 50 (mi/h) • Field data 1.2 mífh

Average inline skater
soeelÚmi1h) • Reld data 10.1 mífh

~~nn~lskater speed SO
mífh • Field data 2.7 mí/h

Average child bicyclist

speed (mi/h) • Field data 7.9 mí/h

~\~~~icydist speed SO
• Field data 1.9 mí/h

5egment length (mí) • Field data, aerial photo Must be provided

Walkway grade 5 5%
(vesornó)~ • Field data Must be provided

rredestrian flow ty:~

random or platooned (! • Field data Must be provided

~:terl:~~
es or no
stripe
• Field data Must be provided

Notes: Ped. '" peóestrian; E~d. '" e~dus~e; SO '" standard deViation.
Bold Italie indiciltes h~h sensitiYity (>20% chilnge) of servk:e measure te tlle chok:e of value.
Bold il'ldicates mOOerate sensitiYity (10% te 20% change) of service measure te tlle choice of value .
• Not required for pedestrian analysis when peak 15'min demand volumes are provided.
b Not required when directional demand voIume5 are provided.
< Pedestrian speeds reduce when grades e~ceed 5%; the service measure is highly sensitive to average
pedestriarl speed.
~lOS 1etter result is h~hly sensitive to lhe se1ection of peóestrian flow type.

Source: Default values from Hummer et al. (5), e~cept for effective facility width.

Chapter 24/Off-Street Pedesbian and Bicycle Facilities Core Methodologies

VersiOn 6,0 Page 24-9
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

EXCLUSIVE OFF-STREET PEDESTRIAN FACILITIES

OvelView
Exhibit 24-7 illustrates the steps required for the exclusive off-street
pedestrian facility methodology .

. Exhibit 24.7
FJowchart for Analysis of
5tep 1: Dd:errnine I:heeffective wlllkway width
ExclusiveOff-5treet
Pedestrian Fadlities

5tep 2: úllculate tne pedestrilln Row rate

Step 3: Úllculate the average pedestrilln space

Step 4: Determine LOS

Exhibit 24-1 (wlllkways without platooning)
Exhibit 24-2 (walkways with platooning)
Exhibit 24-3 (stZlirwllYS)

Step s: Úlrcullltel:he voIume-to-capacity ratio

Step 1: Determine Effective Walkway Width

Walkways and Cross-Flow Areas
Effedive walkway width is the portion of a walkway that can be used effectively
by pedestrians. Various types of obstructions and linear features, discussed
below, reduce the walkway area that pedestrians can effectively use. The effective
walkway width at a given point along the walkway is computed as follows:
Equation 24-1 WE = Wr - Wo
where
Wl effective walkway width (ft),
Wr total walkway width at a given point along walkway (ft), and
Wo sum of fixed-objed effective widths and linear-feature shy distances at
a given point along walkway (ft).
5hy di~nce ÍSa buffer
Exhibit 24-8 illustrates a portian of a sidewalk or walkway. The general
pedestrians leave between
themselves and linear objects concepts shown are applicable both to sidewalks along urban streets and to
along a walkway, such as exclusive off-street paths not located adjacent to a street. Linear features such as
euros and building faces.
the street curb, the low wall, and the building face each have an associated shy
distance, which is the buffer pedestrians give themselves to avoid accidentally
stepping off the curb, brushing against a building faee, or getting too clase to
ather pedestrians standing under awnings or window shopping. Fixed objects,
such as the tree, have effective widths associated with them. The jixed-objeet
effretive width indudes the object's physical width, any functionally unusable
space (e.g., the space between a parking meter and the curb or the space in front

Core MethodoIogíes Chapter 24¡Off-Street Pedestrian and BiCyc:1eFacilities

Page 24-10 Version 6.0
 Highway Capacity Manual; A Guide for Multimodal Mobility Analysis

of a bench occupied by people's legs and bclongings), and the buffer given the
object by pedestrians.

Street Exhibit 24-8

Width Adjustments for Fixed
Objects and linear Features

The concept of effedive width applies

Total w~lkw~y wldth, W,
1
Effective w~lkw~y width,WE
to both on-street and off-street
facilities.. Because of (he proXl/1JItyof
¡ 1 (he street in Exhibit 24-8, the
sidewalk here would be considered
an on-street pedestIian fadlity .
\
.-.--.. .. • '""---
Object llne (fence o' lo•• w~II) • Building f~(e i Building f~ce witl!
wlndo•• dl5play
Tñe street is show" so a/f fac::torsthiJt
can inffuence the effective wkfth of
walkways can be depicted in one
•• Shy di5tance • • FI~ed-(lbjecteffeetlve wldth pface .

Exhibit 24.8 also shows that the effective width of a fixed object (here, a tree)
extends over an effective length that is considerably longer than the object's
physicallength. The effective length represents the portion of the walkway that
is functionally unusable becausc pcdestrians need to move to one side ahead of
time to get around a fixed object. The effectivc length of a fixed object is assumed
to be five times the object's effective width.
Typically, a walkway operational analysis cvaluates the portion of the
walkway with the narrowest effective width because this section forms the
constraint on pedestrian flow. A design analysis identifies the minimum effective
width that must be maintained along the length of the walkway to avoid
pedestrian qucuing or spillover.
Exhibit 24-9 gives the effective widths of a variety of typical fixed objects
found along on- and off-street pedestrian facilities. The values in Exhibit 24-9 can
be used when spccific walkway configurations are not available.

Stairways
Pedestrians tend lo walk in
A stairway's capacity is largely affccted by its width. Unlike walking on a
fines or Janes on stairways;
level surface, traversing stairs tcnds to make people walk in lines or lanes. The lhus, meaningful inoei15eS in
width of a stairway determines both the numbcr of distinct lines people can form capacity are re/ated to the
number of pedestTian Janes
on the stair and the side-to-side spacing bctween them, which affect both the avaiJable.
ability of faster pedestrians to pass slower-moving pcdestrians and the level of
interferenee between adjacent lines of people. Conscquently, meaningful
increases in capacity are not linearly proportional to the width but occur in
increments of about 30 in. (1).
Smalf reverse ffows on
On stairways (in contrast to walkways), a minor pedestrian flow in the stairways should be assumed
opposing direction can result in redueed eapacity disproportionate to the ro use one pedestr"ian lime (30
magnitude of thc reverse flow. As a result, a small reverse flow should be in.) of width.

assumed to oecupy one pedestrian lane, or 30 in. of the stair's width. For a
stairway with an effective width of 60 in. (5 ft), a small reverse flow eould
consume half its capacity (1). The allowance for small reverse flows, when used,
is included as part of the Wo term in Equation 24-1.

Chapter 24jOff-Street Pedestrian and Bic.ycle Facilities Core Metltodologies

Version 6.0 Page 24-11
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Exhlbit 24-9 FlxedObject

.. . ._.'-ff-ect
... lv-e.Wj~th(ftL
Typical Fixed-Object Effective Street Fumiture
Widttls L1ghtpole 2.5-3.5
Traffic signal poIes and boxes 3.0-4.0
Fire alarm boxes 2.5-3.5
Are hydrants 2.5-3.0
Traffic signs 2.0-2.5
Parking meters 2.0
see Exhlbit 24-8 for shy Mail boxes (1.7 ft x 1.7 ft) 3.2-3.7
diStilnces associated with ruros Telephone booths (2.7 ft x 2.7 ft) 4.0
and building faces. Trash cans (1.8 ft diameter) 3.0
Benches 5.0
Bus shelters (on sidewalk) 6.0-7.0
Pub/k Underground Aa::'ess
Subway stalrs 5.5-7.0
Subway ventilation gratings (raised) 6.0+
Transfonner v~lt ventilation gratings (raised) 6.0+
l..Qndsca.E!!!9.., _
T,.., 3.0-4.0
Planter boxes 5.0
Commerciaf Uses
Newsstands 4.0-13.0
Vending stands Variable
AdvertiSing and store displays Variable
Sidewalk cafés (two rows of tables) 7.0
Building Protrosions
Columns 2.5-3.0
5too", 2.0-6.0
(ellar doors 5.0-7.0
Stand pipe connections 1.0
Awning poles 2.5
Truck docks (tnJcks protruding) Variable
Garage entrancejexit Variable
Driveways Variable
SOurce: Pushkarev and Zupan (9).

Step 2: Calculate Pedestrian Flow Rate

Walkways and Cross-Flow Areas
Hourly pedestrian demand is used as an input to the analysis. Consistent
with the general analysis procedures used throughout the HCM, hourly dcmand
is usually converted into peak 15-min f1ows, so that LOS is based on the busiest
15 consecutive minutes during an hour:
v,
Equation 24-2 v -
15 - 4 x PHF
where

VI5 pedestrian flow rate during peak 15 min (plh),

Vh pcdcstrian demand during analysis hour (plh), and

PHF peak hour factor.

If peak 15"11Jinpeáestrian
flowever, if peak 15-min pedestrian volumes are available, the highest 15-
voIumes are available, the
highest 15-min voIume can be min volume can be used directly without the application of a PHF.
used directfy in the method
without the appfieatian of a Next, the peak 15.min f10w is con verted into a unit flow rate (pedestrians per
PHF. minute per foot of effective path width):

Core Mettlodologies O"Iapter 24/Off-Street Pedestrian and Bicyde Fadlities

Page 24'12 Ver.s,OO6.0
 Highway Capacity Manual: A Guide for Mulfjmodal Mobility Analysis

Vl5
V =---- Equation 24-3
p lSxWf"
where vp is pedestrian flow per unit width (p/ít/min), and all other variables are
as previously defincd.

Stairways
Becausc pedestrians usc morc cncrgy to ascend stairs than to dcscend them, CritiCal pedestrian fJows on stairs
OCClJrin the up dlrection.
lower flow rates typically occur in the ascending direction. For this rcason, when
stairs serve both directions simultaneously or when the same stairway will be
used primarily in the up direction during somc time periods and primarily in the
down dircction duríng othcr time periods, the upward flow rate should be uscd
for analysis and dcsign (1). The calculation of pedestrian flow rate for stairways
is otherwise the same as that described for walkways and cross-flow areas.

Step 3: Calculate Average Pedestrian Space

The scrvicc measure for walkways is pedestriall space, the inverse of densíty. 1
Space = Density
Pedestrían spacc can be directly obscrved in thc fieId by measuring a sample
area of the fadlity and determining the maximum number of pcdestrians at a
given time in that area. Pedestrian space is rclated to pedestrian speed and unit
flo\\! ratc:
Sp
A =- Equation 24-4
p v
p

where
Ap pedestrian spacc (ft2/p),
Sp pedestrian spccd (ft/min), and
vp pedestrian flow per unit width (p/ft/min).

Step 4: Determine LOS

Walkways with Random Pedestrian Aow
Where pedestrian flo\\!on the path is not influenccd by platooning (sce next
subsection), Exhibit 24-1 should be used to determine pedestrian LOS.
Ramps with griJdes of S% (K
Research (9-11) has shown that pedestrian speeds on ramps with grades up less can be treated as
to 5% are not significan tI)' different from speeds on level walkways but that walkways for the purpose of
determining LOS.
spceds decrease at higher grades. Therefore, thc walkway LOS values are also
applicable to ramps with grades of 5% or less. Ramps with stccpcr grades are
díscussed in $ection 4. The walkway LOS values can also be adapted to
pedestrian plazas and pcdestrian zones (exclusive pedestrian strcets), as
discusscd in Section 4.

Wa/kways with Platoon F/ow

Platooning on wa/kwa)l.S".
lt is important for the analyst to dctermine whether platooning a1tcrs the
underlying assumptions oí random flow in the LOS calculation. Platoons can
arise, for example, if entf)' to a walkway segmcnt is controlled by a traffic signal
at a street crossing or if pedcstríans arrive at intervals on transit vehicles.

Chapter 24/Off-Street Pedestrian and Bicycle Facilities Core Metl1odologies

Version 6.0 Page 24-13
Highway Capacity Manual: A Guide for Multimodal MobiJity Analysis

Where platooning oecurs, the pedestrian f10wis concentrated over short time
periods rather than being distributed evenly throughout the peak 1S-min
analysis periodo The available space for the typical pedestrian under these
circumstanees is much more constrained than the average space available with
random arrival would indicate. There is no striet definition for what
differentiates platooning from random f1ow;observations of local conditions and
engineering judgment should be used to determine the most relevant design
eriteria (Le., platoons versus random flow).
If platooning oecurs during.the analysis period, Exhibit 24-2 should be used
to determine LOS. Researeh (9) indicates that impeded f10wstarts at S30 ft2/p,
which is equivalent to a f10wrate of 0.5 p/min/ft. This value is used as the LOS
A-B threshold. The same research shows that jammed f10win platoons starts at
11 fWp, which is equivalent to 18 p/min/ft. This value is used as the LOS E-F
threshold.

Cross-Flow Areas
ws
CrGlS5-tk:lw threshokfs are
A cross f10wis a pedestrian f10wthat is approximately perpendicular to and
kienticiJl to rhose for wafkways,
except for the LOS E-F erosses another pedestrian stream, for example, at the interscetion of two
I1>ce<hoId.
walkways or at a building entrance. In general, the lesser of the two f10ws is
referred to as the eross-f1ow eondition. The same proeedure used to estimate
walkway spacc is used to analyze pedestrian facilities with eross f1ows.As
shown in the notes to Exhibit 24-1 and Exhibit 24-2, the LOS E threshold (Le.,
eapadty) in eross-f1ow situations oecurs at a lower density (higher average space)
than that for walkways without cross f10ws (12).

Stairways
Researeh (13) has developed LOS thresholds based on the Institute of
Transportation Engineers' stainvay standards, whieh provide the space and flow
values given in Exhibit 24-3. As with walkways, stairway LOS is deseribed by the
service measure of pedestrian spacc, expressed as square feet per pedestrian.

Step s: Calculate Volume-to-capacity Ratio

The volume-to-eapadty (v/e) ratio can be eomputed by using the following
values of capadty for various exclusive pedestrian facilities:
• Walkways with random f1ow:23 p/min/ft;
• Walkways with platoon f10w(average over 5 min): 18 p/min/ft;
• Cross-f1ow areas: 17 p/min/ft (sum of both f1ows);and
• Stairways: 15 p/min/ft in the aseending direction.

PEDESTRIANS ON SHARED-USE PATHS

LOS for pedestrians on shared-use off-street paths is based on the number of
events during whieh a pedestrian either meets an oncoming bicyclist or is passed
by a bicyclist. As the number of events inereases, the pedestrian LOS deereases
because of redueed eomfort. Exhibit 24-10 shows the steps taken to determine
pedestrian LOS on shared-use paths.

(ore Me1:hodologies Chapter 24/Off-Street Pedestrian and Bicyde Fadlitie5

Page 24.14 Verron 6.0
 Highway Capacify Manual: A Guide for Mulfimodal Mobility Analysis

Exhibit 24-10
Step 1: Gather Input data
Flowchart for Analysis of
Hourly or peak 1S-min pedestrian and bicyde dem¡mds by dired:ion Pedestrian LOS on Shared.Use
Average bicyde and pedestri&n speeds Paths

1
Step 2: Úllculate the number of bicyde passing and meeting events I
1
Step 3: Detennlne LOS
Exhibit 24-4

Shared-use paths typically are open to users of nonmotorized modes such as

bicyclists, skateboarders, and wheelchair users. They are often constructed to
serve areas without city streets and to provide recreational opportunities for the
publicoThese paths are also common on university campuses, where motor
vehicle traffic and parking are often restricted. In the United States, there are few
paths exclusively for pedestrians-most off-street paths are for shared use.
Bicyeles-because of their markedly higher spceds-have a negative effect
on pedestrian capacity and LOS on shared-usc paths. However, it is difficult to
establish a bicycle-pedestrian equivalent because the relationship between the
two differs depcnding on their respective f10wsand diredional splits, among
other factors. This section covers pcdestrian LOS on shared-use paths. Bicyelists
have a different perspective, as discussed in the following section.

