INTERNATIONAL JOURNAL OF

PROJECT
MANAGEMENT
International Journal of Project Management 24 (2006) 145–155
www.elsevier.com/locate/ijproman

Selecting a delay analysis method in resolving construction claims

David Arditi *, Thanat Pattanakitchamroon
Department of Civil and Architectural Engineering, Illinois Institute of Technology, Construction Engineering and Management Program,
Alumni Memorial Hall, Room 229, 3201 South Dearborn Street, Chicago, IL 60616, USA

Received 14 June 2005; accepted 5 August 2005

Abstract

Delays occur commonly in construction projects. Assessing the impact of delay is sometimes a contentious issue. Several delay anal-
ysis methods are available but no one method can be universally used over another in all situations. The selection of the proper analysis
method depends upon a variety of factors including information available, time of analysis, capabilities of the methodology, and time,
funds and effort allocated to the analysis. The paper reviews 20 research studies that discuss various aspects of delay analysis methods
and summarizes the advantages and disadvantages of widely used delay analysis methods, including the as-planned vs. as-built, impact
as-planned, collapsed as-built, and time impact analysis methods. The paper also discusses the most important issues in delay analysis
that affect the results of the analysis. The selection of a suitable analysis method depends heavily on the availability of scheduling data,
the familiarity of the analyst with the capabilities of the software used in the project, clear specifications in the contract concerning the
treatment of concurrent delays and the ownership of float.
Ó 2005 Elsevier Ltd and IPMA. All rights reserved.

Keywords: Claims management; Delay analysis; CPM; Bar charts

1. Introduction project completion date. A delay may sometimes contrib-

ute to the formation of other delays.
Delays in construction can cause a number of changes in In construction contracts, schedule delay analysis is
a project such as late completion, lost productivity, acceler- commonly conducted to demonstrate cause and effect rela-
ation, increased costs, and contract termination. The party tionships of time-related disputes. Schedule delay analysis
experiencing damages from delay needs to be able to recog- makes use of the as-planned schedule, the as-built schedule,
nize the delays and the parties responsible for them in order and schedule updates. Sometimes, subnetworks or ‘‘frag-
to recover time and cost. However, in general, delay situa- nets’’ are used to present details about delay events. The
tions are complex in nature. A delay in an activity may not schedules may be in the form of bar charts or network dia-
result in the same amount of project delay. A delay caused grams. The network diagram is more effective than the bar
by a party may or may not affect the project completion chart as it shows the logical sequences between activities.
date and may or may not cause damage to another party. CPM schedules add another dimension to schedule analysis
A delay can be caused by more than one party; however, as they provide schedule analysts with a critical path, float
it can also be caused by none of the parties (such as unusu- consumption, and the opportunity of utilizing what-if
ally severe weather conditions). A delay may occur concur- methodology. CPM has long been accepted by courts as
rently with other delays and all of them may impact the an effective tool to evaluate the impact of delays [1].
There are many delay analysis methods that have been
used by researchers and practitioners including the Global
*
Corresponding author. Tel.: +1 312 567 5751; fax: +1 312 567 3519/
Impact Technique which simply plots delays on a bar chart
3540. and determines the global impact of delay by summing up
E-mail address: arditi@iit.edu (D. Arditi). the total duration of all delays. The global impact

0263-7863/$30.00 Ó 2005 Elsevier Ltd and IPMA. All rights reserved.

doi:10.1016/j.ijproman.2005.08.005
146 D. Arditi, T. Pattanakitchamroon / International Journal of Project Management 24 (2006) 145–155

