United States

Department of Part 630

Agriculture

Natural
National Engineering Handbook
Resources
Conservation
Service

Chapter 19 Transmission Losses

Rain clouds
Cloud formation

Precipitation

n
iratio
Evaporation
n

ean
tio

Transp
s o il
ta

from oc
ge
ve

fr o m
m

n
fro

ti o

s
am
n

ir a
ti o

tr e
sp
ra
po

an

ms
va

Tr

Surfa E
fr o

ce ru
noff

Infiltration

Soil
Percolation

Rock Ocean
Ground water
Deep percolation

(210VINEH, August 2007)

Chapter 19 Transmission Losses Part 630
National Engineering Handbook

Issued August 2007

The U.S. Department of Agriculture (USDA) prohibits discrimination in all

its programs and activities on the basis of race, color, national origin, age,
disability, and where applicable, sex, marital status, familial status, parental
status, religion, sexual orientation, genetic information, political beliefs, re-
prisal, or because all or a part of an individuals income is derived from any
public assistance program. (Not all prohibited bases apply to all programs.)
Persons with disabilities who require alternative means for communication
of program information (Braille, large print, audiotape, etc.) should con-
tact USDAs TARGET Center at (202) 7202600 (voice and TDD). To le a
complaint of discrimination, write to USDA, Director, Ofce of Civil Rights,
1400 Independence Avenue, SW., Washington, DC 202509410, or call (800)
7953272 (voice) or (202) 7206382 (TDD). USDA is an equal opportunity
provider and employer.

(210VINEH, August 2007)

Acknowledgments

Chapter 19 was originally prepared by Victor Mockus (retired) in 1956 and

revised in 1972. In 1983, it was revised by Leonard J. Lane, Agriculture Re-
search Service (ARS), Tucson, Arizona, in cooperation with George Comer
(retired), Gary Conaway (retired), Herman McGill (retired), and Harry
Millsaps (retired) of the Soil Conservation Service; and Virginia Ferreira
(retired), Edward Shirley (retired), and Delmer Wallace (retired) of the
Agriculture Research Service. This version was prepared by the Natural
Resources Conservation Service under guidance of Donald E. Woodward,
(retired), Washington, DC, and reviewed and edited by Helen Fox Moody,
hydraulic engineer, Beltsville, Maryland.

(210VINEH, August 2007) 19i

Chapter 19 Transmission Losses

Contents: 630.1900 Introduction 191

630.1901 Assumptions and limitations 191

(a) Assumptions..................................................................................................191
(b) Limitations ....................................................................................................191

630.1902 Symbols and notation 192

630.1903 Applications 193

(a) Summary of procedure ................................................................................193
(b) Estimating parameters from observed inow-outow data ...................194
(c) Estimating parameters in the absence of observed inow- ....................195
outow data
(d) Summary of parameter estimation techniques .........................................196

630.1904 Examples 196

191 No lateral inow or out-of-bank ow ...................................................197
192 Uniform lateral inow ..........................................................................1910
193 Approximations for out-of-bank ow .................................................1911
194 Transmission losses limited by available storage .............................1914

630.1905 Summary 1916

630.1906 References 1917

Appendices
19A Derivation of Procedures for Estimating Transmission Losses ..... 19A1
When Observed Data are Available
19B Analysis of Selected Data Used to Develop the Procedure ........... 19B1
for Estimating Transmission Losses
19C Estimating Transmission Losses When No Observed Data ........... 19C1
are Available

(210VINEH, August 2007) 19iii

Chapter 19 Transmission Losses Part 630
National Engineering Handbook

Tables Table 191 Relationships between bed material characteristics 195

and parameters for a unit channel

Table 192 Procedures to use when observed inow-outow data 196

are available

Table 193 Procedures to use when no observed inow-outow 196

data are available

Table 194 Outline of examples and comments on their 1916

applications

Table 19B1 Hydrologic data used in analyzing transmission losses 19B1

Table 19B2 Parameters for regression model and differential 19B2

equation model for selected channel reaches

Table 19B3 Unit length, unit width, and unit length and width 19B3
parameters for selected channel reaches

Table 19C1 Data for analysis of relations between effective 19C2

hydraulic conductivity and model parameters

Table 19C2 Auxiliary transmission loss data for selected 19C3

ephemeral streams in southern Arizona

Table 19C3 Range of seepage rates in unlined canals 19C4

Figures Figure 19A1 Observed versus computed peak discharge of the 19A3
outow hydrograph

Figure 19C1 Relation between KD and regression intercept 19C2

Figure 19C2 Relation between KD/P and decay factor 19C2

19iv (210VINEH, August 2007)