Step 1: Gather Input Data

The following input data are required for the analysis:
• Hourly or peak 15-min pedestrian and bicycle demands by direction, and
• Average pedestrian and bicycle speeds.

Step 2: Calculate Number of Bicycle Passing and Meeting Events

LOS is based on !he overtaking
LOS foc shared-use paths is based on hindrance. Research (14) has of pedestrians by bicydists.
established LOS thresholds for pedestrians based on the frequency of passing (in Passing occurs in the same
the same diredion) and of meeting (in the opposite direction) other users. diredion; meeting occurs trom
the opposite direction.
Because pcdestrians scldom overtake other pedestrians, pedestrian LOS on a Pedestrian-to-pedestrian
interaction is typically
shared-use path depcnds on the frequency with which the average pcdcstrian is negligible.
met and overtaken by bicyclists (14). However, the analyst should observe
pedestrian behavior in the field before assuming that pedestrian-to-pedestrian
interaction is negligible. The shared.use path methodology docs not account foc
events with users other than bicyclists (e.g., inline skaters).
The average numbers oí passing and meeting events per hour are calculatcd
by Equation 24-5 and Equation 24-6, respectively. These equations do not
account for the range of bicycle speeds encountered in practice; however,
because of the Iimitcd dcgrce of overlap between the speed distributions of
bicyelists and pedcstrians, thc resulting difference is practically insignificant.

Chapter 24/Off'Street Pedestrian .lOO Bicycle Facilities Core Methodologies

Version 6.0 Page 24-15
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

For one-way paths, there are no meeting events, so only Fp, the number of
passing events, needs to be calculated. Paths 15 ft or more in width may effectively
operate as two adjacent one-way facilities, in which case F m may be set to zero.

Equation 24-5 F = Q,b

p PHF
(1 _ Sp)
Sb

Equation 24-6 F
m
= Qob
PHF
(1 + Sp)
Sb
where
Fp number of passing events (events/h),
Fm number of meeting events (events/h),

Q<b bicycle demand in same direction (bicycles/h),

Qdo bicyclc demand in opposing direction (bieycles/h),
PHF peak hour factor,
Sp mean pedestrian speed on path (mi/h), and
Sb mean bicycle spccd on path (mi/h).
If peak 15-min volumes by direction are known, they should be substituted
for the Q,JPHF and Q.JPHF terms in the aboye equations. If only two-directional
volumes are known, a directional distribution factor can be applied to the two-
directional volume to estimate thc directional volumes. (However, as mentioned
previously, thc LOS resuIts are highly sensitive to the choice of directional factor,
and field measurement of the directional distribution is recommended when
possible.)
Meebng events create Iess Meeting events allow direct visual contact, so opposing-direction bicydes
hindrance tllan oW!ftaking
events. tend to cause less hindrance to pedestrians. To account for the reduced
hindrance, a factor of 0.5 is applied to the meeting events on the basis of theory
(14). When sufficient data are available on the relative effects of meetings and
passings on hindrance, this factor can be calibrated to local conditions.
Because the number of events calculated in the previous step was based on
hourIy demand, a PHF must be applied to convert them to the equivalent
demand based on peak 15-min conditions. The total number of events is
Equation 24-7
F = (Fp + O.5Fm l
where F is the total number of events on the path in events per hour, and the
other variables are as defined previously.

Step 3: Determine lOS

Exhibit 24-4 is used to determine shared-use path pedestrian LOS bascd on
the total events pcr hour calculated in Step 2. Unlike the case for exclusive
pedestrian facilities, the LOS E-F threshold does not reflect the capacity of a
shared-use path but rather a point at which the number of bicyde meeting and
passing events results in a severely diminishcd experience for the pedestrians
sharing the path.

Core Methodologies Chapler 24jOff-Street Pedestrian and Bicycle Fad!ities

Page 24-16 Version 6.0
 Highway Capacify Manual: A Guide tor Multimodal Mobility Analysis

OFF.STREET BICYCLE FACILmES

The off.stree( bieyele facillty
On shared-use paths, the presence of other bicydists and other path users
analysfs is rosee! on several
can be detrimental to bicyclists by increasing bicycle delay, dccreasing bicyde fact0r5 thilt affect USfY
capacity, and redudng bicyclists' freedom oi movement. Research (5) corrclating perceptions.
user pcrceptions of cornfort and enjoyrnent of path facilities with an objective
measure of path and user charactcristics serves as the basis for thc LOS
thresholds and methodology described in this section. Thc following key crHeria
are considered through this methodology:
• The abiJity of a bicydist to maintain an optimum speed,
• Thc number of times bicydists meet or pass other path users, and
• The bicydist's freedom to maneuver.
The rcsults of a perception survey were used to fit a linear regression model
in which the survey results served as thc dependent variable. The methodology
incorporates the effects of five path modes that may affect BLOS:other bicydists,
pedestrians, runncrs, inline skaters, and child bicydists. Five variablcs-
mt."etingsper minute, active passings per minute, path width, presence of a
centerline, and delayed passings-arc used in the model. In the special caS('of an On exdusive off-stree( bfcyde
facilities the number of
exdusive off.strcet bicyele facility, the volume for all nonbicyde modes is
passings and meetings fs
assumed to be zero, and the numbcr of passings and meetings is determined determined soIe/y by the
solely by the volume of bicydes. bfcyde vofume.

BLOSon exclusive and shared-use off-street bicyde facilities is based on uscr

perceptions of how the LOS of shared-use paths changes according to several
factors. These factors are combined into a single BLOSscore. LOS thresholds
relate to a specific range of LOS score values. Exhibit 24-11shows the steps taken
to determine the LOS of off-street bicyde facilities.
The following sections describe the steps for calculating BLOSfor an off-
street facility.

Step 1: Gather Input Data

The methodology addresses five types of path users, or mode grOl/ps:
bicydists, pedestrians, runners, inline skaters, and child bicydists. The following
input data are required for each mode group:
• Hourly or peak 15-min demand by direction in modal users per hour,
• Average mode group speed in miles per hour, and
• Proportion of aH path users reprcscnted by a particular mode group (i.e.,
rnode split).
In addition, the following data are required for the facility:
• Path width in fcct, and
• Presence oi a centerline stripc (yes or no).

Qlapter 24¡Off-Street Pedestrian and Bicycle Facilities Core Methodologies

VersiOn 6.0 Page 24-17
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Exhibit 24.11
F10wchart for Analysis of BLOS Step 1: Gather Input data
on Off-Street Fadlities For Mch mad~lI'USN (pM~ns, biC''lc!L~, in/int>s1alt~rs. run~ ltndchl1d bicyc/m):
Hourfy or peak 15-min demands by direction, spee¿s, and mode split
Forth~ fiKi/ity:
Path widt:h, ~nterline prese!nct

Step 2: ÚIlculate active passings per minute

Step 3: Ú1\a.;late meeting s per minute

Step 4: Determine the number of effective lanes

Step 5: Calculate the probability of delllyed passing

Step 6: Úllculate delayed passings per minute

Step 7: Determine alOS

Exhibit 24.5

Step 8: Adjust lOS for Iow.volume paths

If peak 15-min directional volumes are known for each user group, the
analysis can proceed directly to Step 2. Otherwise, the hourly directional flow
rate on the path is calculated for each of the five modes on the basis of the hourly
directional demand for the path and the path mode split

Equation 24.8
QT xPi
q¡ = PHF
where
q; hourly directional path flow rate for user group ¡(modal userslh),
Qr total hourly directional path demand (modal users/h),
Pi path mode split for user group ¡(decimal), and
PHF peak hour factor.
If only two-directional total path volumes are known, a directional
distribution factor can be applied to the two-directional volume to estimate the
dire<:tionalvolumes prior to entering them in Equation 24-8. (As mentioned
above, LOS results are highly sensitive to the choice of directional factor, and
field measurement of the directional distribution is rerommcnded when
possible.)

Core Methodo1ogies O1apter 24/Off-Street Pedestrian and Bieyele Facilities

Page 24-18 VetSiOO 6.0
 Highway Capacity Manual: A Guide for Multimodal MobUity Analysis

Step 2: Calculate Active Passings per Minute

Active passings are defined as the number oE other path users traveling in
the same direction as an average bicyclist (Le.,a bicyclist traveling at the average
speed oE aHbicycles) who are passed by that bicyclist. The average bicyclist is
assumed to move at a constant speed U. The value oE U should be set to the
average speed oE bicyclists on the Eadlity in question. The methodology Eor
determining active passings incorporates separately the effects oE each oEthe Eive
mode groups described in Step 1. The speeds oE path users oE each mode group
are assumed to be normalIy distributed with a mean JI., and standard deviation 0;,
where i represents mode.
The average bicyclist passes only those users who (a) are present on the path
segment when the average bicyclist enters and (b) exit the segment after the
average bicyclist does. Thus, for a given modal user in the path when the average
bicyclist enters, the probability of being passed is expresscd by

P(v,) = plv, < U(l-j)J Equation 24-9

where
P(v,) probability of passing user of mode i,
U speed of average bicyclist (mi/h),
v, speed of a given path user of mode i (mi/h),
L length of path segment (mi), and
x = distance from average bicyclist to user (mi).
Exhibit 24-12 provides a schematic of active passing events.

Exhlbit 24-12
5chematic of Active Passing
Events

Source: Adilpt:ed from HlImmer et al. (5).

Because Vi is distributed normally, the probability in Equation 24-9 can be

calculated from the integral under the standard normal curve. By dividing the
full length of the path L into 11 small discrete pieces each of length dx, the average
probability of passing within each piece j can be estimated as the average of the
probabilities at the start and end oE each piece:
P(v,) = O.5[F(x - dx) + F(x)] Equation 24-10

where F(x) is the cumulative probability of a normal distribution of speeds with

mean f.J and standard deviation O; and the other variables are as defined
previously.
The expected number of times that the average bicyclist passes users of mode
i over the entire path segment is determined by multiplying P(v,) by the density
of users of mode i and summing over aH portions of the segment. The number of
passings per minute is then obtained by dividing the result by the number of
minutes required for the bicyclist to traverse the path segment:

O1apter 24¡Off-Street Pedestrian and Bicycle Facilities Core Methodologies

Vel'SlOn 6.0 Page 24-19
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

" q, 1
Equation 24-11 A- =
f L
}=1
P(v.) x - x -dx.
1". t J
,...1
where
A, expected passings per minute of mode j by average bicyclist,
q, directional hourly flow rate of mode j (modal users/h),
Ji, average speed of mode j (mi/h),
path segment travel time for average bicyclist (min), and
dx¡ length of discrete segment j (mi).
The other variables are as previously defined.
Research (5) has found that setting dx equal to 0.01 mi is appropriate for the
purposes of the cakulations shown in Equation 24-11 and below.
Equation 24-11 provides expected active passings by the average bicyclist for
mode i. To determine total active passings of aH modes, Equation 24-11 must be
repeated for each individual mode and then summed:
Equation 24.12 A ¿A,
T = i

where AT is the expected active passings per minute by the average bicyclist
during the peak 15 min, and the other variables are as defined previously.

Step 3: calculate Meetings per Minute

Meetings are dcfined as the numbcr of path users traveling in the opposing
direction to the average bicyclist that the average bicyclist passes on the path
segmento AHusers present on the path when the average bicyclist enters will be
passed by the average bicyclist, assuming no user enters or exits the path at an
intermediate point:

Equation 24-13 M -!!... '\ q,

1-60¿Jli
,
whcre MI is the meetings per minute of users already on the path segment, and
the other variables are as prcviously defined.
In addition to users already on the path segment, users who have yet to enter
the segment will meet the average bicyclist within the segmento The probability
of this occurrence is

Equation 24-14

where
probability of meeting opposing user of mode j,

v, speed of path user of made j (mi/h),

x distance of user beyond end of path segment (mi), and
U speed of average bicyelist (mi/h).

Core Methodologies Chapter 24fOff-5treet Pedestrian and Bicycle Facilities

Page 24.20 Vemm6.0
 Highway Capacity Manual: A Guide for Multimodal MobiJjty Analysis

Because vo.; is distributed normally, the probability in Equation 24-14 can be

readily calculated from the area under the standard normal curve. The length of
path beyond the analysis segment that may supply users who will be met by the
average bicyclist is defined as X •• By dividing x. into tl small discrete pieces, each
of length dx, the average probability of meeting a modal user from each piece can
be estimated by applying Equation 24-10,substituting X for x. Although sorne
meetings will occur with very fast path users located greater than L distance
beyond the end of the segment when the average bicyclist enters, setting x. equal
to L is sufficient to guarantee that al least 99% of meetings will be captured (5).
Exhibit 24-13 provides a schematic of meeting events.

L
x. fxhibit 24-13
5chematic of Meeting Events

~u l +4J
¡
X dx
source: Adapte!! from Hummer et al. (5),

Similar lo the process Ear calculating number of active passings (Equation 24-
11), the estimation of number of meetings with users from a particular mode
group nol on the path segment when the average bicydist enters is
"
M2i=
. L P(VOi)xqix~dxl.
~
. t Ili
fquation 24-15

)=:1

where M¡, is the expected meetings per minute of users of mode j located beyond
the end of the path segment at the time the average bicyde enters the segment,
and the other variables are as previously defined.
Finally, the total number oE expected meetings per minute during the peak
15 min Mr is determined by adding M¡ to the sum of M2,j across all mode groups:

Equation 24-16

AlI variables are as previously defined.

In the special case of a one-way path, there are no opposing users to meet;
therefore, Mr is zero.

Step 4: Determine Number of Effective Lanes

The effective numhcr oE lanes on a shared.use path affects the number of
delayed passings: as the number oí lanes increases, delayed passings decrease.
Even paths without painted Jane markings will operate with a de facto number
of lanes. The relationship between path width and the number oí effective
operationallanes is shown in Exhibit 24-14.

_Path _Width (ft) __ E_ff~ective

L.a!!~ Exhibit 24-14
8.0-10.5 2 Effective lanes by Path Width
llJ}-14.5 3
15,0-20.0 4

source: Hummer et al. (5).

Chapter 24/Off.5treet Pedestrian and Bicycle Facilities Core Methodologies

L__
Version 6.0 Page 24-21
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Step 5: Calculate Probability of Delayed Passing

Delayed passing maneuvers occur when there is a path user ahead of the
overtaking average bicyelist in the subject direction and another path user in the
opposing direction, such that the average bicyelist eannot immediately make the
passing maneuver. The probability of a delayed passing depends on the passing
distance required, which in turo depends on both the overtaking mode and the
mode of the user being passed. The passing distances bicyelists require to pass
other user modes are shown in Exhibit 24.15.