technique is an easy way to assess delay when detailed events [5]. This approach is a method of choice when a
calculations cannot be performed. However, courts and contractor lacks an acceptable schedule during the pro-
arbitration panels resist the use of the global impact tech- ject, or when no as-planned schedule was required in
nique because it wrongly assumes that every delay has an the contract [2].
equal impact on the project duration. The results of delay The time impact method relies on the assumption that
analysis may be influenced by the method selected and delay impacts to a project can be assessed by running a ser-
therefore the selection of the most appropriate method is ies of analyses on schedule updates. Time impact analysis is
of importance to all parties concerned. a procedure that uses CPM principles. It assesses delaysÕ ef-
The objective of the study presented in this paper was to fects on the project schedule by analyzing the schedule peri-
establish selection guidelines for delay analysts by compar- odically, generally on a day-by-day basis [3]. Window
ing the characteristics of the most commonly used delay analysis, a variation of time impact analysis, uses weekly
analysis methods, and considering their advantages and or monthly updates to perform the analysis. Delay events
disadvantages under different circumstances. The paper are inserted into the schedule and delay impacts are accu-
starts out with a brief description of the universally ac- mulated every time the schedule is recalculated.
cepted delay analysis methods, and discusses the issues in-
volved in delay analysis. The criteria that need to be 3. Issues in delay analysis
considered when a party tries to select the most suitable
methods are presented after the views expressed by various A number of factors may influence the result of delay
researchers/practitioners in papers/textbooks published be- analysis regardless of which delay analysis method is used.
tween the years 1987 and 2004 are reviewed. These factors are most of the time specified in contract
clauses and include concurrent delays, float ownership, the-
2. Delay analysis methods ories of critical path, and scheduling software options.
Sometimes, the way these factors are presented in the con-
There are four methods often mentioned in the con- tract favor the ownerÕs interests and contractors are reluc-
struction literature that are professionally acceptable. They tant to follow them. Fair solutions for both parties have
include (1) the as-planned vs. as-built schedule analysis been discussed in the literature.
method, (2) the impact as-planned schedule analysis method,
(3) the collapsed as-built schedule analysis method, and (4) 3.1. Concurrent delays
the time impact analysis method. They are sometimes re-
ferred to by different names in the literature. The basic concept of concurrency was defined by Rubin
The as-planned vs. as-built method is the observation of et al. [6] as the situation in which two or more delays occur
the difference between an as-planned schedule and an as- at the same time, either of which had it occurred alone,
built schedule. The method identifies the as-built critical would have affected the ultimate completion date. The
activities, compares these activities with the activities on events are considered to result in concurrent delay when
the as-planned schedule, assesses the impact of delays it is clear that one of the events caused a delay in the sche-
on the project, identifies the sequences which actually de- dule, but even if that event had not occurred, the schedule
fine the duration of the project, and determines the causa- would have been delayed anyhow by the other event. Some
tion and responsibility of delays that impact project argue whether the two delays are required to occur at the
completion [2]. same time to be considered as concurrent delays. Some de-
The impact as-planned method uses only an as-planned lay events may not start and end exactly at the same time
or baseline schedule for delay analysis. It is based on the but their effects may happen at the same time. SCL [5] sug-
theory that the earliest date by which a project is completed gests using the term ‘‘concurrent effect’’ for sequential de-
can be determined by adding the delays into the as-planned lay events to avoid confusion and treat them differently
schedule. New activities that represent delays, disruptions, from a concurrent delay.
and suspensions are added to the as-planned schedule and Any type of delay, i.e., excusable non-compensable (a
are used to demonstrate the reason why the project was contractor is entitled to an extension of time only), excus-
completed later than planned. Contractors who submit able-compensable (a contractor is entitled to an extension
claims that involve a time extension, add only owner- of time and damages), and non-excusable delays (a con-
caused delays to the as-planned schedule in the appropriate tractor is responsible for the delay) can occur concurrently.
sequence to document the total project delay caused by the The possible combinations of concurrent delays are de-
owner [3]. scribed by Kraiem and Diekmann [7]. Arditi and Robinson
The collapsed as-built method is also referred to as the [8] identified the possible variations of concurrent delays by
‘‘but-for’’ schedule method. This analysis is popular in taking into account the timing of the start of each activity,
claim presentations because it is easily understood by tri- the chronological sequence in which delay types occur in
ers of fact [4]. SCL defines it as a method where the ef- the constituent activities that are concurrently delayed,
fects of delays are ‘‘subtracted’’ from an as-built schedule and the criticality of the activities where concurrent delays
to determine what would have occurred but for those occur.
 D. Arditi, T. Pattanakitchamroon / International Journal of Project Management 24 (2006) 145–155 147

The contentious point of discussion is which remedy to [10] of 46 UK professionals employed by owners, contrac-
which party should be assessed from any concurrent delay tors, and claims consultants shows that the majority of
event [9]. Such delay that affects the overall project dura- respondents believed that the contractor should have exclu-
tion is the responsibility of either the contractor or the sive control of float while only few owners supported the
owner. Even if one party is able to prove the causation position that float should be allocated on a first-come
of the project delay, the assessment of the damage to the first-served basis. de la Garza et al. [11] agree with the Brit-
delay raises further questions of how extended time and ish professionals that float is exclusively for the benefit of
compensation would be assessed in case of the ownerÕs the contractor and add that float should be traded as a
fault, and how liquidated damages would be issued to the commodity. According to de la Garza et al. [11], the con-
contractor in case of delay caused by the contractor. In the- tractor is entitled to sell the float in case the owner needs
ory, the possible results of delay concurrency can be sum- to consume such float. The method to convert the value
marized in Table 1. The question marks in this table of the total float into selling price is provided in their
indicate that the solution is not clearly defined more often article.
than not. When concurrent delays exist, the assessment of Householder and Rutland [12] suggest that the use of
delay damages and/or time extensions is difficult and often float should be reserved for the party who loses or gains
results in serious disagreements. Research is necessary to as a result of fluctuation in the project cost. In other words,
develop agreeable methods that will allow the parties to in fixed-price contracts, wherein the contractor has ulti-
reach consensus by examining the root causes of the indi- mate risk or benefit from project cost, the contractor
vidual concurrent delay events. should exclusively control float usage. Conversely, wherein
the owner has the ultimate risk or benefit from project cost
3.2. Float ownership in a cost-plus contract, the owner should be entitled to own
float to minimize cost to the owner. Ponce de Leon [13]
In a construction project, float is valuable for both con- suggests a compromise position regardless of the type of
tractors and owners. Float provides flexibility to the con- contract: allocating float in a shared way. Activities would
tractorÕs time and budget management. On the other be allocated a percentage of the float available to the path
hand, the owner also needs float to accommodate the im- based on the individual activityÕs duration. If an activity is
pact of change orders on the project. Float ownership is delayed beyond its allocated float, then time extension may
usually specified in the conditions of contract. Contracts be justified to preserve the allocated float of other activities
typically state that float belongs to the project or on a in the approved schedule. Another compromise solution is
‘‘first-come first-served basis’’. In other words, when an offered by Pasiphol and Popescu [14] who propose a qual-
owner-caused delay occurs first and uses up the total float, itative method to distribute total float into each activity
a contractor becomes responsible for contractor-caused de- prior to executing a project. The qualitative factors pro-
lays that put off the project completion date, and that could posed by the authors are subjectively assessed and there-
have been accumulated if the project float had not been ex- fore may be subject to the preparerÕs manipulation.
hausted by the owner. Similarly, when a contractor uses up The varied positions concerning who owns float can
all the float at the beginning of the project, the owner be- influence the result of delay analysis. The sample design/
comes responsible for all delays caused by change orders, build project presented in Fig. 1 is composed of three activ-
a situation that could have been averted had the contractor ities, two of which involve contractor provided design and
not exhausted the total float. construction, and one is ownerÕs approval of design. The
There are a number of suggestions seeking a fair resolu- design needs to be approved by the owner before construc-
tion for the float ownership problem. The UK Delay and tion can start. The contract duration is 12 days while the
Disruption ProtocolÕs position on float ownership is that contractor has planned to execute the project in only 10
if the clause stating the entitlement of float is not specified days, which yields two days of total float. Two alternative
in the contract, float should belong to a project [5]. In other scenarios of actual activities are also shown in Fig. 1. In the
words, the first-come first-served principle should govern first scenario, the contractorÕs design activity experiences a
the use of float. On the other hand, Scott et al.Õs survey 2-day delay, followed by an owner-caused delay of 2 days.