Chapter 19 Transmission Losses

630.1900 Introduction 630.1901 Assumptions and

limitations
Natural stream channels in arid and semiarid regions
are generally ephemeral. Flow is occasional and fol- (a) Assumptions
lows storms, which are infrequent. When ows occur
in normally dry stream channels, the volume of ow The methods described in this chapter are based on
is reduced by inltration into the bed, the banks, and the following assumptions:
possibly the ood plain. These losses to inltration,
called transmission losses, reduce not only the volume Water is lost in the channel; no streams gain
of the hydrograph, but also the peak discharge. water.
Inltration characteristics and other channel
This chapter describes a procedure for estimating properties are uniform with distance and width.
the volume of runoff and peak discharge for ephem-
eral streams; it can be used with or without observed Sediment concentration, temperature, and ante-
inow-outow data. If available, observed inow-out- cedent ow affect transmission losses, but the
ow data can be used to derive regression equations equations represent the average conditions.
for the particular channel reach. Procedures based The channel reach is short enough that an aver-
on the derived regression equations enable a user to age width and an average duration represent the
determine prediction equations for similar channels of width and duration of ow for the entire channel
arbitrary length and width. reach.
Chapter 19 also gives the procedures for estimating Once a threshold volume has been satised, out-
parameters of the prediction equations in the absence ow volumes are linearly proportional to inow
of observed inow-outow data. These procedures are volumes.
based on characteristics of the bed and bank material. Once an average loss rate is subtracted and the
Approximations for lateral inow and out-of-bank ow inow volume exceeds the threshold volume,
are also presented. peak rates of outow are linearly proportional to
peak rates of inow. Moreover, the rate of change
in outow peak discharge with changing inow
peak discharge is the same as the rate of change
in outow volume with changing inow volume.
Lateral inow can be either lumped at points of
tributary inow or uniform with distance along
the channel.
For volume and peak discharge calculations, lat-
eral inow is assumed to occur during the same
time as the upstream inow.

(b) Limitations
The main limitations of the procedures are:
Hydrographs are not specically routed along the
stream channels; predictions are made for vol-
ume and peak discharge.
Peak ow equations do not consider storage at-
tenuation effects or steepening of the hydrograph
rise.

(210VINEH, August 2007) 191

Chapter 19 Transmission Losses Part 630
National Engineering Handbook

Analyses on which the procedures are based

represent average conditions or overall trends. 630.1902 Symbols and notation
Inuences of antecedent ow and sediment
concentration in the streamow have not been Upstream inow
quantied.
P = inow volume (acre-feet)
Estimates of effective hydraulic conductivity in p = peak rate of inow (cubic feet per second)
the streambed are empirically based and repre-
sent average rates. Lateral inow
Peak discharge of outow is decreased by the QL = lateral inow volume (acre-feet per mile)
average loss rate for the duration of ow. qL = peak rate of lateral inow (cubic feet per second
per foot)
Procedures for out-of-bank ow are based on the
assumption of a weighted average for the effec- Outow
tive hydraulic conductivity.
Q(x,w) = outow volume (acre-feet)
q(x,w) = peak rate of outow (cubic feet per second)

Channel reach
D = duration of ow (hours)
K = effective hydraulic conductivity (inches per hour)
V = total available storage volume of alluvium in the
channel reach (acre-feet)
w = average width of ow (feet)
x = length of reach (miles)

Prediction equations (parameters)

a = regression intercept for unit channel
(acre-feet)
a(D) = regression intercept for unit channel with a
ow of duration D (acre-feet)
a(x,w) = regression intercept for a channel reach of
length x and width w (acre-feet)
b = regression slope for unit channel
b(x,w) = regression slope for a channel reach of
length x and width w
k = decay factor (foot-miles)-1
k(D,P) = decay factor for unit channel with a ow
duration D and volume P (foot-miles)-1
PO = threshold volume or amount of inchannel
loss above which channel outow occurs
for a unit channel (acre-feet). Channel
outow is 0.0 until the threshold volume is
achieved.
PO(x,w) = threshold volume or amount of inchannel
loss above which channel outow occurs
for a channel reach of length x and width w
(acre-feet). Channel outow is 0.0 until the
threshold volume is achieved.

192 (210VINEH, August 2007)

Chapter 19 Transmission Losses Part 630
National Engineering Handbook

converts from acre-feet per hour to cubic feet per

630.1903 Applications second.

If lateral inow is uniform, the volume equation is that

The simplied procedures are summarized here; shown in equation 194 at the bottom of the page.
additional details and derivations are given in the
appendices. Methods have been developed for two The corresponding equation for peak discharge is
situationswhen observed inow-outow data are shown in equation 195 at the bottom of this page.
available and when no observed data are available. In this equation the factor 5,280 converts cubic feet
per second per foot to cubic feet per second per mile.
(a) Summary of procedure Derivations and background information are in appen-
dix 19A.
The prediction equation for outow volume, without
lateral inow, is For a channel reach with only tributary lateral inow,
equations 191 and 193 would be applied on the
P Po ( x, w )
tributary channel and the main channel to the point
0
of tributary inow. Then the sum of the outows from
Q ( x, w ) = these two channel reaches would be the inow to the
a ( x, w ) + b ( x, w ) P P > Po ( x, w ) lower reach of the main channel.