Exhibit 24-15 OvertaJdngMode ModePass~(t~uired PassiEgD~~~ ..-'!!L

Required Bicycle Passing Bicyele Bicydist 100
Distance Bicycle Pedestrian 60
Bicyde Inline skater 100
Bicyde Runner 70
BieycJe Child bicyclist 70
Source: Hummer et al. (5).

With the values in Exhibit 24.15, the probability that a given passing section
will be vacant oí a given mode for at least the required passing distance Pi can be
estimated by using a Poisson distribution. The probability of observing al least
one modal user in the passing section is the complement of the probability of
observing a vacant sectian. The probability p".; oí observing a bloeked passing
seetion for mode iis
Equation 24-17

where
P",i <!= probability of passing seetion's being blocked by mode i,
Pi distance required to pass mode j (mi), and
k¡ density of users of mode j (users/mi) <!= q//l,'
Equation 24-17 is applieable to both the subject and opposing directions.

Two-Lane Paths
On a two-lane path, delayed passing occurs when, within the distance
required to complete a pass p, the average bicyclist eneounters one of the
following: traffie in both direetions, blocking a single ¡ane in eaeh direetion; ar no
traffie in the subject direetion in eonjunetion with traffie in the opposing direction
that is being overtaken by an opposing bicyclist. Note that these situations are
rnutually exclusive. The delayed passing probabilities in the subject and
opposing directions are
Equation 24-18 Pas = PnoPns + Pno(1- Pns)(1- Pao)
Equation 24-19 Pao = PnoPns + Pns(1- Pno)(1 - Pas)
where
Plb probability of delayed passing in subject direction,
Pdo probability oí delayed passing in opposing direction,
P"" = probability oí blocked lane in opposing direction, and
p., probability oí blocked lane in subject direction.

Core MethodoIogies Olapter 24fOff-5treet Pedestrianal'ld Bicycle Facilities

Page 24-22 Vet3iafl6.0
 Highway Capacity Manual: A Gu;de tor Mulfjmodal MobWty Analys;s

&llving Equation 24.18 and Eguation 24-19 for Pd• results in

PnoPns + Pno(1 - Pns)2
PdS = ----------- Equation 24-20
1 - PnoPns(1 - Pno)(1 - Pns)
Bccause P"" and Pn, are calculated from Eguation 24.17, Equation 24-20 can be
readily solved for Pd,-

T17ree-Lane Paths
Because a greater variety of possible scenarios may occur, the operations of
three-Iane paths are more complicated than those of two-Iane paths. The
methodology includes severallimiting assumptions regarding user behavior:
• Bicyclists in the subject direction use only the two rightmost lanes,
• Bicyclists in the opposing direction use only the two leftmost lanes,
• Passing maneuvers occur only in the middle lane and never in the left
lane, and
• Groups of users may sometimes block the two lanes allocated to that
direction but cannat block aH three lanes.
As a result, a delayed passing occurs in two cases: (a) traffic in the subject
direction blocks the rightmost lane in conjuncHon with oppasing traffie
occupying the other two lanes, or (b) side-by.side users block the t\\'o rightmost
lanes in the subject direction. The probabilities of the occurrence of a delaycd
passing in the subject and opposing directions are given by
Pds = Pns[Pbo + Pno(1 - Pdo)] + Pbs Equation 24-21

Pdo = Pno[Pbs + Pns(1 - Pds)] + Pbo Equation 24-22

where P""is the probability of two blocked lanes in the opposing direction, P", is
the probability of two blacked lanes in the subject direction, and all other
variables are as previously defined.
Equation 24.21 and Equation 24-22 are simultaneous equations with two
unknowns, P J, and PJo' Defining D as P J, - P Jo gives the following equation:
D = [(PbS - PbO) + (PnsPbo - PnoPbs)]/(1 - PnsPno) Equation 24-23

Substituting Equation 24-23 into Equation 24-21 results in

PdS = [Pns(Pbo + Pno(1 + D)) + Pbs]/(1 + PnsPno) Equation 24.24

This model requires determining four probability parameters: specifieally, Pn

and Pb in eaeh direction. Calculating these parameters requires estimating the
fracHon of aHevents in which both lanes are blocked. These parameters were
established through research (5) in which video data of more than 4,000 path
users on U.S. shared-use paths ,vere observed. Exhibit 24.16 shows the blocking
frcquencies by mode.

Mode Freg,ueng of Bloddn91~)_ Exhibit 24.16

Bicycle 5 Frequency of Blocking of Two
Pedestrian
Inline skater
36
8 "'"'"
Runner 12
Child bicyclist I

Source: Hummer et al. (S).

Chapter 24{Off-Street Pedestrian and Bícycle Facilities Core Methodologies

Version 6.0 Page 24-23
Highway Capadty Manual: A Guide tor Multimodal Mobility Analysis

Therefore, P¡",,¡andPM,iI the probabilities that a user Di mode i will block two
lanes in the opposing and subject directions, respectively, are found by
multiplying the irequency oi blocking twa lanes by a particular user oi mode i
(Exhibit 24.16) by the probabiJity that a user oi made i will be encountered,
which is given by Equation 24.17, This process results in
Equation 24.25 PbS,1 = F¡ x P 7IS,í
Equation 24.26 PbO,1 = F¡ x P 7IO,i

where F¡is the irequency with wruch mode i will block two lanes (from Exhibit
24.16), and aH other variables are as previously defined. The probability that a
user oi any mode will block two lanes is thus given by

Equation 24-27 PbS = L, Pbs,í

Pbo = L, pbO,¡

The probabilities that only a single lane will be blocked by a user of a given
mode i, Pqo,; and Pqo,;, are thus derived from the probability that at least one ¡ane
will be blocked (from Equation 24.17) minus the probability that two lanes will
be blocked (from Equation 24.25 and Equation 24.26). These probabilities are

Equation 24.29 P7IS,1. -- 1- eP¡ks.1 - Pk"S,l '

Equation 24-30 Pno, ,= 1- eP¡ko.í - Pk,,0,1.

where k.,¡ and ko,j are the densities oi users of mode i in users per mBe in the
subject and opposing directions, respectively, and aH other variables are as
previously defined.
The probabilities that a user oi any mode will block a single lane are thus
given by

Pns = L, Pns,i

Equation 24.32 Pno = L, Pno,í

The values of PM and PI><>irom Equation 24.27 and Equatian 24.28 and the
values oi p •• and P"" from Equation 24.31 and Equation 24.32 can now be
substituted into Equation 24-23 and Equatian 24.24 to determine the probability
of delayed passing, PJ.. Bccause this delayed passing factor was calibrated by
using peak hour volumes rather than peak 15-min volumes, a PHF is applied to
convert AT from peak 15.min flaw rate conditions to hourly conditions.

Four-Lane Paths
On iour.lane paths, the methadalogy assumes the path operates similarly to
a dividcd four.lane highway, such that the probability oi delayed passing is
independcnt oi opposing users, as no passing occurs in the leftmost lanes. Thus,
the probability oi delayed passing Pdo is equivalent to the probability that both
subject lanes will be blockcd (PI•.>. which can be found by using Equation 24.25
and Equatian 24.27.

Core Methoclologies Chapter 24jOff-Street Pedestrian and Bicyde Facilities

Page 24-24 Version 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Step 6: Calculate Delayed Passings per Minute

The probability of delayed passing Pd; described aboye applies only to a
single pair of modal path users (e.g., a bicydist passing a pedestrian and
opposed by a mnner). The total probability of delayed passing Pu, must be
calculated from all modal pairs. Because there are five modes, five times five (25)
modal pairs require calculation. The total probability of delayed passing is found
by using

PTdS = 1- n{l- Pm•ds) Equation 24-33

m
where PTd, is the total probability of delayed passing, and P m.d, is the probability
of delayed passing for mode pair m. The operator n in Equation 24-33 indicates
the product of a series of variables.
Finally, delayed passings per minute DP m are simply the active passings per
minute Ar multiplied by the total probability of delayed passing PTd;:
DPm = AT X PTds X PHF Equation 24-34

Because the DPm factor was calibrated from peak hour volumes rather than
peak 15-min volumes, a PHF is applied to convert Ar from peak lS-min flow rate
conditions to hourly conditions.

Step 7: Determine SlOS

The BLoo seore (Equation 24-35) uses inputs from Steps 2, 3, and 6 plus
facility data gathered in Step 1. The cquation was dcvcloped from a regression
model of user responses to video dips depicting a variety of off-street bicycle
fadlities (5). The LOS e-o threshold represents the midpoint of the response
scale used in thc survey.
BLaS = 5.446 - 0.00809£ - 15.86RW - 0.287CL - DP Equation 24-35

where
E weighted events per minute = meetings per minute +
10)( (actiw passings per minute);
RW reciprocal of path width = l/path width (ft);
eL 1 if trail has centerline, O if no centerline; and
DP min [DP"," 1.5/(180/60),1.5] = min lDP"," 0.5, 1.5].
The delayed passings factor DP is calibrated (a) to fal! within the range of
delayed passings (1-180 delayed passings per hour) observed during the
research that developed this factor and (b) to produce a maximum change of
three letters in the LOS result (5).
With the exception of the spedal cases discussed in Step 8, the bieydist
perception index is used directly with Exhibit 24-5 to determine bieydist LOS on
off-street facilities. As with shared pedestrian facilities, the LOS E-F thrcshold
does not reflect the capadty of an off-street bicyde facility but rather a point at
which the number of meeting and passing events results in a severely
diminished experience far bicydists using the path.

Olapter 24/Off-Street Pedestrian and Bícycle Facilities Core Methodologies

Version 6.0 Page 24-25
Highway Capacity Manual: A Guide for Multimodal Mobility Ana/ysis

5tep 8: Adjust LOS for Low-Volume Paths

It is not possible to achieve LOS A or B for narrow (e.g., 8-ft) paths by using
Equation 24.35. Because paths with very low volumes would be expected to
result in a high perceived quality of service, the following adjustments are made
to the LOS results:
• All paths with five or fewer weighted events per minute are assigned
LOSA.
• All paths with more than five to 10 weighted events per minute are
assigned LOS B, unless Equation 24-35 would result in LOS A.

Core Methodologies Olapter 24/0ff-Street Pedestrian and Bicyde Fadlities

Page 24-26 Version 6.0
 Highway Capacity Manual: A Guide tar Multimodal Mobility Analysis

4. EXTENSIONS TO THE METHODOLOGIES

PEDESTRIAN PlAZAS
Pedestrian plazas are large, paved areas that serve multiple functions,
including pedestrian circulation, special events, and scating. The circulation
fundion is of interest here, although the design of a plaza must consider how aH
the functions interact. For example, queues from areas designated for food
vendors may intrude into a pedestrian drculation route, reducing the mute's
effective width, or two circulation mutes may intersect each other, creating a
cross-flow area. In addition, the circulation and amenity functions oí a plaza
sometimes conflict, as people tend to linger longer in plazas that do not act as
thoroughfares (9).
The exclusive pedestrian walkway methodology can be used to analyze
pedestrian drculation routes through pedestrian plazas. The methodology does
not address the need or desire to have space for amenities within a pedestrian
plaza. The effective width of such a mute is not as easily identified as that of a
walkway, because the edges of the circulation area are often undefined.
However, pedcstrians will tend to take the shortest available mute across the
plaza, as iIlustratcd in Exhibit 24-17.

Exhibit 24-17
Pedestrian Círculation Space
in a Pedestrian Plaza

The effedive width of a circulation route is influenced by the widths of the

entrance and exit points to the plaza and by the prcsence of obstacles (e.g., walls,
poles, signs, benches). Effedive width may also be influenced by whether a
change in texture or color is used to mark the transition betwcen circulation and
amenity space. Between 30% and 60% of pedestrians will use plaza space that is
flush with a sidewalk, with the higher percentages applying to wider plazas and
those that help cut a comer and the lower percentages applying to narrower
plazas and those with obstacles (9).
For design applications, peak pedestrian demands through the plaza would
need to be estimated. Given this information and a design LOS, a minimum
effective width could be determined for each circulation route. Multiplying the
width of the route by the length of the mute and summing for aHroutes results
in thc space required for pedestrian circulation. Space requirements for seating
arcas and other plaza fundions are added to the circulation space to determine
the total plaza space required for pedestrian circulation and amcnities.

Otap1er 24{Off-Street Pedestrian and Bicyde Facilities Extensions 10 the Methodologies

Version 6.0 Page 24-27
Highway Capacity Manual: A Guide for Multimodal MobilUy Analysis

For operational applications, ao average effective width can be determined

through field observation of the space occupied by pedestrians on a drculation
route during peak times. Dividing an average per minute pedestrian volume by
the effective width gives the pedestrian f10w rate far the circulation route, from
which LOS can be delermined.

PEDESTRIAN ZONES
PedestTian zones are streets Pedeslrian zones are streets dedicated to exclusive pedestrian use on a full-
dedicated lo exclusive
pedestJian use on a fu/{- or ar part-time basis. These zones can be analyzed fram an operational standpoint
part-time basis. by using the exclusive pedestrian walkway methodology, as long as the kinds of
obstructions listed in Exhibit 24-9, suro as sidewalk café tables, are taken into
account. Altemative performance measures may be considered that assess the
street's attractiveness to pedestrians, because a successful pedestrian zone is
expected to be relatively crowded (i.e., to have a lower LOS). Although an
uncrowded zone would have a high LOS, it could be perceived by pedestrians as
being a potential personal security risk because of the lack of other users.
The HCM methodoIogy is not
The HCM methodology is not suitable for pedestrian zones during times
suitable for pedestrian zones
duríng times when defivery when delivery vehicles are allowed to use the street. The HCM methodology is
vehicles are affowed lo use the
also not applicable to the analysis of a low-speed street (e.g., a Dutch-style
<lreet.
woonerft shared by pedestrians, bicycles, and automobiles.

WALKWAYS WITH GRADES ABOVE SOlo

Ccnsult the IiJtest version of
the ADA Accessibility
Research (9-11) has shown no appreciable impact on pedestrian speed for
Guklelines for guidance on the grades up to 5%. As shown in Exhibit 24-18, aboye a 5% grade, walking speeds
maximum sJope afJowed on an
iJCCl!SSlbferoute.
drop as grade increases, with travel on a 12% grade being about 30% slower than
travel on a level surface. Grade may not have an appredable impact on capacity,
however, because the reduction in pedeslrian speed is offset by closer pedestrian
spadng (9). The stairway LOS table (Exhibit 24-3) would provide a conservative
estimate of pedestrian LOS on steeper walkways.

Exhibit 24-18
Effect of Vertical Oimb on
Horizontal Distance Walked

••
-
100 UG 2GCl
V.rtlulll ••.•••••• Qlmbecl (ft)

SOurce: Municipal Plannlng Association (11).

Extensions to the Methodologies Olapter 24/Off-Street Pedestrian and Bicyde Facilities

Page 24-28 Version 6.0
 Highway Capacify Manual: A Guide for Multimodal Mobilify Analysis

PATHS SEGREGATING PEOESTRIANS ANO BICYCUSTS

Sorne paths are signed or striped, or both, to segregate bicyclists from
pedestrians. When field observation on the path (or similar paths in the same
region) indicates path users generalIy comply with the regulations, up to aH the
bicycle-pedestrian passing events could be con verted to meeting events in Whete suffk:ient physical
segregation of bicydiStS and
proportion to the path users' compliance rate; this would result in an improved pedestrians OCCVIS,it mar be
LOS. Where sufficient physical segregation oí bicyclists and pedestrians occurs, it ro
aPfXO{JfÍate treat the path
as tIfl7 separate f8dfities.
may be appropriate to treat the path as two separate facilities.