Table 1
Possible remedies for concurrent delays
Delay concurrency Possible remedies
Excusable non-compensable Non-excusable – ! EOT?
Excusable non-compensable Excusable compensable – ! EOT and COMP?
Non-excusable Excusable compensable – ! EOT? and COMP? or LD? or Nil?
Excusable non-compensable Non-excusable Excusable compensable ! EOT? and COMP? or LD? or Nil?
EOT, extension of time to the contractor; Comp, compensation to the contractor; LD, liquidated damages assessed by the owner; NIL, neither contractor,
nor owner recover damages.
148 D. Arditi, T. Pattanakitchamroon / International Journal of Project Management 24 (2006) 145–155

PROJECT DURATION
ACTIVITY (Party Responsible) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Plan Completion date
Design (Contractor)
Approval (Owner)
Construction (Contractor)

Actual Scenario1 Contractor delay

Owner delay
Design (Contractor)
Approval (Owner)
Construction (Contractor)

Actual Scenario 2
Owner delay
Design (Contractor) Contractor delay
Approval (Owner)
Construction (Contractor)

Fig. 1. Effect of ownership of float.

In the second scenario, the contractor-caused delay of 2 on CPM principles. For example, the critical path can be
days occurs after the 2-day delay caused by the owner. defined by the longest path(s) in the network diagram,
The different position of float ownership would influence and by the amount of total float in the different paths.
the results of delay analysis significantly as presented in These two methods become contentious when considered
Table 2. It is therefore imperative that float ownership be in delay analysis because only critical delay should be con-
clearly defined in the contract in order to avoid disagree- sidered when awarding time extensions [15]. The example
ments in delay-related claims. If the party that owns float shown in Fig. 2 presents concurrent contractor and owner
is identified in the contract, both the owner and the con- delays in a situation where the contract time has expired.
tractor can anticipate the outcome of potential actions Under the longest path theory, the owner is responsible
and can adjust their activities accordingly. for 2 days of delay that extended project completion be-
cause only the delay caused by the owner is on the critical
3.3. Scheduling options path. However, under the float theory, a concurrent delay
is justified on day 14 because the contractorÕs activity is
Scheduling issues also influence the result of the analysis also critical since it has one day of negative float caused
because network-based delay analysis substantially relies by the contractorÕs delay. The resolution of the issue of

Table 2
Results of delay analyses
Float ownership Scenario 1 Scenario 2
Extension of time Liquidated damages Extension of time Liquidated damages
granted to assessed to contractor granted to contractor assessed to contractor
contractor in days in days in days in days
Contractor exclusively owns float 2 0 2 0
Owner exclusively owns float 0 2 0 2
Project owns float (first-come, first-served basis) 2 0 0 2
Distribute float to both parties proportionally (duration) 1.6 0.4 1.6 0.4
Ignore float 2 2 2 2

Fig. 2. Longest path vs. float.

 D. Arditi, T. Pattanakitchamroon / International Journal of Project Management 24 (2006) 145–155 149

negative float has a significant impact on the entitlement to tion reflects realistic uses of specific resources governed by
time extensions, compensation for the extended perfor- different working schedules, it makes activities in the same
mance period, and the assessment of liquidated damages. path contain different amounts of float based on their indi-
CPM software packages offer various functions that vidual calendar. A path may contain critical and non-crit-
provide flexibility for scheduling tasks, but some of these ical activities, and as a result, may make the critical path
functions sometimes cause problems when incorporated less definable. It should also be noted that some of the well
into delay analysis. For example, some CPM software established software packages generate inconsistent and
packages such as Primavera Project Planner allow users sometimes incorrect results under certain multiple calendar
to select between retained logic and progress override. The conditions discussed in detail by Kim and de la Garza [17].
progress override mode calculates an updated schedule In addition to duration and logic errors contained in
regardless of the logic relationships of the activities that analyzed data, different uses of scheduling software bring
have started in order to avoid creating unrealistically long about a number of intricate problems in delay analysis.
paths caused by the out-of-sequence activities that would The most mentioned include the use of constraint/manda-
happen in retained logic mode. tory functions that force activities to start or finish by
The delay analysis shown in Fig. 3 is conducted during specific dates, the use of unconventional logic such as
the performance of the activity ‘‘Approval’’ while ‘‘Con- start-to-finish relationships, and the use of long or negative
struction’’ has started out of sequence. The retained logic lag times [16]. These software functions often yield anoma-
mode maintains the logic relationship between ‘‘Approval’’ lous results in delay analysis and should be considered with
and ‘‘Construction’’. The computed start date of the caution by analysts.
remaining portion of ‘‘Construction’’ is still imposed by
the finish date of ‘‘Approval’’ even though ‘‘Construction’’ 4. Comments on delay analysis methods
has actually started out of sequence. As a result, the re-
tained logic mode computes the new completion date as The views of some researchers and practitioners who
day 11 and ‘‘Approval’’ is on the critical path. On the other published about standard delay analysis methods in the
hand, the progress override mode ignores network logic years 1987–2004 are presented in Table 3. Twenty sets of
that has been out of sequence and allows ‘‘Construction’’ comments were obtained from various sources, including
being driven without the predetermined logic. As a result, 19 papers and 3 textbooks about construction claims. These
the progress override mode displays day 10 as the new articles and book chapters discuss how the methods work
completion date and ‘‘Approval’’ is not on the critical path. based on theoretical and practical considerations. They
When delay analysis is conducted, disputes may arise if the may not constitute the entirety of the body of work pub-
contract does not specify which logic mode is to be used. lished in this field but they represent a good cross-section
While the retained logic mode is preferred by some for of what is available. The responses to the methods are var-
updating schedules, others claim that it makes it impossible ied. Some authors recommend using some methods, while
to get an accurate update schedule [16]. others discourage using them, and some others have a neu-
Some software packages allow different calendars to be tral position. Substantial comments are made on advanta-
assigned to each activity and resource. Although this func- ges and disadvantages. Table 3 contains the preferences of