(eq. 191) The procedures described by equations 191, 193,
where the threshold volume is 194, and 195 require that the upstream inow vol-
umes and lateral inow volumes along the channel
( )
a x , w reach be estimated using the procedures described in
(
Po x, w = ) b( x, w ) (eq. 192)
National Engineering Handbook, part 630 (NEH 630),
chapter 10. Peak ow rates and ow durations are
estimated by use of procedures described in NEH 630,
The corresponding equation for peak discharge is chapter 16.
shown in equation 193 below. In this equation, 12.1

0 Q ( x, w ) = 0

q ( x, w ) = 12.1

D
{ }
a ( x, w ) 1 b ( x, w ) P + b ( x, w ) p Q ( x, w ) > 0
(eq. 193)

0 P Po ( x, w )


Q ( x, w ) = (eq. 194)
QL
a ( x, w ) + b ( x, w ) P + kw 1 b ( x, w ) P > Po ( x, w )

0 Q ( x, w ) = 0


q ( x, w ) = (eq. 195)
12.1 q L (5, 2800 )
{ }
D a ( x, w ) 1 b ( x, w ) P + b ( x, w ) p + kw
1 b ( x, w ) Q ( x, w ) > 0

(210VINEH, August 2007) 193

Chapter 19 Transmission Losses Part 630
National Engineering Handbook

(b) Estimating parameters from observed ticular attention to any data points far from the
inow-outow data regression line, especially those points that may
be strongly inuencing the slope or intercept.
If a channel reach has an assumed length x and aver- 4. Correct data points that are in error; remove
age width w, then n observations on Pi and Qi (without points that are not representative.
lateral inow) can be used to estimate the parameters
in equation 191. Parameters of the linear regression 5. Recompute the regression slope and intercept
equation can be estimated as using equations 196 to 199 and the corrected
data.
(Qi Q )( Pi P )
n

b( x, w ) = i =1
(eq. 196) A great deal of care and engineering judgment must be
( )
n 2 exercised in nding and eliminating errors from the set
Pi P
and i =1 of observed inow-outow observations.

a( x, w ) = Q b( x, w )P (eq. 197) (1) Unit channels

A unit channel is dened as a channel of length x =
where: 1 mile and width w = 1 foot. Parameters for the unit
Q = mean outow volume channel are required to compute parameters for chan-
P = mean inow volume nel reaches with arbitrary length and width. The unit
n = number of observations on Pi and Qi channel parameters are computed by the following
equations:
Alternative formulas recommended for computation ln b( x, w ) (eq. 1910)
are k=
xw
n n n
n Pi Qi Pi Qi b = e k (eq. 1911)

(Qi Q )( Pi P ) =
n
i =1 i =1 i =1
i =1 n a( x, w )(1 b)
(eq. 198) a=
[1 b(x, w )] (eq. 1912)

and
2 where a(x,w) and b(x,w) are the regression param-
n n
n Pi2 Pi eters derived from the observed data. In this case the
i =1 (eq. 199)
( Pi P )
n 2
= i =1 length x and width w are xed known values. Par-
i =1 n ticular care must be taken to maintain the maximum
number of signicant digits in determining k, b, and a.
Linear regression procedures are available on most Otherwise, signicant round-off errors can result.
computer systems and on many handheld calculators.
(2) Reaches of arbitrary length and width
Constraints on the parameters are Given parameters for a unit channel, parameters for a
channel reach of arbitrary length x and arbitrary width
a( x, w ) < 0 and 0 b( x, w ) 1 w are computed by the following equations:

When one or both of the constraints are not met, the b( x, w ) = e kxw (eq. 1913)
following procedure is suggested:
1. Plot the observed data on rectangular coordinate a( x, w ) =
a
1 b
[
1 b( x, w ) ] (eq. 1914)
paper: Pi on the X-axis and Qi on the Y-axis.
( )
a x , w
2. Plot the derived regression equation on the graph
with the data.
(
Po x, w = ) b( x, w )
(eq. 192)

3. Check the data for errors (such as events with

lateral inow or computational errors). Pay par-

194 (210VINEH, August 2007)