Olapter 24/Off-Street Pedestrian and 6icyde Facilities Extenslons to tne MethodoIogies

V~6.0 Page 24-29
Highway Capacity Manual: A Guide for Multimodal Mobilify Analysis

5. APPLICATION5

EXAMPLE PROBLEMS
$ection 2 of Chapter 35, Pedestrians and Bicycles:Supplemental, provides
two example problems that ilIustrate the application of the off-street pedestrian
and bicycle facility methods:
1. Comparison of pedestrian LOS on shared-use and exclusive paths, and
2. BLOSon a shared-use path.

EXAMPLE RESULTS
This scction presents the results of applying this chapter's methods in typical
situations. Analysts can use the illustrative results presented in this section to
observe the sensitivity of output performance measures to various inputs, as well
as to help evaluate whether their analysis results are reasonable, The exhibits in
this seetion are not intended to substitute for an actual analysis and are
deliberately provided in a format large enough to depict general trends in the
results-but not large enough to pull out specific results.

Exclusive Off-Street Pedestrian Facilities

Exhibit 24-19 presents i1lustrative results showing how average pedestrian
space relates to the eombination of two-directional pedestrian demand volume
and (a) effective path width or (b) average pedestrian speed. It can be seen that
average pedestrian space increases with both increasing effective path width and
increasing average pedestrian speed.

Exhibit 24.19
IIIustrative Effect of
Pedestrian Volume, Effective
Path Width, and Average
Pedestrian Speed on Average
Pedestrian Space

o-+-- --.-
........•
---1 ~
o I,OlXI 1,(lOO J,OlXI ••0lXI S,OlXI 6.OlXI 7,OlXI LOlXI l,OlXI 3,OlXI ••0lXI S,OlXI 6,OlXI 7.0lXI
~"_~_ll>/lIl T~_n_(p/hl
I_~ •..•
~...",.-11. -, •• 1 I _>.l •••••• -- •••••••••••.••••••_ .•••••••,

(a) Effective Path wldth (b) Average Pedestrian Speed

Note: Calculated uslllg this chapter's methods, using PHF = 0,85. In Exhibit 24-19(a), average pedestrian speed
•• 3.4 mi/h. In Exhlbit 2+19(b), effective path widttl = 10 ft

Exhibit 24.20 presents illustrative results demonstrating how the ealculated

average pedestrian space during the peak 15 min (altematively, the maximum
hourly pedestrian volume that achieves a particular pedestrian space) relates to
the choice of PHF. lt can be secn that average pedestrian space is moderately
sensitive to the choice of PHF.

Applications Chapter 24/Off-street Pedestrian ancl Bicyde Facilities

Page 24-30 Version6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

Exhibit 24-20
lIIustratil/e Effect of
Pedestrian Volume and PHF
on Average Pedestrian Space

1.000 1.000 3.000 4.000 ~.OOO6.000 7.000

~_ol""'moionVOlu,,", (p¡hl
I-.~••
-.••
.,.-
..•
"'._-~
••~
••
I
Note: Cakulated l)Sjng tt1is dlapter"s mettlods, using an average pe«>strian speed of 3.'1 ml/h and an effective
path width of 10 ft.

Pedestrians on Shared-Use Paths

Exhibit 24-21 prcsents iIIustrative results showing how the weighted event
rate (Le., weighted number of bicycle-pedestrian passings and meetings per
hour) relates to the combinabon of two-directional bicycle demand volume and
(a) average pedestrian speed or (b) average bicycle speed. lt can be seen that the
weighted event rate is relatively insensitive to path user spL'Cdand that even
relatively low bicycle volumes produce diminished pedestrian guaUty of service
(e.g., LOS D or worse, equivalent to 103events per hour or more).

-~ Exhibit 24-21
i'~
¡,~
IIIustrative Effect of Bicycle
- ,~ Volume, AI/erage Pedestrian
Speed, and AI/erage Bieycle
! 120
Speed on Weighted EI/ent
"00
!~ R,te
! ~
~ 010
J ~
• o so 100 ISO 200 2~0 lOO o so 100 ISO 200 2!>O lOO
Two-Oft<tI<>NoIai<y<loo
Voh.om.(bI<y<loo/hj T--oweet\oftol ai<y<loo
VoIumo(l>i<rdo's/hl
I '.••_--._-_ .••.••...•
---._1 1- •.••.•.••
_10 ••••••• _."._- ••••••••• 1
(a) AI/erage Pedestrian Speed (b) Average Bicycle Speed

Note: Calculated using this chapter"s methods, using PHF = 0.8S and a SalSa tlicyde directional ~Iit. In Exhibit
24.21(a), the aVff(}Qebicyde speed is 12.8 mi/h. In Exhibit 24-21(bl, the average pedestrian speed is M
mi/h.

Exhibit 24.22 presents illustrative results demonstrating how the weighted

event rate varies with (a) PHF and (b) the directional distribution of passing
bieyclists. The weighted event rate is moderately sensitive to PHF and highly
sensitive to the diredional distribution. A directional imbalance with more
passing bicyclists has a greater impact than an imbalance with more meeting
bicyclists; note that Equation 24-7 weights a meeting event as equivalent to half a
passing evento The directional distribution can be rc1evant to shared-use paths
used as bicycle commuter routes, as well as recreationa! trails where bicyclists
travel out and back via the same route, and thus a greater proportion of bicyclists
travel in a given direction at a given time of day.

Chapter 24jOff-Street Pedestrian and Bieyele Facilities Applications

Version 6.0 Page 24-31
Highway Capacity Manual: A Guide for Multimodol Mobility Analysis

Exhibit 24-22
IIIustrative Effect of B~
¡ -:
,~-
VoIume, PHF, and Directlonal
Distributlon of Passing - ,~
Bicydists on Weighted EYeflt ! 120 ,

Rate :r:
1:
1»
o ~+
o 50 100 150 :lOO 250 lOO
r-DINclIonIl ticyde Vokome(blcvdnlhl
I-,,---~._.-I
(a) PHF (b) Distribution of Passing Bicyclists

Note: Calculated usIng this cnapter's methods, using <In average pedestrian speed of 3.4 mifh and an average
bicyde speed of 12.8 mi/h. In Exhibit 2+22(a), the directional distribution is 50/SO. In Exhibit 24-22(b),
the PHF is 0.85.

Off.5treet Bicycle Facilities

Exhibit 24-23 presents illustrative results showing how the BLOS score
relates to the combination of two-directional path volume and (a) path width or
(b) eenterline width.

It can be seen in Exhibit 24-23(a) that the BLOS score improves substantially
as path width increases, which allows bicyclists to maneuver more freety. On the
two-Iane (Le., 8- and 10-£1)paths, the effect of the LOS adjustment for low-
volume paths (Step 8 of the methodology) can be seen as stair steps in the curves,
as the LOS is automatieally set to LOS A or LOS B under low-volume eonditions.
On the three- and four-lane (Le., 12- and 15-£1)paths, an inflecnon point can be
seen in the curve at higher path volumes. This inflection point is an effeet of the
delayed passings (DP) variable in Eguation 24-35, which eaps the delayed
passing rate at 1.5 delayed passings per minute. Once thís point ís reaehed, the
BLOS seore declines mueh more slowly.
Exhibít 24-23(b) shows that the BLOS seore 00 paths with centerlioes is
always 0.28710wer thao 00 paths without centerlines, corresponding to the
coefficient for this factor in Eguation 24-35, except under low-volume eonditions,
when the Step 8 LOS adjustment applies.

Exhibit 24-23
IIIustrative Effect of Path
Volume, Path Wtdth, and
centerline Presence on BLOS
:r
ªj'. +-1'-- ------
5<0,.
2 """ -~
o~ , t- o~
o :lOO '*00 600 !lOO UlOO l,:lOO o 200 '*00 600 !lOO 1.000 l,:lOO
_1 "thVOlume(pl"1 T~I ,.thVOlume{p/h1
1_ ••.•••__ .12. ".1 1 ••• _.-... 1

(a) Path Width (b) Centerline Presence

Note: Calculated using this chilpter's methods, uslng the default mocle splits and modalllSef speeds given In
Exhlblt 24-6, a PHF of 0.85, and a 50/50 patfl use!" directional distribution. In Exhlblt 24.23(a), no
centerline is pres.ent. In Exhibit 24-23(b), the path width is 10 tt

ApplicatiOrlS OJapter 24jOff-street Pedestrlan and Bicycle Facilities

Page 24-32 Vef3'iOn 6.0
 Highway Capacity Manual: A Guide for Multimodal Mobility Ana/ysis

Exhibit 24-24 presents iIIustrative results demonstrating how the BLOSseore

varíes with the percentage of path users who are bicyclists. It can be seen that the
BLOS seore varies widely for a given path volume depen~ing on the pereentage
of bieyclists (altematively, the maximum path volume that achieves a given LOS
varies greatly depending on the percentage of bicyclists). The relative
pcrcentages of other path users also affect the BLOSseore, with the slowest
modal users (e.g., pedestrians, child bicyclists) having the biggest effect.
Howevcr, accuratcly estimating the bicyclist percentage is more important, in
terms of the impaet on the final result, than accurately estimating the relati\'('
proportions of thc othcr path uscr types.

, Exhibit 24-24
IIIustrative Effect of Path
Volume and Bicyclist
~.-~ Percentage on BLOS 5core

1:~~:.:.:------
•

o
o 200 400 liIXI 800 I.lXlQ 1.200
Two-Oi_n.1 P.lh V"'"""' (PI")
__ ._"""' ••. ""." ..•• " ,"'-- .•••• 1

Note: CaIOJlated using this chapter's methods with the default modal user speeds given in Exhibit 24,6, a PHF of
0.85, and a 50/50 path user directional distriootion. The mode splits for runners, inline skatefS, and ehikl
bicyelists are as given in Exhibit 24,6, with pedestrians making up the balance of the path users.

1YPES OF ANAL YSIS

Operational Analysis
A eommon application of operational analysis is to compute the LOS of a
facility under exístíng or future demando The effective width of the facility ís an
input to the calculatíon, and LOS is an output.

Design Analysis
Designing for an effective
Design applications require that a LOS goal be established, with the primary
width.
output being the facility design eharaeteristics requircd or the maximum user
volumcs allowable for the LOS goal. For instance, a design analysis for a
pedestrian walkway may estimate the minimum effective width Wr needed to
achieve a design LOS value. In this ease, the maximum pedestrian unit flow rate
for the desired service level would be determined from Exhibit 24-1 or Exhibit
24-2. The effective wídth would be computed by solving the pedestrian unit
flow-rate equation backward. To avoíd pedestrian spillover (i.e., where
pedestrians walk outside the path to pass other users), it is desirable to design a
vvalkway to aehieve LOS e or better (Le., a maximum of 10 p/min/ft). Stairways
are desirably designed to achieve LOS e or D.
Oetermining service voIumes.
Similarly, the achievable path flow rate QT ean be solved as the primary
output. For exclusive bicycle facilities, the minimum LOS perception sean' for the
design LOS would be determined from Exhibit 24-5. By holding all but one path-
user group's demand eonstant and solving the events equation backward (e.g.,

Chapter 24jOff-Street Pedestrian and Bicycle Facilities Applications

Version 6.0 Page 24-33
Highway Capacity Manual: A Guide for Multimodal Mobility Analysis

13. Virkler, M. Quality of Flow Along Pedestrian Arterials. Presented at 18th

Annual Australian Road Research Board Transport Research Conference-
Transit New Zealand Land Transport Symposium, Christchurch, New
Zealand, Sept. 1996.
14. Botma, H. Method to Determine Level oí Service for Bieycle Paths and
Pedestrian-Bieycle Paths. In Transportafioll Research Record 1502,
Transportation Research Board, National Research Council, Washington,
D.C., 1995, pp. 38-44.

References O1apter 24{Off-Street Pedestrian and Bicycle Facilities

Page 24-36 Ve¡sioo 6.0
 Highway Capacify Manual: A Guide tor Multimodal Mobilify Analysis

VOLUME 3 INDEX

The index to Volume 3 Iists the text citations oí the terms defined in the
Glossary (Volumc 1, Chapter 9). Volumes 1, 2, and 3 are separately indexed. In
the index listings, the first number in each hyphenated pair oí numbcrs indicates
the chapter, and the number after the hyphen indicates the page \vithin the
chapter.

AltemJtive tool, 16-32, 17-43, 17-44, 18-SO lo

A 18-82, 19-94, 19.%, 20-1, 20-7 1020-9, 20-37,
Acn'lcrJtion I,me, 18-71 20-38,21-1, 20-10, 22-3, 22-10 to 22-13,
Aceess point, 16-30. 18-6, 18-13, 18-1510 22-24, 22-26, 23-2, 23-11, 23-12, 23-18, 23-19,
18-19,18-21,18-22,18-25 to 18-28, 18-30 to 23-28,23-50,23-61, 23-6.1lo 23-65, 23-70,
18-33,18-38,18-59,18-60,18-64,18-82, 23-79,23-81,23-82,23-92
19-33,23-35 Analysis hour, 24-12
Acn'Ssibility, 16-10, 18-10, 22-30, 24-28 Analysis period, 16-1, 16-3, 16-5, 16-6, 16-10
Aeruracy, 16-4, 16-9, 18-4, 18-9, 18-14, 18-25, to 16-13, 17-.1,17-7, 17-11, 17-12, 17-14,
18-43,18-59,18-67,18-80,19-14.19-22, 17-17,17-25 to 17-27, 17-29, 17-38, 17-39,
19-73, 19-87, 19-94,23-29,23-72,23-85 17-41,17-4.1,17-44,18-1, 18-4, 18-5, 18-lOto
Active access point, 18-16, 18-18, 18.19 18.12, 18-14 10 18-17,18-20,18-21, 18-25,
Active passings, 24-3. 24-17, 24-]9 to 24-21, 18-27,18-33, 18-44, 18-45, 18-59, 18-68,
24-25 18-69,19.2,194,19-14,19-18 to 19-20,
Active traffie and deffiand mana¡;ement 19-22 to 19-32, 19-34, 19-39, 19-4910 19-53,
(Am~I), 16-2, 17-210 17-6, 17-34, 17-36, 19-56,19-65,19-67 to 19-69, 19-73, 19-74,
17-41, 17-42, 18-2 19-87.20-7,20-10,20-12,20-20,20-30, 20-32,
Actuated contro!. 18-34, 19-4 to 19-6, 19-34, 20-37,21-10,21-11, 21-18, 21-19, 22-.1,22.10,
19-38,19-41, 19-5O,19-56, 19-76, 19-~, 22-14, 22-16, 22-23, 24-6, 24-14
23-29,23-63 Analytical model, 20-30, 22-2, 22-3
Adjuslmenl factor, 17-12, 17-13, 17-15, 17-18, Annual aVl'rage dai1y traffie (AADn. 16-6,
17-19,18-2,18-27 tu 18-30, 18-32, 18-35, 17-12,17-15,17-16,23-93
18-52,18-63,19-2,19-12,19-23,19-2910 Approach dday, 19-24, 19-32
19-31, 19-451019-49, 19-51, 19-84, 19-89, Approoeh grade, 19-23, 19-34, 19-46, 20-10
19-90,20-18,20-19,20-22, 20-24, 20-25, Arca typc, 17-23, 18-69, 19-24, 19-40, 1945,
20-27, 21-13, 21-16, 21-21, 22-13, 22-16, 1947,2.1-29,23-.11
22-2\, 22-22, 13-27, 23-29, 23-3110 23-33, Arriva1 rale, 16-11, 18-11, 19-13. 19-19, 19-62,
23-37,23-38, 23-44. 23-78. 23-82, 23-85 19-63,19-64,19-74,21-5,21-6
Algorilhm, 17-35 Arriva1 type, 16-29, 19-28, 19-29, 23-29, 23-58,
AIl-way STOP-controlled (AWSC), 16-17, 23-76,23-82
16-22, 16-29, 16-30, 18-23, 18-42, 18-43, Arrival-depaTturc po1ygon, ]t1-37
18-57, 18-58, Chapler 21, 22-9 Automobile, 16-1, 16-9, 16.10, 16-16, 16-3.1,
A1temalívedataset, 17-9, 17.12, 17-15, 17-18, 18-1,18-9,18-10, 11:1-13, 18-39, 18-71, 18-83,
17-221017-25,17-37 19-17,19-18,19-98,204,23-1,24-28
A11emative inlersection, 18-35, 23-1 to 23-6, Average bicyelist, 24-19 lo 24-22
23-9,23-17,23-65,23-70 to 23-73, 23-87 Average grade, 19-34
Average running sjX't'd, 16-7. 17-9, 18-6.
18-27,18-62
Average Irave1spt-'t'd, 23-13