Fig. 3. Retained logic vs. progress override.

 150
Table 3
Comments compiled from the literature (1987–2004)

D. Arditi, T. Pattanakitchamroon / International Journal of Project Management 24 (2006) 145–155

References Delay analysis methods
As-planned vs. as-built method Impact as-planned method Collapsed as-build method Time impact method
Sandlin et al. [28] N/A Spurious results Erroneous evaluation Overcomes some disadvantages
of others
Lovejoy [25] Fair Good Excellent Very good
Sgarlata and Brasco [19] Worthy method N/A Most acceptable by courts Useful for prospective analyses,
but minimal utility supporting claims
Gothand [23] Major drawbacks Major drawbacks Major drawbacks Reliable
SCL [5] Simple, limited Simple, limited Suitable for some situations, Most reliable when available
subjective
Harris and Scott [24] Least popular N/A Fair, most accepted Make some use by claims consultants
Zack [4] Critical flaws Critical flaws Unreliable, easy to manipulate Accurate but expensive
Stumpf [22] Can be challenged Easy to prepare, fundamental flaws Easy to prepare, fundamental flaws Reliable, but time consuming
Fruchtman [2] Reliable Simple, limited No baseline needed, limited Contemporaneous basis, but no
future changes considered
Finke [26,27] N/A Less reflective of actual events Less reflective of actual events Most reasonable and accurate
Zack [21] Unreliable Many flaws, widely discarded Suitable Suitable
McCullough [20] Not acceptable Not acceptable Useful in some situations but easy Dependent on baseline schedule,
to manipulate accurate
Bubshait and Cunningham [30,31] Acceptable, dependent Acceptable, dependent on Acceptable, dependent on availability Acceptable, dependent on
on availability of data availability of data of data availability of data
Levin [1] N/A Simple, consistently rejected by courts Dependent on quality of as-built schedule Dependent on how the method
is applied
Alkass et al. [32] N/A Some major problems Sound, but ignores changes of critical paths Some drawbacks/propose modified
method
Zafar [18] Reliable Fault analysis Fault analysis N/A
Schumacher [33] N/A Potential shortcoming, Overcome some shortcomings Effective method
one-sided analysis
Baram [29] Dependent on Dependent on Most practical in some circumstances Most desirable approach
Wickwire et al. [3] N/A ‘‘Great lie’’ Alive and well Recommended
Bramble and Callahan [34] Acceptable, dependent Acceptable, dependent on Acceptable, dependent on availability N/A
on availability of data availability of data of data
 D. Arditi, T. Pattanakitchamroon / International Journal of Project Management 24 (2006) 145–155 151

the researchers relative to each method, ordered by date of rent delays and acceleration. It also offers ease of use and
publication. Some of the researchers do not specifically pre- flexibility in delay analysis. However, it lacks a systematic
fer one method over the others. They suggest that the selec- procedure to evaluate the impact of delay events individu-
tion of a suitable method should depend on a number of ally. Insufficient methodological support in the literature
criteria and circumstances including those discussed in the results in the as-planned vs. as-built method being least
preceding section. mentioned in the articles.