Chapter 19 Transmission Losses Part 630
National Engineering Handbook

(c) Estimating parameters in the absence and for the decay factor on ungaged reaches
of observed inow-outow data
KD
k ( D, P ) = 1.09 ln 1.0 0.00545 (eq. 1916)
When inow-outow data are not available, an esti- P
mate of effective hydraulic conductivity is needed to
predict transmission losses. Effective hydraulic con- Given values of a and k from equations 1915 and
ductivity, K, is the inltration rate averaged over the 1916, equations 1913, 1914, and 192 are used to
total area wetted by the ow and over the total dura- compute parameters for a particular x and w. Derived
tion of ow. Because effective hydraulic conductivity relationships between bed material characteristics,
represents a space-time average inltration rate, it effective hydraulic conductivity, and the unit channel
incorporates the inuence of temperature, sediment parameters a and k are shown in table 191. These
concentration, ow irregularities, errors in the data, data can be used to estimate parameters for ungaged
and variations in wetted area. For this reason it is not channel reaches.
the same as the saturated hydraulic conductivity for
clear water under steady-state conditions. Analysis of
observed data results in equations for the unit channel
intercept

a ( D) = 0.00465KD (eq. 1915)

Table 191 Relationships between bed material characteristics and parameters for a unit channel (average moisture condi-
tions)

Bed material group Bed material characteristics Effective - - - - - - - - - Unit channel parameters - - - - - - - - -
hydraulic Intercept 2/ Decay factor 3/
conductivity 1/ a k
K (in/h) (acre-ft) (ft-mi)1

1: Very high loss rate Very clean gravel and large >5 < 0.023 >0.030
sand
2: High loss rate Clean sand and gravel, eld 2.0 to 5.0 0.0093 to 0.023 0.0120 to 0.030
conditions
3: Moderately high loss Sand and gravel mixture with 1.0 to 3.0 0.0047 to 0.014 0.0060 to 0.018
rate low silt-clay content
4: Moderate loss rate Sand and gravel mixture with 0.25 to 1.0 0.0012 to 0.0047 0.0015 to 0.0060
high silt-clay content
5: Insignicant to low Consolidated bed material; 0.001 to 0.10 5 x 10-6 to 5 x 10-4 6 x 10-6 to 6 x 10-4
loss rate high silt-clay content
1/ See appendix 19C for sources of basic data.
2/ Values are for unit duration, D = 1 hour. For other durations, a(D) = 0.00465KD.
3/ Values are for unit duration and volume, D/P = 1. For other durations and volumes, use:

k ( D, P ) = 1.09 ln 1.0 0.00545
KD
P

(210VINEH, August 2007) 195

Chapter 19 Transmission Losses Part 630
National Engineering Handbook

(d) Summary of parameter estimation

techniques 630.1904 Examples
Suggested procedures for use when observed data are
available are summarized in table 192. Procedures for The following examples illustrate application of the
use on ungaged channel reaches are summarized in procedures for several cases under a variety of circum-
table 193. Again, whatever procedure is used, the pa- stances. As in any analysis, all possible combinations
rameter estimates must satisfy the constraints a(x,w) of circumstances are impossible to consider, but the
< 0 and 0 b(x,w) 1. examples presented here should provide an overview
of useful applications of the procedures. Use of these
procedures requires judgment and experience. At each
step of the process, care should be taken to ensure
that the results are reasonable and consistent with
sound engineering practice.

Example 191 illustrates application of the procedures

with and without observed data when ow is within
the channel banks and there is no lateral inow. Ex-
ample 192 is for the same channel reach, but is based
on assumption of uniform lateral inow between the
inow and outow stations. Approximations for out-
of-bank ow are described in example 193.

Table 192 Procedures to use when observed inow-outow data are available

Step Source Result

1. Perform regression analysis Eqs. 196, 197, 192 Prediction equations for the particular reach
2. Derive unit channel parameters Eqs. 1910 to 1912 Unit channel parameters
3. Calculate parameters Eqs. 1913, 1914, 192 Parameters of the prediction equations for
arbitrary x and w

Table 193 Procedures to use when no observed inow-outow data are available

Step Source Result

1. Estimate inow Hydrologic analysis Mean duration of ow, D, and volume of inow P
2. Identify bed material Table 191 Effective hydraulic conductivity, K
3. Derive unit channel parameters Eqs. 1915, 1916, 1911 Unit channel parameters
4. Calculate parameters Eqs. 1913, 1914, 192 Parameters of the prediction equations for
arbitrary x and w

196 (210VINEH, August 2007)

Chapter 19 Transmission Losses Part 630
National Engineering Handbook

Example 191 No lateral inow or out-of-bank ow

Given: A channel reach of length x = 5.0 miles, average width w = 70 feet.