Volume 3/1nterrupted Aow Page V3-1 Volume 3 Index

V~6.0
Highway Capacity Manual: A Guide tor Multimodal Mobility Analysis

Delay, 16-2, 16-7, 16-11, 16-12, 16-16, ]6-20, Diamond interchange, 20-10, 23-2, 23-5, 23-6,
16-24,16-27,17-2,17-5,17-9,17-27 to 17-34, 23-10,23-11, 23-13, 23-18 to 23-20, 23-22,
18-2,18-6,18-7,18-11 to 18-13,18-15,18-19 23-33,23-42,23-47, 23-SO,23-56, ~1,
to 18-23,18-27,18-30 to 18-32, 18-34, 18-35, 23-63, 23-64
18-38,18-44,18-47,18-51,18-54 to 18-56, Díl"'-'Ctionaldislribulion, 17-15, 17-38, 24-16,
18-59,18-61,18-62,18-70 lo 18-75, 18-81, 24-18,24.31 to 24-33
19-1,19-2,19-7,19-8,19-11,19-15,19-16, Dirt.'Ctional flow fatI', 24-18
19-19,19-20,19-23,19-25,19-29,19-30, Dírectional split, 24-5, 24-15, 24-31
19-32, 19-41, 19-491019-57,19-62,19-64 lo Dísplacl-d. Idt tum interscctíon (DLn, 23-5,
19-70,19-79,19-83 to 19-86, 19-89, 19-90, ~,23-9 to23-11, 23-17, 23-65, 23-68,
19-94 to 19-99, 20-6, 20-7, 20-9, 20-21, 20-30 23-69,23-71, 23-73, 23-81 to 23-86, 23-89 to
to 20-33, 20-37, 20-39 lo 20-44, 20-4710 23-92, 23-94
20-49, 21-9, 21-18, 21-19, 21-22 to 21-24, Distributl'd ínteTSt.'Clion,23-1, 23-2, 23-7, 23-9,
21-26,21-27,22-2,. 22-9, 22.22, 22-23, 22-29, 23-10 to 23-13
23.1,23-2, 23-6, 23-9, 23-12, 23-13, 23-16 to Diverge, 23-66, 23-67, 23-75
23-18,23-27,23-28,23-31, 23-45, 23-50, Diverging díamond ínterchange (001), 23-2,
23-55,23-56,23-60,23-62,23-65, 23-70, 23-5,~, 23-8 to 23-11, 23-16, 23-19, 23-20,
23-71,23-78,23-79,23-81, 23-83, 23-86, 23-22, 23-26, 23-32, 23-33, 23-36 to 23-39,
23-88,23-89,23-92 to 23-94, 24-17 23-42 to 23-47, 23-SOto 23-54, 23-5610
Oclaye<! crossing, 20-41, 2{}-43 23-<1
Oclaye<! passing maneuver, 24-22 Divide<! mlodian typc, 18-60
Ocmand adjustmcnl factor (DAF), 17-15, Driver population, 20-4
17-23,17-26 Dual entry, 19.23, 19-37
Demand flow rate, 18-}4, 18-15, 18-16, 18-17, Dwell time, 17-11, 18-67, 16-68, 18-73
18-30,18-32,18-35,18-52,18-63,18-64, Dynamic speed limil, 17-5
18-82, 19-22 to 19-24, 19-26, 19-30, 19-50,
19-52, 19-54, 19-59, 19-65, 19-67, 19-73,
19-74, 19-tll, 19-84, 19-87, 19-91, 20-7, 20-10, E
20-12, 21-10, 21-11, 22-14 to 22-16, 22-22, Effective grl'l'n time, 18-32, 18-36, 18-73, 19-9,
23-32, 23-34, 23-55, 23-71, 23-73, 23-80 19-11,19-13,19-27,19-38,19-49,19-51,
Demand multiplíer, 23-91, 23-92 19.52,19-64,19-89,23-39,23-44,23-53 to
Demand starvation, 18-12, 19-20, 20-7, 21-10, 23-56, 23-59
23-5,23-10,23-11, 23-30, 23-38, 23-39, 23-47 Effective red time, 19-9, 19-11, 19-64, 19-74,
to 23.SO,23-55, ~1 23-54
Demand volume, 16-3, 16-7, 16-11, 17-1, 17-7, Effective walk time, 19-78, 19-79, 19-81
17-15,17-16,17-37,18-6,18-11,18-25,19-15, Efk'Ctive walkway width, 24-10
19-19,19-26,19-59,20-10,20-12,20-46, 85th percentile speed, 19-73, 19-74, 19-85
21-11, 21-15, 22-14 lo 22-16, 22-28, 23-7, Empírical model, 23-35, 23-62
23-11, 23-29, 23-64, 23-69, 23-73, 23-83, Entry fiow, 22-2, 22-3, 22-8, 22-17, 22-18, 22-29
23-84, 23-92, 24-9, 24-30, 24-31 Excess wait time, 18-67, 18-68, 18-76, 18-77
Ocnsity, 16-6, 18-15, 18-16, 18-19, 18-27 to Exclusive Off-stfl'l't bicycle path, 24-1
18-29, 18-82, 22-3, 23-2,23-43, 23-44, 24-7, Exclusive tum lane, 18-14, 18-17, 19-7, 19-21,
24-13,24-14,24-19,24-22,24-24 19-22,19-24,19-33,19-40,19-43
Ocpartull' headway, 21-3 lo 21-7, 21-9, 21.11, Exit f!ow, 18-24
21-14, 21-17 to 21-19, 21-26 Expcrienced travel time, 23-2, 23-12, 23-13,
[A'Sígn analysis, 16-31, 18-79, 18.80, 19-29, 23-26, 23-SO,23-70, 23-80, 23-84
19-93,20-46,21-25, 22-28, ~1, 24-11, Extensíonofeffectivegrl'~n, 19-11, 19-12,
24-33,24-34 19-14,23-39
Desígn hour, ~1 Extra distance travel time, 23-12, 23-13, 23-15
[A'Sígn spt-'l'd, 23-26, 23-46, 23-56 lo 23-17, 23-26, 23-56, 23-70, 23-75, 23-76,
Detection mode, 19-24, 19-40 23-79,23-80
Detection zone, 19-35, 19-39, 19-40
Detector, 17-4, 17-5, 17-11, 17-34, 19-4, 19-20,
19-39,19-40,19-56,19-98,23-63
[),>termirustic model, 17-34,22-12,22-13,
23-62
D-factor, 16-30, 19.92, 23-93

Volume 3 Index Page V3-4 Volume 3/Interrupted Flow

VetS.O? 6.0
 Highway Capacity Manual: A Guide tor Multimodal Mobility Ana/ysis

F G
Facility, Chapter 16, Chapler 17, 18-2, 18-39, Gap, 18.51, 18-54, 18-AA,18-70, 19-7, 19-13,
18-41,18-42,18-48, ]tI-SS, 18-67, 18-79, 19-21, 19-23, 19-37, 19-38, 20-2 to 20-6, 20-8,
18-82, 19-70, 20-6, 20-7, 20-11, 20-37, 21-10, 20-9,20-12,20-13,20-15,20-19,20-2210
22-5,22-14, 23-2, 23-5, 23-7, 23-8, 23-12, 20.24, 20-35, 20-40 lo 20-43, 20-45, 20-47,
23-13,23-19,23-20,23-24,23-61,23-64, 20-49,22-2,22-5,22-11 to 22-13, 23-51 to
23-80, 23-91, 24-3, 24-610 24-11, 24-13, 23-54, 23-58
24-14,24-16,24-17,24-19,24-25,24-29, Gap aCCt'plance, 20.1, 20-2, 20-4, 20-8, 20-9,
24-33,24-34 20-19,22-2,22-5,22-12,22-13,23-51 to
Fixcd-objecl l,ffeclive width, 18-49,24-10 23-54, 23-SS
F1ared approach, 20-3, 20-8, 20-10, 20-46 Generalized servíce volume lable, 16-30,
Flow profi1e, 18-33, 19-9, 19-27, 19-49, 23.3, 19-92
23-76,23-77, 23-82, 23-84 lo 23-86 Geomelriccondition, 16-31, 18-79, 18-80,
Flow rale, 16-11, 17.26, 17-44, 18-11, 18-1410 19-93,20-11,20-37,20-45, 20-46, 21-7, 22-27
18-17,18-30,18-33,18-34,18-44,18-45, c..'ometric delay, 18-34, 18-35, 23-79
18-SO,18-59, 18-80, 19-9, 19-12, 19-19, 19-22 Gn't'n time, 17-5, 19.9, 19-38, 19-48,23-6,23-7,
to 19-27, 19-30 lo 19-32, 19-41, 19-44 to 23-29,23-39,23-44,23-47,23-55,23-82
19-46,19-48,19-56, 19-5S to 19-60, 19-63 lo
19-65,19-67,19-73. 19-74, 19-78, 19--80lo
19-82,19-85,19-87,19-89,19-90,19-94, H
20-13,20-16,20-20,20-22,20-24,20-25, Headway, 17-44, 18-29, 18-68, 18-75, 18.78,
20-27 to 20-30, 20-32, 20-38, 20-40, 20-41, 19-10, 19-11, 19-13, 19-47, 19-97, 20-5, 20-1R,
20-47,21-3,21-7,21-11,21-12, 21-14, 21.17, 20.19, 20-39 lo 20-41, 20-43, 20-45, 21-3 to
21-19, 21-20, 21-8, 22-12, 22.16 to 22-21, 21-6, 21-8, 21.9, 21-13, 21-14, 21.16, 21-17,
22-24,22-27,23.2,23-7,23-9,23-10,23-17, 22-3,22-5, 23-2, 23-32, 23-43, 13-51, 23-63,
23-29, 23-32, 23-37, 23-52, 23.54, 23-55, 23-77, 23.97
23-58,23-59,23-73,23-77,23-78,24-4,24-12 Heavy vehide, 18-59, 18-63, 18-64, 19-12,
lo 24-14, 24-18, 24-20, 24-24, 24.25, 24-33, 19-23,19-25,19-30,19-45,19-46,19-92,
24.34 20.11, 20-18, 20-19, 20-46, 21-4, 21-10, 21-13,
Flow ratio, 19-44, 19-51, 19-58 to 19-62 22-4,12-6, 22-8, 22.14, 22-16, 22-22, 22-25,
Follow-up h...adway, 20-3, 20-4, 20-5, 20-7, 22-28,23-29,23-31,23.95
20-10. 20-18, 20-19, 20-20, 20-21, 22.3, Hindrance, 24-15, 24-16
22-25,23-52 lo 23-54
Force-otL 19-5, 19.37, 19-39
Four-phase pattl'm, 21-2 to 21-4 1
Frec-tlow spt-'l'd (FFS), 16-7, 16-1410 16-16,
lmpt->dance,20-14, 20-16, 20-21, 20-24, 20.26,
17_9,17.12,17_27,17_28,17.35,17_36,17-44,
20-33 lo 20-36, 20-49, 22-20, 22-22
18-3,18-6,18-7,18-13,18-27 to 18.29, 18-30,
1ncidenl, 17-1, 17-7, 17-8, 17-11 to 17-14,17-18
18_31,18-38,18_82,20_30,23-29,23-43,
to 17-23, 17-26, 17-30 lo 17.33, 17-36, 17-38,
23-44,23-46,23-59,23-71,23.79,23-82,
17_39,17-41,17-43,17-44,20_30,21.18
23.93
Incident dearance time, 17.20
FTl'l'-tlow traVl'1time, 23-14, 23-82
Incident detection time, 17-19, 17.20
Fn'l'way, 16-6, 17-13, 17-42, 19-31, 22-11, 23-1,
Incident n-spon"", lime, 17-19, 17-20
23-2,23.4,23-5, D-B, 23-14 to 2H6, 23-19,
Incremental delay, 19-50 to 19-53, 19-90, 19-95
23-20,23-22,23-24,23-26,23-27,23-33,
Influence arca, 16.10, 18-4, 18-10, 19.14, 19-18
23-44,23-45,23-47,23-50,23--60,23-64
lnitial queue, 16-11, 17-27, 18-11, 19-2, 19-18,
Frecway facility, 23-2, 23-60 19-19,19-23,19-29,19-30,19-50,19.51,
Freíghl, 17-35, 17-45 19-53, 19-56, 19-65, 19-67 to 19-69, 19.90,
Fun stop, 16-16, 18.361018-38
19-94, 23-48
Fully acluatl>d control, 19-4, 19-56, 19-76,
Initial queue delay, 19-2, 19-50, 19-51, 19-53,
19-88
19-65,19-69,19-90
Inputs, 17-7, 17-12, 17.23, 17-]8,20-21,22-12,
23-6,23-61,23-62,23-71, 23-91, 23-97, 24.2,
24-25, 24-30, 24.34
lnlelligenl transportation system (ITS), 19-34,
19-99