4.1. As-planned vs. as-built method

4.2. Impact as-planned method
There seems to be consensus that the as-planned vs. as-
The impact as-planned method is the least favored meth-
built method is the simplest form of analysis among the
od among the four methods discussed the articles/books
four methods. But the majority of the researchers listed
listed in Table 3 due to its theoretical flaws. The articles/
in Table 3 have negative opinions of the reliability of this
books listed in Table 3 consistently criticize the impact
method for the very reason that this method is only a sim-
as-planned method, and not one of them recommends the
ple comparison of as-planned and as-built schedules and
use of the impact as-planned method to prove delay impact.
that there is no advanced technique being applied. SCL
Many courts have not accepted this method since the 1990s.
[5] suggests that the as-planned vs. as-built method is useful
One of the reasons is because the method relies only on an
as a starting point in relation to other complex methods of
as-planned schedule to determine the impact of delay. In-
analysis.
deed, this method does not measure the effect of actual work
McCullough [20], Zack [21], Stumpf [22], and Gothand
performed, but relies heavily on the validity of a baseline
[23] do not recommend using the as-planned vs. as-built
schedule. An analysis based on an unrealistic baseline sche-
method because this method simply determines a net im-
dule will not only suffer from faulty logics, but also from
pact of all delay events as a whole rather than scrutinizing
overestimated project durations.
each individual delay event separately. The as-planned vs.
Another reason that undermines the reliability of the im-
as-built method assumes the as-planned and as-built sched-
pact as-planned method is that the contractor, (i.e., the
ules are correct in both activity durations and logic rela-
party that files the claim) inserts only owner-caused delays
tionships even though there are no intermediate updates
into the as-planned schedule to prove the case. Wickwire
available. In addition, the method requires that additional
et al. [3] consider this method to be a ‘‘great lie’’ because
activities be treated separately to make sure that the com-
the analysis fails to incorporate the delay events caused
parison is valid [24]. It is also difficult to make a detailed
by all parties to the contract.
comparison when an as-built schedule has been modified
compared to an as-planned schedule.
On the other hand, Zafar [18] and Fruchtman [2] 4.3. Collapsed as-built method
encourage the use of this method because it takes into ac-
count both as-planned and as-built schedules to evaluate The collapsed as-built method is based on the concept
delay impacts and because it identifies and quantifies both of what-if methodology similar to the impact as-planned
owner and contractor delays. Sgarlata and Brasco [19] re- analysis, but it has evolved to overcome some of the
port that courts and boards have found this a worthy meth- drawbacks of the impact as-planned analysis. The as-
od of analysis. The as-planned vs. as-built analysis method built schedule depicts the factual information concerning
is thought to be capable of addressing concurrency and the work that has been undertaken. Courts and boards
compensability. in the US consider the collapsed as-built method to be
A significant advantage of this method is that the anal- useful because the activities in this method are consistent
ysis requires only existing materials associated with general with actual occurrences on the project [19]. Likewise,
administrative procedures, i.e., as-planned and as-built most UK professionals approve of this method [24]. In
schedules. In a project with less restricted general require- case an as-planned schedule does not exist or is not up-
ments where CPM schedules may not be available, only a dated, an as-built schedule can be initiated from records
bar chart diagram with an as-built schedule or updated re- such as monthly progress reports. The collapsed as-built
cords are sufficient to conduct the analysis. The as-planned method is often selected when reliable schedules cannot
vs. as-built method does not require generating an adjusted be readily obtained from project records or the project
or a newly created network such as in additive or subtrac- does not have scheduling information. The other advan-
tive methods. This prevents analysts from incorporating a tage of this method is that it incurs less time and cost
biased position into the analysis [18]. than time impact analysis. According to Lovejoy [25],
In conclusion, the as-planned vs. as-built analysis relies the collapsed as-built analysis is the most practical ap-
on common sense, a comparison of before-and-after delay proach since it offers a good combination of benefits.
events. The analysis incorporates both as-planned and as- On the other hand, Zafar [18], Finke [26,27], Fruchtman
built schedules, and both contractor and owner delay [2], Stumpf [22], Zack [4], Gothand [23], and Sandlin et al.
events, which supports the ability for recognizing concur- [28] criticize the premise of the collapsed as-built analysis.
152 D. Arditi, T. Pattanakitchamroon / International Journal of Project Management 24 (2006) 145–155

First, when a contractor conducts a collapsed as-built anal- vidually evaluated in detail. Using the CPM algorithm,
ysis, the analyst considers only the delays caused by the the time impact analysis method follows up on the pro-
owner to prove the effects of owner-caused delays on the ject day-by-day from beginning to completion date,
project completion. The analyst does not include contrac- including consumption of float, concurrent delays, recov-
tor-caused delays in the analysis. Therefore, concurrent de- ery time and acceleration or resequencing accurately.
lays cannot be recognized using this delay analysis and this According to SCL [5], this technique is therefore the pre-
constitutes a drawback. This flaw is similar to the short- ferred technique to resolve complex disputes related to
coming observed in the impact as-planned method as the delay and its compensation.
contractorÕs analyst chooses which delays to analyze. Sec- Time impact analysis is distinguished from the impact
ond, the collapsed as-built method does not consider the as-planned and collapsed as-built analyses in the fact that
dynamic nature of the critical path method. Indeed, it as- it incorporates both party delays into the analysis. The
sumes that the as-built schedule makes use of the contrac- excusable compensable, excusable non-compensable, and
torÕs original as-planned intentions to execute the project, non-excusable delays can be separately identified. In addi-
using the same sequence of activities and the same produc- tion to this advantage, Wickwire et al. [3] describe indirect
tivities [4]. Consequently, the events that cause delay along benefits generated by the use of time impact analysis in that
the course of the project may not be detected. Third, the it also provides a disciplined basis for the contractual par-
collapsed as-built analysis is highly subjective and subject ties to keep a project schedule up-to-date and properly
to manipulation. The analyst is required to recreate logic adjusted.
relationships into an as-built schedule from project records Some limitations that exist in some actual construction
in order to perform the CPM analysis. Indeed, an as-built projects may weaken the power of this method. First of
schedule no longer depends on the logic of the original net- all, time impact analysis requires a large amount of infor-
work but on actual dates of activity progress. This process mation in order to perform the analysis. An as-planned
is subjective because the records, including logical se- schedule in CPM format is necessary; additionally, the
quences, lag times, etc., can be subjectively interpreted. schedule needs to be periodically updated. The projects
Both parties are expected to examine the records and agree that lack strict administrative procedures and/or updated
with the interpretation of a recreated as-built network be- schedules are not good candidates for this method. Baram
fore performing the analysis. [29] suggests that the use of time impact analysis is the most
To conclude, the collapsed as-built schedule can deter- desirable approach to handle a delay claim, but only when
mine delay impact in case of limited time and resources data and source documents are available in the required
available for analysis. This method will be useful when format and in the required time frame.
both the contractor and the owner have access to the de- Second, the analysis may not be appropriate when re-
tail of as-built records and reasonably concur in inter- sources or time allowed are limited. As to the detail in-
preting the information used to construct the as-built volved in the methodology, time impact analysis
network. consumes much more time compared to the other methods.
Examining periodic updates is burdensome as actual data
4.4. Time impact analysis method associated with many activities may need to be verified
and compared for every updated period. Added or deleted
Time impact analysis is the most credible delay analysis activities have to be documented. In situations where time
method among the four methods discussed in this paper. and budget are limited, time impact analysis may not be the
The majority of the viewpoints cited in Table 3 agree that method of preference.
the method yields the most reliable results. Time impact Third, the result of the analysis may be influenced by a
analysis does not display the shortcomings of the other variety of factors because time impact analysis is intricate
methods discussed. This approach uses fragnets to analyze as it determines accumulative results from a number of
individual delay events. The durations of the delays and the contemporaneous data. For example, when a window per-
relationships of delays to project activities are reviewed in iod is treated separately from a delay event, the analysis
detail with contemporaneous information. The delay is may require approximation if the delay is divided between
then inserted into the project. This process provides both two window periods [26]. Also the different outcomes
parties an opportunity to scrutinize the delay and reduce between the retained logic and progress override methodol-
disputes. ogies can influence the result of the analysis whenever
Baram [29], Finke [27], Zack [4], and Stumpf [22] ad- on-going schedules in time impact analysis are analyzed
dress the importance of the dynamic nature of project in each window update.
critical paths. Time impact analysis performs a series of In conclusion, time impact analysis is a refined method
analyses throughout a project period, in contrast to the that determines delay impact in construction projects. It
major disadvantage of the previously mentioned methods incorporates contemporaneous data to simulate actual cir-
that observe a schedule at a single point in time. The cumstances at the time the delay occurs and accumulates
analysis is able to trace the causes and effects of delay impacts of delay events by using a series of windows. How-
events systematically. The impact of a delay event is indi- ever, it requires significant time and effort.
 D. Arditi, T. Pattanakitchamroon / International Journal of Project Management 24 (2006) 145–155 153