Bed material consists of sand and gravel with a small percentage of silt and clay.
Assume a mean ow duration D = 4 hours and a mean inow volume of P = 34 acre-feet.

Find: The prediction equations for the channel reach. Estimate the outow volume and peak for an
inow P = 50 acre-feet and p = 1,000 cubic feet per second.

Case 1 Observed inow-outow data

- - - - - Observed inow-outow data (acre-feet) - - - - -

Pi 20.0 100.0 25.0 10.0 15.0 P = 34

Qi 6.0 75.0 9.0 0.1 2.5 Q = 18.52

Solution: Follow the procedure outlined in table 192, step 1, for x = 5.0 miles and w = 70 feet.

b( x, w ) =
( )(
Qi Q Pi P ) = 0.850
( )
2
Pi P

a ( x, w ) = Q b ( x, w ) P
= 18.52 0.850 ( 34 ) = 10.38 acre-ft

a ( x, w ) 10.38
Po ( x, w ) = = = 12.21 acre-ft
b ( x, w ) 0.850
Substituting these values in equation 191, the prediction equation for volume is

0 P 12.21

Q ( x, w ) =
10.38 + 0.850P P > 12.21

and the prediction equation (from equation 193) for peak discharge is

0 Q ( x, w ) = 0

q ( x, w ) =
31.4 0.454P + 0.850 p Q ( x, w ) > 0

For an inow volume P = 50 acre-feet and an inow peak rate p = 1,000 cubic feet per second,
the predicted outow volume is
Q(x,w) = 10.38 + 0.850(50) = 32.1 acre-ft

(210VINEH, August 2007) 197

Chapter 19 Transmission Losses Part 630
National Engineering Handbook

Example 191 No lateral inow or out-of-bank owContinued

and the predicted peak rate of outow is

q ( x, w ) = 31.4 0.454 (50 ) + 0.850 (1, 000 )

= 796 ft 3 /s

Case 2 No observed inow-outow data

Solution: Follow the procedures outlined in table 193. From table 191, estimate K = 1.0 inch per hour,
with D = 4.0 hour, P = 34 acre-feet, so:

a = 0.00465KD = 0.01860 acre-ft

KD
k = 1.09 ln 1.0 0.00545
P
= 0.000699 ( ft-mi )
1

and
b = ek = e0.000699 = 0.999301

are the unit channel parameters. From equations 1913, 1914, and 192, the parameters for the
given reach with x = 5.0 miles and w = 70 feet are
( )( )( )
b( x, w ) = e kxw = e 0.000699 5.0 70
= 0.783
a
a ( x, w ) = 1 b ( x, w )
1b
0.01860
= (1 0.783 )
(1 0.999301)
= 5.788 acre-ft

and
a ( x, w )
Po ( x, w ) =
b ( x, w )
( 5.78 )
= = 7.38 acre-ft
0.783
The prediction equation for the volume is
0 P 7.38

Q ( x, w ) =
5.78 + 0.783P P > 7.38

198 (210VINEH, August 2007)

Chapter 19 Transmission Losses Part 630
National Engineering Handbook

Example 191 No lateral inow or out-of-bank owContinued

and the prediction equation for peak discharge is

0 Q( x , w ) = 0

q( x, w ) =

17.5 0.656 P + 0.783 p Q( x, w ) > 0

For an inow volume of P = 50 acre-feet and an inow peak rate of p = 1,000 cubic feet per sec-
ond, the predicted outow volume is

Q ( x, w ) = 5.78 + 0.783 (50 ) = 33.4 acre-ft

and the predicted peak rate of outow is

q ( x, w ) = 17.5 0.656 (50 ) + 0.783 (1, 000 )

= 733 ft 3 /s

(210VINEH, August 2007) 199

Chapter 19 Transmission Losses Part 630
National Engineering Handbook

Example 192 Uniform lateral inow

Given: The channel reach parameters from example 191 and a lateral inow of 21.3 acre-feet at a peak
rate of 500 cubic feet per second. Assume the lateral inow is uniformly distributed.

Find: The volume of outow and peak rate of outow if P = 50 acre-feet and p = 1,000 cubic feet per
second.