Volume 3/Interrupted Flow Page V3-S Volume 3 Index

Version 6.0
Highway Capacity Manual: A Guide tor MultimOdal Mobility Analysis

Interchange, 16-6, 16-12, 16-15, 16-17, 16-22,. Le.•.e! of scrvice (LOS), 16-1 lo 16-4,16-6 to I
18-1,18.12, 18-13, 18-23, 18-42, 18-43, 18-57, 16-9,16-12,16-14 lo 16-28, 16-30, 16-31,
18-58,19-99, 2(}-]0,22-1, Chapler 23
I
17-1,17.7,17-30,17-40,17-42,17-43,18-1,
Interchange ramp terminal, 16-15, 16-17, 18-2,18-4, 1S-6 to 18-9, 18-13, 18-3810
16-22,18-13,18-23,18-42,18-43,18-57, 1842, 18-44, 18-48, 18-51 lo 18-54,18-56 to
18-58,22-1, 23-1, 23-6, 23-19, 23-24, 23-28, 18-59,18-61 lo 18-67, 18-70, 18-71, 18-78 to
23-30. 23-31, 23-39, 23-49, 23-63 1S-80, 18-82, 19-1, 19-2, 194, 19-14 to 19-17,
Intemallink, 23-10, 23-35, 23-42, 23-43, 23-47 19-20,19-41,19-44,19-54,19-55,19-70,
to 23-49 19-76, 19-81, 19-.84lo 19-86, 19-&\, 19-90 lo
Intcrrupted flow, 20-11 19-95,20-1,20-2,20-6,20-7,20-9,20-32,
Intersection deJay, 19-24, 19-32 20-37,20-39,20-44 to 20-47, 21-1, 21-9,
Inler .•.aJ, 16-5, 17-39, 17-40, 18-5, 18-34, 19-4, 21-11, 21-18, 21.19, 21-22, 21-25, 21-26, 22-1,
19-9, 19-1I, ]9-13, 19-14, 19-23, 19-28, 19-29, 22-9,22-10,22-12, 22-23, 22-28, 23-2, 23-6,
19-351019-39,19-51,19-52,19-63 to 19-65, 23-10,23-12 to 23-14, 23-16 to 23-18, 23-27,
19-73, 19-76, 22-11, 23-40, 23-41, 23-4710 23-28,23-55,23-56,23-60 to 23-62, 23-70,
23-49,23-57 to 23-59, 24-13 23-78,23-80,23-83, 23-88, 23-89, 23-93,
Island, 19-73, 19-74, ]9-.84, 19-85, 20-13, 20-16, 23-94,24.1. 24-3 lO24-7, 24-9, 24-12 to
20-17,20-39,22-1, 23-24, 23-26 24-18,24-25 to 24-34
lsolated inlel"S('(:tion,16-5, 16-12, 18-13, LcveJ-of-service srore (LOS score), 16-1, 16-8,
19-29,19-94, 20.3, 21-1 16-171016-28,18-1,18-7,18-8,18-41,18-42,
18-44,18-48,18-511018-54,18-56,18-58,
18-59,1$-61 to 18-67,18-70,18-71, 18-78,
J 19-1,19-70,19-.84 to 19-86, 19.90, 19-91,
Jam density, 23-43 23-89,24-17
Jughandle, 23-65, 23-87 Limih..'<ipriority, 22-2, 22-3
Link, 16-3, 16-6, 16-7, 16-17, 16-20, 16-22,
16-24,17-41, ]8-4 to 18-8, 18-13, 18-]9,
K 18-20,18-28,18-39, 18-tI, 18-42, 18-45,
1847 lo 18-49, 18-52 lo 18-54, 18-56 to
K-faclor, 16-30
18-58,18-61,18-63,18-64, 1S-67, 18-70,
18-72, 18-73, 18-78,23-2,23-6,23-10,23-12
L to 23-14, 23.16 to 23-18, 23-27, 23-28, 23-55,
23-56,23-60 to 23-62, 23-70, 23-78, 23-80,
Lme 1. 21-14, 21-15, 21-20 23-83,23-88,23-89, 23-93, 23-94
Lane2, 21.14, 21-15, 21-20 Link Jenglh, 18-19, 18-20, 18-28, 18-45, 23-43
Lane addition, 19-31 Load factor, 18-67, 18-69, 18-77
Lane distribulion,. 20-14, 20-19, 20-37, 23-33, Local street, 16-9, ]6-12, 18-9, 22-29
23-62, 23-63, 23-74 Loop ramp, 23-5, 23-6, 23-22, 23-23, 23-56
Lane group, 16-14, 18-13, 18-23, 18-42, 18-43, Lost time, 18-32, 18-82, 19.9 lo 19-12, 19-38,
18-57,18-58,19-12, ]9-15, 19-21, 19-24, 19-58 to ]9-60,19-62,19-92,23-2,23-7,23-8,
19-40 to 19.56,19-58,19-59,19-61 lo 19-65, 23-10,23-11. 23-30, 23-38 lo 23-45, 2347 to
19-67 to ]9-69, 19-96, 23-3, 23-16, 23-29, 23-SO,23-55, 23-56
23-31. 23-33 lo 23-35, 23-38, 23-55, 23-73,
23-80, 23-81, 23-87
Lane gmup delay, 19-54 M
Lanc ulillzation,. 17-36, 19-23, 19-30, 19-31,
Macrosropic mool'J, 22-13
19-45,19-47,22-14,22-18,22-19,23-3,23_5,
Mainlinc, 23-23, 23-65
23-81023.10,23-22,23-27,23-30,23-32 to
23.36, 23-49, 23-62, 23-63, 23-73, 23-74, Major slrt'l't, 16-2, 16-14, 17-2, 18-2, 18-21,
23-81 19-2,19-8,19-59,19-61,19-71, 19-78, 19-80,
Lane width, 18-44, 18-45, 18-59, 19-23, 19-30, 19-82 to 19-85, 20-1, 20-3 lO20-7, 20-12 lo
19-31, 19-45, 19-47, 22.11, 23-29, 23-31 20-21, 20-23, 20-24, 20-30, 20.31, 20-33,
20-34, 20-37, 20-38, 20-40, 20-42, 20-45,
20-47,21-22,22-29,23.9,23-13, 23-65, 23-71,
23-74 lo 23-79, 23-81, 23-87 to 23-89, 23-91
Maximum green,. 19-13, 19-23, 19-35, 19-52,
19-56
Maximum n.'CaJl,19-37

VoIume 3 Index Page V3-6 VoIume 3/Interrupted Flow

V~6,O
 Highway Capacity Manual: A Guide for Multimodol Mobility Anolysis

Median, 17-27, 18-15, 18-19, 18-27, 18-2!!, Multimodal, 16-1, 16-3, 16-4, 16-29, 16-33,
18-59,18-60,18-82,19-36,19-75,20-10, 18-1, 18-42, 18-58, 18-83, 19-1, 19-98,24-35
20-15,20-18, 20-19, 20-25, 20-27, 20-37 lo
20-39, 20-46, 2.3-2, 2.3-5, 23-11, 23-65, 23-66,
23-71, 23-73, 23-78 N
Median U-Ium inlersection (MUl), 23-2, Near-side slop, 18-n, 18-73
23-5, 23-6, 23-8 to 23-11, 23-65 lo 23-67, Node, 23-11, 23-24, 23-63, 23-71
23-71, 23-73, 23-74, 23-76 to 23-81, 23-87, Nonreslricti\'e median, 18-60
23-89,23-91,23-94,23-97
Meeting, 17-10, 17-40, 17-45, 19-98, 20-50,
22-30,23-97,24.3,24-5 lo 24-7, 24-15 to
24-17,24-20,24-21,24-25,24-29,24-31
o
Off-linO'bus slop, 18-71
Merge, 20-13, 21-3, 22-8, 23-5, 23-11, 23-22,
Off-ramp, 23-16, 23-20, 23-42 to 23-46
23-66 lo 23-68, 23-73, 23-75, 23-77 lo 23-79,
Offsel, 17-4, 18.32, 19-24, 19-38, 23-6, 23-9,
23-85,23-91,23-93,23-95
23-11, 23-29, 23-76, 2.3-82, 23-83, 24-28
Minimum gn't.'n, 19-23, 19-35 lo 19-37, 19-40,
Off.strt.'t.'t, 16-19, 16-24, 18-48, 18-61, 24-1 to
19-56,23-29
24-3,24-51024-7,24-10,24-11, 24-14, 24-15,
Minimum !\'Cal1, 19-37
24-17,24-25,24-30,24-34
Minor movement, 19-4, 19-5, 20-3, 20-4, 20-6,
One-stage gap acccplancc, 20-15
20-13, 20-20, 20-JO, 20-32, 20-33
On-ramp, 19-31, 23-64
Minor slrl't.'l, 19.71019-9,19-59,19-61,19-71,
On-time arrival, 17-10
19-79, 19-80, 19-84,20-1,20-4 lo 20-6, 20-10,
Opcrational analysis, 16-13, 16-29 lo 16-31,
20-12, 20-13, 20-16, 20-18, 20-19, 20-21,
17-14,18-20,18-79, 18-tlO, 19-79, 19-92,
20-33,20-47,21-22,22-29,2.3-9,23-13,23-71,
19-93, 20-46, 21-25, 22-28, 22-29, 23-6, 23-26,
23-74 lo 23-77, 23-85, 23-88, 23-93, 23-94
23-28, 23-30, 23-61, 23-62, 23-66, 23-70,
Mobility, 1&-7, 16-10, 1&-12, 17-13, 18-6, 18-10,
23-71, 23-88, 23-89, 23-94, 23-95, 24-11,
2J-.1,24-7
24-33
:'1001',16-1,16-3,16-4,16-61016-8.17-6,
Opposing approach, 19-9, 19-63, 21-2, 21-3,
17-30,17-40, ]7-41, 17-43, 18-1, 18-6 lo 18-8,
21-7 to 21-9, 21-14, 21-20, 21-26, 23-69
18-41, 19-1, 19-4, 19-7, 19-8, 19-1410 19.16,
OutpulS, 17-43, 20-46, 21-25, 12-28, 23-28,
19-21,19-23,19-24,19-34 lo 19-40, 19-64,
23-61,24-34
19-74,19-75,19-98,20-6,20-37,20-45,20-46,
Overflow queue, 18-36, 19-52, 19.69
21-9,21-11, 21-22, 21.24, 21-25, 12-9, 22-26
lo 22-29, 23-1, 23-28, 24-1, 24-9, 24-15, 24-17
to 24-25, 24-32, 24-33 p
ModO'group, 24-17, 24-19, 24-21
Model, 17-37, 17-38, 17-41, 17-43, 17-44, 18-4, Partial cloverleaf inlerchange (parclo), 23-10,
18-12,18-14, 18-22, 18-27, 18-33, 18--34, 23-11, 23-15, 23-19, 23-20, 23-22, 2.3-23,
18-36,18-39,18-80,18-81, 19-22, 19-40, 23-26, 23-34, 23-35, 23-61
19-62, 19-94, 20-1, 20-3, 20-4, 20-8 to 20-10, rartial diamond intcrchange, 23-20
20-19, 20-38, 20-45, 21-5 lo 21-8, 21-27, 22-1 Passagl' time, 19-20, 19-23, 19-35, 19-52, 19-56,
to 22-5, 22-11 to 22-13, 22-20 to 22-23, 22-25 23-29
to 22-27, 23-28, 23-33 to 23-37, 23-43, 23-44, PasSl'ngef car, 16-29, 19-30, 19-46, 21-4, 22-7,
23-49,23-52, 23-53, 23-63, 23-70, 24-17, 22-16, 22-17, 23-42
24-23, 24-34 Passcnger-car equivalt'Tlt (PCE), 22-21, 22-22
Monle Carlo melhod, 17-38 Passeng",r sen/ice lime, 18-74
Motorized vehide mode, 16-4, 16-6, 16-7, l'assenger lrip l",ngth, 18-67, 18-69, 18-77,
18-6,18-7,18-22, 18-59, 18-60, 19-14, 19-16, 18-78
19-73 lo 19-76, 19-87, 19-88, 20-1, 20-3, Pavement condition fatin}; 18-60, 18-63
20-11,20-21, 20-37 P",ak hour, 16-U, 16-29, 16-30, 17-7, 17.35,
Movement capacity, 16-14, 18-21, 18-25, 17-41,17-44,18-11,18-16,18--17,19-19,
18-32, 18-39, 20-3, 20-11, 20-21, 20-22, 20-24, 19-23 to 19-26, 19-92,20-2,20-10 to 20-12,
20-25, 20-27, 20-28, 20-33 1020-35, 20-47 20-46, 20-47, 21.10, 21-11, 21-25, 22-14 to
Movcmenl group, 16-13, 18-13 to 18-15, 18-21, 22-16,22-28,23-2,23-29,23-73,23-95,24-2,

19-21, 19-22, 19-24, 19-26 lo 19-34. 19-42 to 24-9, 24-12, 24-16, 24-18,24-24,24-25

19-44,19-46,19-67,23-73
Move.up time, 20-30, 21-3, 21-18
Multilanl' mundabout, 22-1, 22-5, 22-7, 22-19,
22-28