5. Recommendations for selecting a delay analysis method owner and often updated by the contractor, some are seldom
or never updated. The varied degree of availability of differ-
The delay analysis methods discussed in this paper may ent types of schedules may constrain the selection of the most
or may not be well suited in different situations and their appropriate delay analysis method. While the as-planned vs.
selection depends on four criteria including data require- as-built method can be conducted by making use of simple
ments, time of analysis, capability of methodology, and bar charts, the remaining three methods require that a
time and effort required (see Table 4). CPM schedule be available. While CPM-based analyses
are more detailed and therefore more reliable, a bar chart-
5.1. Availability of information based analysis has the advantage of being simple, practical,
and easily understandable by the parties.
Delay analysis methods determine the impact of delay The time impact method is more complex than the
by using different types of schedules. As-planned and as- remaining three methods because it requires multiple runs
built schedules and updates are commonly required as part that examine the situation one delay at a time or from up-
of administrative procedures, while adjusted schedules and date to update. The availability of update information
fragnets will be created specifically for delay analysis pur- greatly affects the use of this method whereas the remaining
poses. The selection of the most appropriate method partly three methods rely on only one run using the delay infor-
depends on the type of schedule that was used throughout mation at hand.
the project. For example, the impact as-planned analysis
can be used with projects that have only an as-planned 5.2. Time of analysis
schedule and no updates. On the other hand, the collapsed
as-built method is applicable in situations where little Delay analysis can be performed at different points in
scheduling information is available since it is always possi- time relative to a delay. Foresight analysis is conducted to
ble to create an as-built schedule at the end of the project present the effects of potential delays caused by proposed
by making use of commonly available project records. events such as future change orders. For delays that have
Some small projects may be managed by simple bar charts occurred, the best time to evaluate them is when they actu-
whereas larger projects may make use of network schedules. ally arise [20]. Hindsight analysis is a common practice in
While some of these network schedules are approved by the construction due to the hesitation of project parties to take
Table 4
Comparison of delay analysis methods
As-planned vs. as-built Impact as-planned Collapsed as-built Time-impact
analysis analysis analysis analysis
Availability of information
Type of schedule
As-planned schedule X X X
As-built schedule X X X
Updated schedules X
Adjusted schedules X X X
Fragnets Depends Depends X
Type of information
No CPM (bar chart) X
No CPM but progress record X
CPM approved/not updated X X
CPM approved/updated X X X X
Time of analysis
Foresight X
Real time X X
Hindsight, during performance period X X X X
Hindsight, after project completion X X X X
Capabilities
Float consumption/critical path X/depends X X X
Time extension X X X X
Compensation Depends Depends Depends X
Concurrent delay X X
Resequencing Depends X X
Dynamic nature of CPM X
Acceleration X X
Time-cost-effort
Type of analysis Observative Additive Subtractive Additive
Level of effort Low Low Moderate High
154 D. Arditi, T. Pattanakitchamroon / International Journal of Project Management 24 (2006) 145–155