Solution: Compute the lateral rates as follows:

21.3 acre-ft
QL = = 4.26 acre-ft/mi
5.0 mi

and

500 ft 3 /s
qL = = 0.0189 ft 3 /s/ft
(5.0 mi )(5, 280 ft/mi )
Using a(x,w) = 5.78, b(x,w) = 0.783, k = 0.000699, and w = 70 from case 2 of example 191 in
equation 194, the result is
QL
Q ( x, w ) = 5.78 + 0.783P + (1 0.783 )
kw
= 52.3 acre-ft

The corresponding calculations for peak discharge of the outow hydrograph (eq. 195) are
q L (5, 280 )
q ( x, w ) = 17.5 0.656P + 0.783 p + (1 0.783 )
kw
= 1,175 ft 3 /s

1910 (210VINEH, August 2007)

Chapter 19 Transmission Losses Part 630
National Engineering Handbook

Example 193 Approximations for out-of-bank ow

Given: A channel reach of length x = 10 miles and an average width of inbank ow w1 = 150 feet with
inbank ow up to a discharge of 3,000 cubic feet per second. Once the ow exceeds 3,000 cubic
feet per second, out-of-bank ow rapidly covers wide areas. The bed material consists of clean
sand and gravel, and the out-of-bank material is sandy with signicant amounts of silt-clay.

Find: Determine the outow if the inow is P = 700 acre-feet with a peak rate of p = 4,000 cubic feet
per second. Assume the mean duration of ow is 12 hours and the total average width of out-of-
bank ow is 400 feet. Also, estimate the distance downstream before the ow is back within the
channel banks.

Solution: Using the procedures outlined in table 193, make the following calculations:

Inbank ow:
w1 = 150 ft
K1 = 3.0 in/h (average hydraulic conductivity from table 191)

Out-of-bank ow:
w2 = 400 ft (includes width w1)
K2 = 0.5 in/h for width w2 - w1 (average hydraulic conductivity from table 191)

The weighted average for effective hydraulic conductivity is

w1 K 1 + ( w 2 w1 ) K 2
K= (eq. 1917)
w2
K = 1.44 in/h

Using this average value of K, D = 12 hours, and P = 700 acre-feet in equations 1915 and 1916,
the unit channel parameters are
a = 0.00465KD = 0.08035 acre-ft
KD
k = 1.09 ln 1.0 0.00545
P
= 0.000147 ( ft -mi )
1

and

b = e k = e 0.000147 = 0.99985

Given the unit channel parameters and w2 = 400 feet, the parameters for the channel reach are
( )( )
b( x, w 2 ) = e kxw 2 = e 0.000147 400 x = e 0.0588 x
and
a( x, w 2 ) =
a
1 b [
1 b( x, w 2 )]
0.08035 (
= 1 e 0.0588 x )
(1 0.99985)

(210VINEH, August 2007) 1911

Chapter 19 Transmission Losses Part 630
National Engineering Handbook

Example 193 Approximations for out-of-bank owContinued

Now, estimate the distance downstream until ow is contained within the banks (from equation
193) as

q ( x, w ) =
12.1
D
{ }
a ( x, w ) 1 b ( x, w ) P + b ( x, w ) p

Use an upper limit as

q ( x, w ) = 3, 000 ft 3 /s b ( x, w ) p = e 0.0588 x ( 4, 000 )

which means
3, 000
e 0.0588 x = 0.75
4, 000

1.0
x ln 0.75 = 4.89 mi
0.0588

Then a trial-and-error solution of the volume and peak discharge equations for various values of
x < 4.89 miles produces a best estimate of x = 3.6 miles. Based on this value, the parameters are

b( 3.6, w 2 ) = 0.809

and
a ( 3.6, w 2 ) = 102.3 acre-ft
Therefore, the predictions for x = 3.6 miles are
Q ( 3.6, w 2 ) = 102.3 + 0.809 (700 )
= 464.0 acre-ft
for the volume, and

q ( 3.6,w 2 ) =-238.0+0.809 ( 4,000 ) =2,998 ft 3 /s

for the peak rate.

For distances beyond this point, the ow will be contained in the channel banks. Using K = 3.0,
D = 12, and P = 464.0 acre-feet (the inow from the upstream reach), the parameters for inbank
ow with a distance of x = 10.0 3.6 = 6.4 miles are
a = 0.00465KD = 0.1674 acre-ft
KD
k = 1.09 ln 1.0 0.00545
P
= 0.000461 ( ft -mi )
1

and
b = e k = e 0.000461 = 0.99954

1912 (210VINEH, August 2007)

Chapter 19 Transmission Losses Part 630
National Engineering Handbook

Example 193 Approximations for out-of-bank owContinued

With these unit channel parameters, the parameters for inbank ow are

( ) ( )( )( )
b 6.4, w 1 = e kxw 1 = e 0.000461 6.4 150 = 0.642
and
a
a (6.4, w1 ) = 1 b ( x, w1 )
1b
( 0.1674 )
= [1 0.642]
(1 0.99954 )
= 130.3 acre-ft

The predicted outow is

Q (6.4, w1 ) = 130.3 + 0.642 ( 464.0 )
= 167.6 acre -ft
for the volume and
q (6.4,w1 ) =-298.9+0.642 ( 2,998 )
=1,626 ft 3 /s

for the peak discharge. Therefore, the prediction is out-of-bank ow for about 3.6 miles and
inbank ow for 6.4 miles, with an outow volume of 168 acre-feet and a peak discharge of
1,626 cubic feet per second.