Volume 3{Interrupted Flow Page V3-7 Volume 3 Index

Version 6.0
Highway Capacity Manual: A Guide for Mulfimodol Mobility Anolysis

Peak hourfactor (PHF), 16-11, 16-29, 16-30, Phase, 16-30, 17-4, 17-35, 18-2, 18-12, 1S-23,
17-35,17-44, 18-1l, 18-16, 18-17, 19-19, 18-32 to 1S-34, 19-2, 19-4 to 19-13, 19-15,
19-23 to 19-26, 19-92, 20-2, 20-10 to 20-12, 19-20 to 19-24, 19-26, 19.27, 19-34 lo 19-39,
20-46,21-10,21-11,21.25,22-141022-16, 19-41, 19-49 to 19-52, 19.55, 19.56, 19-58 to
22-28,23-2,23-29, 23-73, 23-95, 24-2, 24--8, 19-63,19-67,19-68,19-71 to 19.73, 19-75,
24-9,24-12,24-16,24-18,24-24,24-25,24-30 19-76,19-78 to 19--81, 19-83, 19-86 to 19..sB,
t024.33 19-92, 19-94, 19-97,21-1 to 21-4, 21-22 to
P,,'dL'Slriall, 16-2 to 16-4, 16-6, 16-8, 16-9, 16-17 21-24,23-7,23-9,23-17,23-29,23-39,23-41,
lo 16-21, 16-24, 17.5, 17.34, 17-35, 18-2, 18-4 23-42, 23-44, 23-45, 23-47, 23-48, 23-50,
lo 18-9, 18-31, 18-32, 18-38, 18-41 lo 18-56, 23-53,23-54,23-57,23-58,23-61,23-65,
18-591018-61, 18-67, 18-70, 18-78, 18-83, 23-68,23-81 to 23-83, 23-90, 23-92
19-2,19-5 to 19-7, 19-9, 19-14, 19-16, 19-17, Phase Ilow ratio, 19-59, 19-60, 19-62
19-23,19-25,19-31,19.36,19-37,19-40, Phase lost time, 19-10 lo 19-12, 19-60, 19-62
19-44, 19-45, 19-47, 19-48, 19-70 to 19--85, Phase pair, 19-7, 19-59
19--88,19-90,19-98,19-99,20-3 to 20-6, Phase pattem, 21-3
20-13,20-15,20-21, 20-33 to 20-46, 20-SO, Phase l'ffall, 19-23
21-10,21-22,21-23,22-7,22-8,22-10, 22-11, Phase sequence, 19-4, 19-6, 19--8, 19-9, 19.23,
22-20 to 22-22, 22-26, 22-27, 22-29, 22-31, 19-34,19-37,19.38,19-56,19-59 to 19-62,
23-19,23-22,23-28,23-29,23-32,23-SO, 19-94,21-22
23-60, 23-65, 23-88 to 23-90, Chapter 24 Planning and preliminary enginl'ering
Pcdestrian circulation mute, 24-27 analysis, 16-6, 16-13, 16-31, 17-14, 18-5,
Pedestrian clcar interval, 19-36, 19-37, 19-75, 18-16, 1S-17, 18-20, 18.24, 18-31, 18-80,
19-78,19-79,19--81 19-14,19-24 to 19-26, 19-39, 19.79, 19-93,
Pedestrian f10w rate, 18-44,18-49, 18-SO, 20-46,21.25,22-28,24-34
19-23, 19-31, 19.73, 19-74, 19-83,20-13, Planning time index, 17-10
20-38,20-40,23-29,24-12,24-13, 24-28 Platoon, 16-2, 16-5, 16-18, 17.2, 17-4, 18-2,
Pedestrian mode, 16-6, 18-6, 18-8, 18-59, 18-4,18-26, 1S-32, 18-33, 18-43, 18-80, 18-81,
18-60,19-77,22-26,24-9 18-82, 19-2, 19-18, 19.23, 19-26 to 19-29,
PL'lÍL'Strian plaza, 24-6, 24-13, 24.27 19-63,19-64,19-82, 2().3, 2()'7, 2().20, 20-21,
Pedcstrian l'ffall, 19-37 20-39,20-40,24-3,24-4,24-13,24-14
Pedestrian service time, 19--82 Platoon dispersioll, 18-33, 18-8(1, 18--81
PL'lÍestrian space, 16-1, 16-6, 16--8, 16-17 to Platoon ratio, 1S-32, 19-23, 19-26 to 19-28,
16-21,18-1,18-6,18-7,18-41,18-42,18-49, 19-64
18-51,18-56,19-70,24.3,24-6,24-13,24-H Point, 16-3, 16-6, 16-13, 16-30, 17-10, 17-23,
24-30 17-30,17-34,18-4 to 18-6, 1S-14 to 18-19,
Pedcstrian start-up time, 20-38, 20-39 18-25 to 18-28, 18-30 to 18.33, 18-37, 18-43,
Pedestrian strcct, 24-1, 24-3, 24-13 18-47,18-54,18-59,18-64,18-67,18-68,19-6,
Pl'dcstrian walkway, 24-1, 24.27, 24-28, 24-33 19-22,19-24,19-28,19-32 to 19-34,19-37 to
Pedestrian zone, 24-3, 24-13, 24-28 19-39,19-49,19-55,19-63,19-73, 19-83,
Performance measure, 16-1, 16-2, 16-6 to 16-9, 19-87, 19..sB, 19-94, 19-95, 2().29, 20-42,
16-12 to 16-14, 16-17, 16.18, 16-21 to 16-23, 21-26,22-13, 24-2, 24-7, 24-10, 24-16, 24-20,
16-25,16-26,16-28,16-31,17.1,17-3,17-7to 24-25,24-27, 24-32
17-9,17-15,17.25,17-27,17-38 to 17-42, Potential capacity, 20-3, 20-11, 20-19 lO 20-22,
18-1,18-3,18-5 to 18-7, 1S-13, 18-20, 18-23, 20-25,20-33 to 20.35
18-38,18-39,18-41. 18-42, 18-47, 1S-S6, Pn'cision, 16-4, 16-31. 17-7, 18-4, 18-8(1, 19-14,
18-58,18-60,18-66,18-70,18-71,18-75, 19-93
18.78,18-79,18-81,19-1,19-3,19-14,19-16, Preemptioll, 16-9, 17.H, 18-9, 19-16, 19-21
19-20,19-41,19-55,19-57,19-70,19_85, Preposilioning. 19-30
19-86,19-91,19-93,19-94,19-95,20-1,2()'7, Presence detection, 19-23, 19.24
20-9,20-37,21-25,22-1,22-2, 22-10, 22-29, Pretimed control, 17-35, 19-4, 19-34, 19-38,
23-6,23-9,23-10,23-12, 23-18, 23-26, 23-28, 19-41, 19-49, 19-92
23-55,23-61, 23-62, 23-70, 23-71, 23-86, Prevailingevndilion, 19-11, 19-12, 19-41
23-88,23--89,24-2,24-6, 24-28, 24-30 Priority reversa\, 22-3, 22-10
Pennitted mode, 18.37, 19-7, 19-8, 19-74 Progression, 16-7, 17-11, 18-6, 18-7, 18-12,
Pcnnitted mm, 19-7, 19-94 19-5,19-15,19-261019-29,19-38,19-51,
19-56,19-94,19-98,22-14,23-7,23-9,23-42,
23-44,23-74,23-76, 23-82

VoIume 3 Index Page V3-8 Volume 3/Interrupted F10w

Vt"rWn 6.0
 Highway Capacity Manual: A Guide tor Multimodal Mobility Analysis

Reliabilily rating. 17-10, 17-27

Q Reliability reporting period, 17-4, 17-7 to
Quality of S\.'rvic•.•
, 16-1, 16-3, 16-8 to 16-10, 17-9,17-12 to 17.15, 17.18, 17-24 to 17-29,
16-17,16-19,16-22,16-24,16-26,18-1, 18-7 17-31,17-371017-40
to 18-10, 18-41, 18-43, 18-53, IR-57, 18-58, R~'Sidualqueue, 16-11, 17-27, 18-11, 19-19,
18-66,18-68,19-1, 19-16to 19-18, 19-70, 19-30,19-52,19-55,19-57,19-69,22-23,
19-71, 19-86, 22-9, 23-2, 23-{>{), 24-1, 24-3, 23-12, 23-57, 23-58
24-6,24-7,24-26,24-31 Rest-in-walk mod •.• , 19-21, 19-36, 19-75
Queue, 16-7, l6-lOto 16-12, 16-14, 16-29, Re-;Iricled crossing U-tum intersection
16-30,16-32, 17-36, 18-2, 16-6, 18-10 to (RCUT), 23-vii, 23-viii, 23-ix, 23-2, 23-5,
18-12,18-15,18-21, 18-25 to 18-27, 18-30, 23-6,23-8,23-9,23-10,23-11, 23-12, 23-13,
18-36,18-37,18-70,18-71,18-82,19-2,19-10 23--65,23-66,23-67,23-71,23-73,23-74,
to 19-13, 19-15, 19-16, 19-18 to 19-20, 19-22, 23-75,23-76,23-77,23.78,23-79,23-80,
19-29,19-30,19-35,19-31'1,19-40,19-41, 23-81,23-83,23-87,23-88,23-89,23-91,
19-49 to 19-56, 19-62 lo 19-67, 19-69, 19-79, 23-93, 23-94, 23-95, 23-97
19.98,20-3,20-7 to 20-11, 20-22 to 20-26, Restricti- .•.•m•..
-dian, 16-29, 16-30, 18-15, 18-19,
20-28 to 20-32, 20-34, 20-35, 20-45 lo 20-49, 18-28
21-3,21-10, 2H9, 21-22, 21-24, 21-26, 21-27, Right-of-way, 16-3, 16-9, 16-17, 16-22, 16-27,
22-10 to 22-12, 22-23, 22-24, 22-29, 23-7, 18-4,18-9,18-41,18-57,18-66,18-67,19-5 to
23-9,23-11,23-12, 23-16, 23-27, 23-28, 23-30 19-7,19-17,19-40,19-89,19-90,20-4,20-5,
to 23-32, 23-36, 23-31'1to 23-44, 23-47 lo 20-9,21-2 to 21-4, 22-20, 22-26
23-50, 23-52, 23-53, 23-55 lo 23-59, 23-61, Right-tum-on-red, 18-82, 19-20, 19-23, 19-25,
23-62,23-70,23-78,23-80,23-82,24-27 19-44,19-73, 19-81, 19-85, 19-%,23-1,23-6,
Queue accumulalion polygon, 19-13, 19-62 to 23-20,23-28,23-62,23-68,23-29,23-72,
19-<>4 23-94, 23-97
Queue delay, 19-51, 20-9 Ring, 19-6, 19-7, 19-37, 19-58 to 19-62
Qul'ue jump, 18-71 Roundabout, 16-9, 16-15, 16-17, 16-22, 16-29,
Queue I•.•nglh, 18-2, 19-2, 19-18, 19-41, 20-3, 16_30,16_32,18-8,18_14,18_22,18-23,18-34,
20-7,20-11,20-28, 20-29, 20-32, 20-46, 2048, 18-35,18-42,18-43,18-57,18-58,18-67,
21-19,22-10, 22-24, 22-29, 23-36, 23-41, 18-71 lo 18-73, 18-82, 21-26, Chapter 22,
23-43, 23-50, 23-52, 23-57, 23-58, 23-62, 23-2,23-7, 23-18 to 23-20, 23-22, 23-24,
23-78 23-26,23-27,23-50,23-52, 23-SS, 23-57,
Queue spillb<lck. 19-16, 19-20,20-7,21- 10, 23-61,23-62,23-74
23-7,23-12,23-49,23-61, 23-78 Running sf't->t-od, 16-16, 16-24, 16-27, 17-26,
Queue storage ratio, 19-20, 19-,54,19-55, 18-30,18-32,18-44,18-47,18-52,18-53,
23-27,23-31, 23-55, 23-56 18-59,18-61 lo 18-64, IR-72 lo 18-75,18-81,
Queued stat •.•, 20-9 19-10
Queuing area, 24.1 Running time, 16-7, 16-14, 16-23, 16-27, 17-27,
17-35,17-36,18-6,18.21,18-27,18-29 to
18-33,18-38,18-47,18-61,18-62,18-67,
R 18-70 to 18-72, 18.74, 18-75
Rural. 17-10, 18-19, 19-55,21-22,21-26,22-26,
Ramp, 16-6, 16-12, 17-43, 18-5, 18-12, 18-35,
23-20,23-66
19-99,20-10,22-11,23-3,23-5,23-6,23-10,
23-15, 23-19, 23-20, 23-22 to23-28, 23-30,
23-33, 23-38 lo 23-46, 23-48, 23-61, 23-62,
23-64,23-87,23-91, 23-96, 24-3, 24-13
s
Ramp meter, 17-43,22-11,23-27,23-64 Saturation flow rat •.•, 17-26, 17-44, 18-13,
Rank, 204, 20-5, 20-13 to 20-16, 20-20 lo 18-36,18-37,18-80,18-82,19-3,19-9,19-11
20-27,20-30 to 20-36 to 19-13, 19-30, 19-41, 19-44 to 19-49, 19-58,
Receiving lan~'S,19-23, 19-33, 19-48, 21-8 19-59,19-63,19-64,19-67,19-1;9,19-94,
Recrealional vehide (RV), 16-9, 18-9, 19-17 19-96,20-5,20-10,20-24,22-3,23-3,23-9 to
R•.•curring congestion, 17-33 2J..11, 23-22, 23-27, 23-31 to 23-33, 23-35,
Red dearance interval, 19-5, 19-9 to 19-13, 23-37,23-51, 23-54, 23-55, 23-SIl,23-59,
19-36 to 19-38, 19-79, 19-88,23-45 23-63,23-72,23-77,23-78,23-82,23-87
Red time, 19-9, 19-11 Saturation headway, 19-10, 19-11, 19-40,21-3,
Reentry delay, 18-67, 18-70 to 18-74 21-5 to 21-8, 21-10, 21-11, 2H3, 21-16,
Reference phasc, 19-24, 19-38, 23-82 21.17,21-20,21-21,23-47, 23-48, 23-54,
Regression model, 20-1, 22-2, 22-3, 24-17, 23-63
24-25

Volume 3/Interrupted FIow Page V3-9 Volume 3 Index

Ve~1On6,0
Highway Capadty Manual: A Guide for Multimodal Mobility Analysis

Scenario, 17-3, 17-7, 17-8, 17-l3, 17-14, 17-24 Space, 16-3, 16-8, 16-18 10 16-21, 18-7, 18-8,
1017.27,17.31, 17.32,. 17.35, 17-36, 17-39 to 18-22,. 18-411018-43,18-49,18-51, 18-56,
17-41. 17-43, 19-26, 19-57,20-7,20-12,. 20-15, 19-32,. 19-46, 19-55, 19-71, 19-75, 19-77,
20-43, 21.10, 21.11. 21.22, 22.16, 23-82, 19-79 to 19-81, 19-83, 20-31, 23-27, 23-41,
24-23 23-44,23-78,24-4,24-10,24-13,24-14,24-27,
Scenario generaliun, 17.8, 17-14, 17-26 24-28, 24---30,24-31. 24-35
Section,. 17-3, 17-29, 24-11, 24-22 Space meiln speed, 17-10
Segm",nt, 16-3, 16-510 16-10, 16-12 lo 16-29, Spacing.. 16-5, 16-30, 18-15, 18-27, 18-29, 18-44,
16-32,17-9,17-12,17-18,17-20 to 17-27, 18-46,18--53,19-28,23-1. 23-5, 23-7, 23-20,
17-29,17-37, Chapler 18, 19-20, 19-28, 23-42,. 23-43, 24---11.24-28
19-29,19-33,19-38,19-49,19-81. 19-96,20-9, Spatiill stop rate, 16-16, 17-9, 18-38, 18-39
20-46,21-25,22-28,23-7,24-2,24-3,24-6, Spedal evenl, 17-4, 17-710 17-9, 17-12, 17-18,
24---9,24-13,24-19 to 24-21 17-22 to 17-25, 17-37, 19-49, 24---3,24-27
Segmenl delay, 19-%, 20-9 Spt.ftl, 16-1. 16-4, 16-6 to 16-8, 16-12, 16-14 10
Sr.'miactualoo control, 19-4, 19-5 16-16,16-18 to 16-20, 16.23, 16-24, 16-27,
SensitivilY analysis, 22-24 16-29,16-30,17-1. 17-4, 17-5, 17-9, 17-10,
Service Row rate, 24-5 17-24,17-26,17-28,17-30,17-34 lo 17-36,
Servic", measure, 16-2, 16-3, 18-1, 18-4, 19-2, 17-43,17-44,18-1, 18-6, 18-7, 18-15, 18-19,
19-4,20-39,20-44,22-9,23-1,23-12,24-3, 18-20,18-27 to 18-32,. 18-36, 18-38, 18-43,
24-8,24.9,24.13,24---14 18-47 lo 18-53,18-55,18-60 to 18-63, 18-68,
Service lime, 18-71, 19-13, 19-24, 19-41. 19-44, 18-71,18-72, 18-75, 18-77, 18--80 10 18-82,
19-82,21-3, 21.5, 21-6, 21-8, 21-11. 21-14, 19-5,19-11,19.12,19-16,19-23,19-24,19-35,
21-18 19-36,19-38 10 19-40, 19-71, 19-73, 19-74,
Service volume, 18--79, 19-92,. 24-33, 24-34 19-81,19-84,20-30,20-33,20-38,20-39,22-8,
Sr.'vere weather, 17-26, 17-31 22-26,23-2,23-12,23-27,23-29,23-32,23-43,
Shared lane, 16-13, 18-17, 18-35, 19-21.19-24, 23-44, 23-46, 23-66, 23-77, 23-78, 23-90,
19-27,19-33,19-41, 19-43, 19-44, 19-47, 23-93,24-4,24-71024-9,24-13, 24-15 10
19-63,19-67,19-94,20-3,20-22 to 20.24, 24-17,24-19,24-20,24-28,24-30 to 24-33
20-28 lo 20-31, 20-47, 23-36, 23-73 Spt-oedadjustment faclor (SAF), 18-52, 18-63,
Shared-lanecapacity, 20-22, 20-28 19-84
Shared-use path, 24-1, 24-2,. 24-6 to 24-8, 24-14 Sptoed hannonization, 17-36
to 24-17, 24-21, 24-23, 24-30, 24-31, 24-34 Spíllback, 16-2, 16-12, 17-2, 18-2, 18-12, 18-24,
Shock wave, 23-43, 23-44 18-26,18-27,19-2,19-49,23-7,23-24,23-27,
Short lenglll, 18--12 23-49
Shoulder, 16-6, 17-11 to 17-13, 17.20 to 17-22, Spillover, 23-12, 23-27, 24-7, 24-11, 24-33
17-26,18-44 10 18-46, 18-52, 18-53, 18--59, Splil, 17-4, 17-35, 19-8, 19-9, 19-24, 19-27,
18-60,18-64,19-31,19-871019-89,19-91 19-34 to 19-36, 19-38, 19-39,23-20,23-83,
Shy distance, 18-49, 18-SO, 24-10, 24-12 24---5,24-9,24-17,24-18,24-32,24-33
Side streel, 19-29,23-8,23-65,23-82,23-87, Split phasing. 19-8, 19-9, 19-27, 19-34
23-91, 23-92 Splitler island, 22-17
Sidqh1lll, 24-1 Stairway, 24-6, 24.11, 24-13, 24-14, 24-28
Sidewalk, 16-8, 16-171016--21, 18-4, 18-7, Start.up losl time, 18-32, 19-10 to 19-13, 19-22,
18-41 to 18-44, 18-46 lO 18-SO, 18-52, 18-53, 19-62, 19-96, 23-39, 23-44, 23-46, 23-63,
18-56,19-70,19-7], 19-75 10 19-77, 19-80, 23-85
19-81,19-88,23-60,24.1. 24-6, 24-10 to Slochastíc mude!' 22-12, 22-13
24---12,.24-27, 24-28 SIOp rate, 16--1, 16--12, 16-16, 17-27, 18-1, 18--13,
Simulalion,. 17-43, 17-44, 18-81, 18-82,. 19-32, 18-21, 18-36, 18-38
19-941019.96,20-8 lo 20-10, 20-37, 20-38, SIOp spacing.. 18-69
22-10,22-13,22-31,23-27,23-28,23-SO, Slop-linedetector lenglh, 19-24, 19-39
23-59,23-611023-63,23-70,24-34 Stoppt-'d delay, 20-3(}
Singleentry, 19-23, 19-37, 22-4 Storage lenglh, 18-18, 19-20, 19-33
Single-Iane roundaboul, 22-1. 22-4, 22-5, Street comer, 19-77
22-28 Study perlud, 16-5, 16-10 to 16-12, 17-3, 17-7,
Single-poinl urban inlerchange (SPUI), 23-2, 17-121017-15,17-25,17-29,17-38,17-41,
23-5, 23-10, 23-11, 23-19, 23-20, 23-23, 23-26, 18-4,18-5,18-10 to 18-12,19-14,19-18 to
23-30, 23-31, 23-33, 23-61 19-20,19-22,19-73,19-87
Subject approach. 18-62, 19-27, 19-31, 19-47,
19-54,19-88,19-91,20-15,21-21021-4,21-6,
21-7,21-17,21-18,21-26,23-34