action at the time the delay began. Because of their method- and forth over the course of the project, using a series of
ological characteristics, some delay analysis methods can- analyses (such as in time impact analysis) can demon-
not be used to perform foresight or hindsight analyses at strate the collective cause and effect of delays and pro-
wish. For instance, the collapsed as-built method cannot vide an accurate result. On the other hand, an analyst
be used to assess the delay impact caused by a potential should avoid using delay analysis methods that consider
change order because the method relies on as-built informa- delays caused by only one party (e.g., impact as planned
tion which does not exist at that time. Impact as-planned or analysis conducted by contractor to prove the impact of
time impact analyses may be used in these circumstances. only owner-caused delays) if concurrent delays are
numerous throughout the project.
5.3. Capabilities of the methodology When a contractor seeks compensation for constructive
acceleration, the contractor is not only required to prove
The outcome of delay analysis is expected to be reliable that the excusable delay exists, but also that the contractor
in that it should represent the actual events that took place actually improved its performance relative to the plan and
throughout the project. But in situations where a project incurred additional cost. The delay analysis method se-
schedule is composed of a complex network and contains lected has to recognize schedule acceleration. A contractor
numerous delays, different delay analysis methods may would benefit from using time impact analysis to examine
yield different results. Stumpf [22], Bubshait and Cunning- the impact of acceleration claims because this method al-
ham [30], and Alkass et al. [32] have tested delay analysis lows the analyst to assess the relationship between delays
methods and found that owner-caused delays had a higher and corresponding accelerations. The delay analysis meth-
impact on project completion when they were analyzed od that uses only the as-planned or only the as-built sche-
using the time impact method. But this finding cannot be dule may not represent acceleration properly.
generalized because it appears that the outcome of the In some cases, schedule disruption does not obviously
methods depend on the individual effect of each type of de- interrupt the schedule but may cause a decline in working
lay, the effect of concurrent delays, scheduling changes dur- resourcesÕ efficiency. For example, if a disruption forces
ing construction, and the effects of delays at different points an activity to stop and start over and over, workers may
in time during performance. experience lost productivity caused by a slowdown in the
Some analysis methods may sometimes be misleading. ‘‘learning effect’’. This kind of delay is unlikely to be de-
For example, activities that have a great deal of float in tected and evaluated by schedule analysis. It is necessary
the early portion of a project may end up on the critical for an analyst to incorporate productivity observation in
path during the course of the project and then be returned the analysis in order to prove cause and effect of delay.
to a non-critical status with float time at the end of the pro-
ject [4]. As a result, some experts challenge the reliability of 5.4. Time and funds available for analysis
the single-run schedule analysis methods, including the as-
planned vs. as-built, impact as-planned, and collapsed as- The as-planned vs. as-built method is the simplest of the
built methods since they fail to consider the critical paths methods because it uses an ‘‘observative’’ approach that
at different times during the project. Some researchers crit- simply compares the as-planned against the as-built sche-
icize the method that uses only one schedule at a single dule, whereas the remaining three methods have a basic
point in time [22,23,27]. A mere as-planned schedule may methodology that is either ‘‘additive’’ or ‘‘subtractive’’
not actually reflect contractor intent to execute the sche- depending on whether delays are added to an as-planned
dule. The contractor may thereafter have performed more schedule or subtracted from an as-built schedule. These last
or less productively than planned, and sequences of work three methods are based on a ‘‘what-if’’ concept that allows
may have been readjusted due to unforeseen changes. An assessing the impact of each individual delay on the final
analyst should therefore avoid a delay analysis method that project duration and may involve greater or lesser effort
uses only an as-planned schedule when the schedule experi- depending on the method used.
enced drastic changes and resequencing because such a The most sophisticated delay analysis method (i.e., time
method does not faithfully represent the impact of the de- impact analysis) may consume much time and may incur
lay events in such circumstances. Conversely, a mere as- high cost compared to simpler methods such as the as-
built schedule assumes that there was a single unchanging planned vs. as-built, impact as-planned, and collapsed
critical path throughout the life of a project, which is unli- as-built methods. It also requires specialized expertise to
kely to happen in construction projects. perform the analysis. The selection of the analysis method
A preliminary observation of delay situations may depends on the degree of detail and accuracy that the ana-
help an analyst to select the proper analysis method. In lyst compromises given the time constraints and budget
a schedule where numerous concurrent delays exist, the limitations. Any delay analysis method should produce
analysis method that incorporates both owner and con- the same result when evaluating a simple non-concurrent
tractor delays (such as as-planned vs. as-built) should delay event. In this sort of situation, it is not necessary
be able to better show a clear picture of the concurren- for an analyst to spend unnecessary time, money, and effort
cies. In a schedule where critical paths have shifted back on an elaborate time impact analysis.
 D. Arditi, T. Pattanakitchamroon / International Journal of Project Management 24 (2006) 145–155 155