(210VINEH, August 2007) 1913

Chapter 19 Transmission Losses Part 630
National Engineering Handbook

Example 193 illustrates the need for judgment in In some circumstances, an alluvial channel could be
applying the procedure for estimating losses in out-of- underlain by nearly impervious material that might
bank ow. Care must be taken to ensure that transmis- limit the potential storage volume in the alluvium (V)
sion losses do not reduce the ow volume and peak and thereby limit the potential transmission losses.
to the point where ow is entirely within the channel Once the transmission losses ll the available stor-
banks. If this occurs, then the reach length must be age, nearly all additional inow becomes outow. The
broken into subreaches, as illustrated in this example. procedure as shown in example 194 is modied to
predict and apply this secondary threshold volume, P1.

Example 194 Transmission losses limited by available storage

Given: The channel reach in example 191 with total available storage (maximum potential transmis-
sion loss) of V = 30 acre-feet.

Find: Given the volume equation from case 1 of example 191, compute equations to apply after the
potential losses are satised. From example 191, a(x,w) = -10.38 acre-feet, b(x,w) = 0.850, and
Po(X,W) = 12.21 acre-feet.

Solution: The total losses are P Q(x,w) computed as

P a ( x, w ) + b ( x, w ) P = a ( x, w ) + 1 b ( x, w ) P

Equating this computed loss to V and solving for the inow volume predicts the inow volume
above which only the maximum alluvial storage is subtracted:
V + a ( x, w )
P1 =
1 b ( x, w )

For this example, this threshold inow volume is 130.8 acre-feet. With this additional threshold,
the prediction equation for outow volume is modied to
0 P Po ( x, w )



Q( x, w ) = a x, w + b x, w P
( ) ( ) Po ( x, w ) P P1


P V P > P1
(eq. 1918)
The solution to this general equation is
0 P 12.21


Q( x, w ) = 10.38 + 0.850P 12.21 P 130.8


P 30 P > 130.8

1914 (210VINEH, August 2007)

Chapter 19 Transmission Losses Part 630
National Engineering Handbook

Example 194 Transmission losses limited by available storageContinued

The slope of the regression line is equal to

Q( x , w )
[P Po (x, w )]
so an equivalent slope, once the available storage is lled, is
(P V )
beq =
[P Po (x, w )]
which for this example is
( P 30)
beq =
( P 12.21)
For an inow volume of P = 300 acre-feet and peak rate of inow p = 3,000 cubic feet per
second, the equivalent slope is beq = 0.938. Using the equivalent slope, the peak equation is
12.1
q ( x, w ) = P Q ( x, w ) + beq p
D
= 90.75 + 0.938 ( 3, 000 ) = 2, 723 ft 3 /s

Therefore, the predicted outow is Q(x,w) = 270 acre-feet and q(x,w) = 2,723 cubic feet per
second.

If the storage limitation had been ignored, the original equations would have predicted an
outow volume of 245 acre-feet and a peak rate of outow of 2,384 cubic feet per second. If a
channel reach has limited available storage, the procedure should be modied, as it was in this
example, to compute losses that do not exceed the available storage.

(210VINEH, August 2007) 1915

Chapter 19 Transmission Losses Part 630
National Engineering Handbook

630.1905 Summary

The examples presented illustrate the wide range

of applications of the transmission loss procedures
described in this chapter. They were chosen to empha-
size some limitations and the need for sound engineer-
ing judgment. These concepts are summarized in table
194.

Table 194 Outline of examples and comments on their applications

Example Procedure Special circumstances Comments

191 Table 192 Observed data available Slope and intercept must satisfy the constraints
(case 1)
191 Table 193 No observed data Typical application
(case 2)
192 Table 193 Uniform lateral inow Importance of lateral inow demonstrated
Eqs. 194, 195
193 Table 193 Out-of-bank ow Judgment required to interpret results
Eq. 1917
194 Table 192 Limited available storage Concept of equivalent slope used
Eq. 1918

1916 (210VINEH, August 2007)