Volume 3 Index Page V3-IO Volume 3jlntell\Jpted FIow

Version 6.0
 Highway Capacity Manual: A Guide for Multjmodal MobiUty Analysis

Suslainro spillback, 16-2, 17-2, 18-2, 18-25, Travellimc, 16-2, 16-14, 16-15, 16-19, 16-20,
18-27,19-2,19-45,19-49 16-23,16-24.16-27,17-1, 17.3, 17.7to 17.10,
System,l6-6, 17.3, 17-4, 17-11, 17.13, 17.20, 17-15,17-25,17-271017-29,17-31,17-40 lo
17-34,17-38,17-40,17-43.17-45,18-5,18-22, 17-43,18-2,18-34,18-68,18-73,18-75,18-76
18-23,18-35,18-60,18-68, HI-oo, 19-5, 19-18, lo HI-71l,18-82, 19-11, 19-15, 19-24, 19-95,
19-27,19-28,19-34,19-381019-40,19-56, 21-7,21-18,23-1, 23-3, 23-12 to 23-15, 23-27,
19-77, 19-9~, 20-8, 20-10, 21-6, 22-13, 23-13, 23-62,23-71,23-77,23-78,23-82,23-93,
23-27, 23-4~, 23-49, 23.57, 23-58, 23-82, 24-20
23-83, 23-92, 23-96 Trav!'l time dislribution, 17-7, 17-15
Syslem !'lement, 19-77 Travel time index, 17-9, 17-28, 17-29
Travel time ratl', 18-75 to 18-77
Travel time reliability, 16-2, Chapler 17, 18-2
T Truck, 16-9, 16-10, 17-11, 18-9, 18-10, 19-12,
Temporal variability, 17-7 19-17,19-18,19-73,20-14,24-12
llm>e-ll'vl'l diamond interchange, 23-20 Tum bay spillback, 18-27
Through whidcs, 16-7, 16-14, 16-16, 17-9, Tum lane, 16-16, 16-29, 16-30, 17-)6, 18-15,
17-29,18-6,18-13,18-14,18-18 to 18-21, 18-17,18-18,18-39,18-40,19-21. 19-25,
18-26,18-27,18-30,18-31,18-34,18-36, 19-33,19-43,19-60 10 19-63,19-67,19-92,
18-38, lB-62, 18-74, 19-20, 19-22, 19-76, 20.10,20-16,20-17,20-23,20-24,20-28,
19-88,20-5,20-6,20-31, 21-3, 21...t 23-6, 22-19,23-8,23-29,23-34,23-35,23-37,23-38,
23-44,23-65,23-73,23-74,23-81 23-63,23-66,23-68,23-74,23-84,23-85,
Throughput, 21-3, 21-7, 21-17, 23-63 23-94
Tight uroon diamond interchange, 23-20 Tuming mO\'t'mt'nt. 19-95, 20-26, 20-46, 21-4,
Time intervai, 16-5, 16-13, 17-3, 17-14, 18-5, 21-5, 21-11, 21-25, 21-26, 22-14, 22-18, 22-28,
tB-20, 19-14, 19-39, 19-64, 20-5, 20-39 23-2,23-5,23-6,13-8,23-9,23-11,23-20,
Timl'...space, 19-9, 19-77, 19-79 to 19-81, 19.83 23-22,23-25,23-26,23-29,23-31,23-32,
TooI, 16-2, 16-23, 16-32, 17-2, 17-4, 17-34, 23-34,23-37,23-48,13-51, 23-57, 23-61 to
17-43,17-44,18-2,18-12,18-13, 18-27, 18-43, 23-63,23-68,23-71,23-73, 23-81, 23-82
18-58,18-66,18-80 10 18-82, 19-2, 19-20, Two-Iane highway, 24-1, 24-6
19-71, 19-86, 19-94 lo 19-96, 20-8 to 20-10, Two-phase patlem, 21-3
20-37,20-38, 21.10, 22-10 to 22-13, 23-27, Two-slagt' crossing, 20-15, 20-38 to 20-41,
23-28,23-59,23-62 to 23-64, 23-70, 23-81 20-44, 23-88
Traffic analysis 1001,18-80, 19-94, 22-12, 23-28 Two-slage gap acceptance, 20-3- 20-8, 20-15,
Traffic cirde, 22-10, 23-26 20-25,20-27
Trafficcondilion, 16-6, 16-9, 16-31, 18-4, 18-5, Two-way left-tum lane [fWLTL), 18-18,
18-9,18-24,18-79,19-12,19-14,19-16,19-45, 18-60,19-33,20-10
19-74,19-93,20-45,21-2,22-27,23-32 Two-way STOP-conlroUed (I1VSC), 16.9,
Traffic control dl'viU', 16-7, 17-9, 17-30, 11:1-6, 16-17,16-22,16-32, 18-5, 18-8, 18-23, 18-42,
18-12, 18-27, 18-30, 18-72, 19-11,22-11, 18-47,18-57,18-61,18-72,18-82, Ch¡¡ptl'r
23-71 20, 21-19, 21-22, 22-2, 23-3, 23-18, 13-26,
Traffic prl'SSllre, 23-11, 23-32, 23-33 23-27,23-51, 23-64
Transit fn.'quency, 18-67, 18-69, 18-75, 18-76
Transil mude, 16-6, 16-8, 18-6 to 18-8
Transition, 17-11, 24-27
u
Travel demand model, 17-40 Uncertainty, 20-6, 21-3, 22-2, 22-5
TraVl'1modl', 16-1, 16-3, 16-6, 16-8, 16-10, Unconlrollt.'d, 16-H 18-5, 18-6, 18-21, 18-32,
16-12,16-17,16-22.16-26,17-4,18-1, 18-7, 18-35,18-36,18-44,18-47,18-51,18-5-t,
18-8,18-10,18-13, 18-41, 18-42, 18-57, 18-58, 18-62, 19-11, 19-71. 20-1, 20-37
18-66,19-1.19-16,19-18,19-20,19-70,19-86, Unifonndt'lay, 19-2, 19-50, 19-51, 19-53,
24-3 19-62,19-64,19-65,19-67,19-68,19-95
Travl'l spl"l'd, 16-1, 16-7, 11>-121016-20, 16-22 Uninlt'rrupted tlow, 13-85
lo 16-24, 16-261016-28,17-5,17-9,17-27, Unitextt'nsion, 19-35
17-29,18-1,18-6, t8-7, 18-13, 18-38, 18-42, Unm('1 d('mand, 19-51, 19-53, 19-65, 19-67 10
18-51,18-58,18-62,18-66,18-69,18-71, 19-69
18-75, 18-76, 18-81, 18-82, 20-42, 23-12, 24-5, Unsignalized inlersection, Chapter 20,
24-7 Chaptt'r 21, Chapter 22, 23-7, 23-19, 23-57

Volume 31Interrupted Flow Page V3-11 Volume 3 Index

Version 6.0
Highway Capacity Manual: A Guide tor Multimodal Mobility Analysis

Urban, 16-1 10 16-3, 16-6 10 ItrIO, 16-12 10 Volume, 16-3, 16-4, 16-610 16-8, 16-11, 16-13,
16-15,16-17,16-19,16-22,16-23,16-26,
16-27,16-29,16-30,16-32,16-33,17-1, 17-2,
16-14,16-16,16-30,16-31,16-33,17-9,17-13,
17-15,17.171017-19,17-23,17-26,17-30,
)
17-4 lo 17-7, 17-910 17-14, 17-17, 17-18, 17-37,17-41,17-43,18-2,16-6,18-7,18-11,
17-21, 17-24 lo 17-28, 17-34, 17-35, 17-37, 18-13,18-14,18-16,18-17,18-231018-27,
17-42. 18-1 1018-10, 18-12, 18-13, 18-19, 18-291018-31, 18-33, 18-35 lo 18-39, 18-42,
18-22 10 18-25, 18-28, 18-30, 18-32, 18-34, 18-43, 18-52, 18-53, 18-63, 18-64, 16-67,
18-391018-41, 18-48, 18-57, 18-61, 18-66, 18-71,18-73, 18-79, 18-80, 19-1, 19-2, 19-15,
18-71, 18-79 lo 18-83, 19-2, 19-17, 19-18, 19-16,19.181019-20,19-22,19-241019-26,
19-27,19-49,19-77,19-88,19-98,19-99,20-7, 19-28,19-29,19-31 lo 19-33,19-36,19-38,
20-10,20-20,20-46,20-47,20-49,20-50, 19-41,19-44,19-50,19-53 10 19-59, 19-65,
21-22,22-8,22-13,22-26,23-3, 23-6, 23-11, 19-73,19-79,19-80,19-84,19-90,19-92 to
23-12. 23-20, 23-23, 23-52, 23-57, 23-59, 19-95, 20-6, 20-7, 20-9, 20-12, 20.15, 2{).23,
23-60,23-66,23-71,23-79,23-81,23-82, 20-31 102(}..33, 20-37, 20-39, 20-42, 20-46 10
23-84,23-97,24-6,24-10,24-35 20-48,21-7,21-9,21-11, 21-12, 21-14, 21-17
Urban streel, Chapler 16, Chapter 17, 1021-19,21.22,21-23,21-25, 21-26, 22-810
Chapter 18, 19-17, 19-27, 19-49, 19-88, 20-7, 22-10, 22-15, 22-16, 22-18 10 22-20, 22-22 10
20-20,20-46,23-12,23-52,23-57,23-59, 22-24,22-28,22-29,23-2,23-3,23-6,23-8,
23-60,23-81, 24-6, 24-10 23+11,23-16,23-33 to 23-37, 23-43, 23+50,
Urban street fadlity, Chapter 16, 17-2, 17-7, 23-56,23-61,23-69,23-71, 23-73, 23-82,
17-91017-12,17-14,17-24,17-27,17-28, 23-86,23-91. 23-94, 24-4, 24-6 to 24-9, 24-12,
18-2. 18-79, 19-49, 23-59, 23-60 24-14, 24-16 lo 24-18, 24-24 to 24-26, 24-28,
Urban street segment, 16-2. 16-6, 17-2, 24-30 10 24-33
Chapler 18, 19-27, 19-49, 20-7, 20-20, 20-46, Volume b"lance, 18-25, 18-26
23+12 Volume-to-cap"cily r"lio, 16-7, 16-8, 16-13,
U-tum, 20-4, 20-12. 20-14, 20-15, 20-18, 20-19, 16-14, 16-16, 17-37, 18-6, 18-7, 18-13, 18-36
20-211020-23, 2(}..25, 20-27, 20-34, 20-46, 1018-39,18-67,18-71,11:\-73,19-1,19-2,
22-19,23-5,23-6,23-81023-11,23-20,23+24, 19-15,19.18,19-20,19-25,19-29,19-31.
23-65 lo 23-67, 23-70 to 23-80, 23-87, 23-89, 19-50,19-53,19-55 10 19-58, 19-94, 19-95,
23+91, 23-93, 23-94 20-6,20-7,20-48,21-25,21-26,22-9,22-10,
22-22 to 22-24, 22-29, 23-2. 23-3, 23-6, 23-8,

v 23+11, 23-16, 23-28, 23-33 to 23-37, 23-43,

23+50,23+56,23-61, 23-69, 23-71, 23-73,
23-82. 23-86, 23-91, 23-94, 23-96, 24-4, 24-6
Validalion, 23-70, 23-96
Variability, 17-1, 17-710 17-10,17-27,17-38,
18-35,24-2
w
Walk inlerval. 19-36
Walkway, 16-18, 18-44, 18-46, 18-49, 18-52,
19-73, 19-75, 19-77, 24-1, 24-3, 24-4, 24-6,
24-7,24.91024-11,24-13,24-14,24-27,
24-28,24-33
Weaving, 23-66, 23-70, 23-77, 23-79, 23-83,
23-93
Work 1.One,17-1, 17-7 to 17-9, 17-12, 17-18,
17-22 10 17-25, 17-30, 17-37, 18-2, 18-3,.
18-18,18-23,19-2,19-3,19-45,19-48,19-49

y
Yellow eh.mge interval, 19-10 lo 19-13, 19-36,
19-79,19-88
Yellow trap, 19-35
Yidd poinl, 23-53, 23-57

VoIume 3 Index Page V3-12 Volume 3fInterrupted Aow

V~6.0