6. Conclusion [5] SCL – Society of Construction Law. Delay and disruption protocol;
2002. Available from: www.eotprotocol.com.
[6] Rubin RA, Guy SD, Maevis AC, Fairweather V. Delay analysis
A fair and effective evaluation of delay impact is possi- construction claims analysis, presentation, defense. New York
ble if the most appropriate delay analysis method is se- (NY): Van Nostrand Reinhold; 1983.
lected that provides a reliable solution with the [7] Kraiem Z, Diekmann J. Concurrent delay in construction projects. J
information available and within the time and cost allo- Constr Eng Manage ASCE 1987;113(4):591–602.
cated for this purpose. Time impact analysis is clearly [8] Arditi D, Robinson M. Concurrent delays in construction litigation.
Cost Eng AACE 1995;37(7):20–31.
accepted by the literature cited in Table 3 as the most reli- [9] Ponce de Leon G. Theories of concurrent delays. Trans AACE Int
able delay analysis method among the four standard meth- 1987:CDH.6.1–CDH.6.5.
ods discussed in this paper. However, the transient nature [10] Scott S, Harris R, Greenwood D. Assessing the new U.K. protocol
of construction projects not often allowing scheduling data for dealing with delay and disruption. J Constr Eng Manage ASCE
being well documented as well as time and budget limita- 2004;130(1):50–9.
[11] de la Garza JM, Vorster MC, Parvin CM. Total float traded as
tions lead a number of researchers to suggest that the commodity. J Constr Eng Manage ASCE 1991;117(4):716–27.
choice of a simpler method may be sensible. The as- [12] Householder JL, Rutland HE. Who owns float? J Constr Eng
planned vs. as-built and collapsed as-built methods are effi- Manage ASCE 1990;116(1):130–3.
cient in some situations. The impact as-planned method [13] Ponce de Leon G. Float ownership – some recommendations.
falls behind these three as it has so many critical flaws. Stratagem 1982;1(1).
[14] Pasiphol S, Popescu C. Qualitative criteria combination for total float
The reliability of delay analysis depends on the selec- distribution. Trans AACE Int 1994:DCL.3.1–DCL.3.6.
tion of a suitable analysis method and on the availability [15] Jentzen GH, Spittler P, Ponce de Leon G. Responsibility for delay
of scheduling data. An analyst should meticulously re- after the expiration of the contract time. Trans AACE Int
view the data obtained from the project records because 1994:DCL.9.1–DCL.9.7.
none of the methods yields reliable results if the informa- [16] Korman R, Daniels SH. Critics canÕt find the logic in many of todayÕs
CPM schedules. PMI, College of Scheduling; 2003. Available from:
tion used is invalid. It is necessary to be very familiar http://www.pmicos.org/fse.asp.
with the capabilities of the software used in project [17] Kim K, de la Garza JM. Critical path method with multiple
scheduling and progress control in order to be able to calendars. J Constr Eng Manage ASCE 2005;131(3):330–42.
generate legitimate schedules for the analysis. The other [18] Zafar QZ. Construction project delay analysis. Cost Eng AACE
controversial issues such as the treatment of concurrent 1996;38(3):23–7.
[19] Sgarlata MA, Brasco CJ. Successful claims resolution through an
delays and the ownership of float should be clearly spec- understanding of the law governing allocation of risk for delay and
ified in the contract. Project participants should settle disruption. CM ejournal, CMAA 2004. Available from: http://
these issues early in the project so as to maintain proper cmaanet.org/ejournal.php .
scheduling administration. [20] McCullough RB. CPM schedules in construction claims from the
The best practice is to be prepared for delay manage- contractorÕs perspective. Trans AACE Int 1999: CDR.2.1–CDR.2.4.
[21] Zack Jr J. Pacing delay – the practical effect. Trans AACE Int
ment throughout the project by adopting these recommen- 1999:CDR.1.1–CDR.1.6.
dations as a routine procedure. Since time impact analysis [22] Stumpf GR. Schedule delay analysis. Cost Eng AACE 2000;
is by far the most effective method in proving time-based 42(7):32–43.
claims for the reasons discussed in this paper, ideally [23] Gothand KD. Schedule delay analysis: modified windows approach.
speaking, all project managers should engage in practices Cost Eng AACE 2003;45(9):18–23.
[24] Harris RA, Scott S. UK practice in dealing with claims for delay. Eng
that will generate adequate information to allow the use Constr Architect Manage 2001;8(5–6):317–24.
of time-impact analysis in case a time-based claim needs [25] Lovejoy VA. Claims schedule development and analysis: collapsed as-
to be proven. However, given the different circumstances built scheduling for beginners. Cost Eng AACE 2004; 46(1):27–30.
in different projects, it is not always possible to generate [26] Finke MR. Window analyses of compensable delays. J Constr Eng
such information. In such cases, the recommendations Manage ASCE 1999;125(2):96–100.
[27] Finke MR. Contemporaneous analysis of excusable delays. Cost Eng
made in this paper should allow a claims analyst to pick AACE 1997;39(12):26–31.
the most effective delay analysis method that is compatible [28] Sandlin LS, Sapple JR, Gautreaux RM. Phased root cause analysis –
with the information at hand at the time of analysis. a distinctive view on construction claims. Cost Eng AACE 2004;
46(6):16–20.
[29] Baram GE. Delay analysis – issue not for granted. Trans AACE Int
1994:DCL.5.1–DCL.5.9.
References [30] Bubshait AA, Cunningham MJ. Comparison of delay analysis meth-
odologies. J Constr Eng Manage ASCE 1998;124(4):315–22.
[1] Levin P. Construction contract claims, changes and dispute resolu- [31] Bubshait AA, Cunningham MJ. Determining schedule impact:
tion. 2nd ed. New York (NY): ASCE Press; 1998. working practice. Practice Periodical Struct Des Constr ASCE
[2] Fruchtman E. Delay analysis – eliminating the smoke and mirrors. 1998;3(4):176–9.
Trans AACE Int 2000:CDR.6.1–CDR.6.4. [32] Alkass S, Mazerolle M, Harris F. Construction delay analysis
[3] Wickwire J, Driscoll T, Hurlbut S. Construction scheduling prepara- techniques. Constr Manage Econ 1996;14(5):375–94.
tion, liability, and claims. New York (NY): Wiley Law Publications; [33] Schumacher L. Quantifying and apportioning delay on construc-
1991. tion projects. Cost Engineering AACE 1995;37(2):11–3.
[4] Zack Jr J. But-for schedules – analysis and defense. Cost Eng AACE [34] Bramble BB, Callahan MT. Construction delay claims. New York
2001;43(8):13–7. (NY): Wiley Law Publications; 1987.