Chapter 19 Transmission Losses Part 630
National Engineering Handbook

Lane, L.J., V.A. Ferreira, and E.D. Shirley. 1980. Esti-

630.1906 References mating transmission losses in ephemeral stream
channels. Proc. Am. Water Resources Assoc.
and Ariz.-Nev. Acad. of Sci., Las Vegas, NV. p.
Babcock, H.M., and E.M. Cushing. 1942. Recharge 193202.
to ground water from oods in a typical desert
wash. Pinal County, Arizona, Trans. Am. Geo- Matlock, W.G. 1965. The effect of silt-laden water on
phys. U. 23(1):4956. inltration in alluvial channels. Ph.D. disserta-
tion, Univ. of AZ, Tucson, AZ. 102 p.
Briggs, P.C., and L.L. Werho. 1966. Inltration and
recharge from the ow of April 1965 in the Salt Peebles, R.W. 1975. Flow recession in the ephemeral
River near Phoenix, Arizona. Ariz. State Land stream. Ph.D. dissertation, Dep. Hydrology and
Dep. Water Res. Rep. No. 29, 12 p. Water Resources, Univ. of AZ, Tucson, AZ. 88 p.

Burkham, D.E. 1970a. A method for relating inltration Peterson, H.V. 1962. Discussion, Transmission losses
rates to streamow rates in perched streams. in ephemeral streambeds, by R.V. Keppel and
U.S. Geol. Surv. Prof. Pap. 700 D. p. D266D271. K.G. Renard. Proc. Am. Soc. Civil Eng., J. Hy-
draulics Div. 88(HY5):339343.
Burkham, D.E. 1970b. Depletion of streamow by inl-
tration in the main channels of the Tucson Basin, Renard, K.G. 1970. The hydrology of semiarid range-
southeastern Arizona. U.S. Geol. Surv. Water Sup- land watersheds. U.S. Dept. Agric., Agric.
ply Pap. 1939B, 36 p. Research Serv. 41162, 26 p.

Jordan, P.R. 1977. Streamow transmission losses in Smith, R.E. 1972. Border irrigation advance and
western Kansas. Proc. Am. Soc. Civil Eng., J. ephemeral ood waves. Proc. Am. Soc. Civil
Hydraulics Div. 103(HY8):905919. Eng., J. Irr. and Drainage Div. 98(IR2):289307.

Keppel, R.V. 1960. Transmission losses on Walnut Texas Board of Water Engineers. 1960. Channel gain
Gulch Watershed. In Joint ARSSCS Hydrology and loss investigations, Texas streams, 1918
Workshop, New Orleans, LA. p. 21.121.8. 1958. 270 p.

Keppel, R.V., and K.G. Renard. 1962. Transmission U.S. Department of Agriculture, Natural Resources
losses in ephemeral streambeds. Proc. Am. Soc. Conservation Service. 2005. Estimation of direct
Civil Eng., J. Hydraulics Div. 88(HY3):5968. runoff from storm rainfall. National Engineering
Handbook 630, chapter 10. Washington, DC.
Kraatz, D.B. 1977. Irrigation and canal lining. United
Nations Food and Agriculture Organization, U.S. Department of Agriculture, Natural Resources
Rome, Italy. 199 p. Conservation Service. 2006. Hydrographs. Na-
tional Engineering Handbook 630, chapter 16.
Lane, L.J. 1972. A proposed model for ood routing in Washington, DC.
abstracting ephemeral channels. Hydrology and
water resources in Arizona and the Southwest. Wilson, L.G., K.J. DeCook, and S.P. Neuman. 1980.
Proc. Am. Water Resources Assoc. and Ariz. Final report: Regional recharge research for
Acad. of Sci., Prescott, AZ. Vol. 2, p. 439453. Southwest alluvial basins. Water Resources Re-
search Center, Dep. Hydrol. and Water Res., Univ.
Lane, L.J., M.H. Diskin, and K.G. Renard. 1971. Input- of AZ., Tuscon, AZ.
output relationships for an ephemeral stream
channel system. J. Hydrol. 13:2240. Wu, I-pai. 1972. Recession ows in surface irrigation.
Proc. Am. Soc. Civil Eng., J. Irr. and Drainage
Div. 98(IR1):7790.

(210VINEH, August 2007) 1917

Appendix 19A Derivation of Procedures for
Estimating Transmission Losses
When Observed Data are Available
In much of the southwestern United States, water- By setting Q(x,w) = 0.0 and solving for P, the threshold
sheds are characterized as semiarid with broad al- volume, the volume of losses that occur before out-
luvium-lled channels that abstract large quantities ow begins is

( )
of streamow (Babcock and Cushing 1942; Burkham
a x , w
1970a, 1970b; Renard 1970). These abstractions or
transmission losses are important because streamow
(
Po x, w = ) b( x, w )
(eq. 192)

is lost as the ood wave travels downstream, and thus,

runoff volumes are reduced. Although these abstrac-
tions are referred to as losses, they are an important Differential equation for changes in
part of the water balance. They diminish streamow, volume
support riparian vegetation, and recharge local aqui-
fers and regional ground water (Renard 1970). Linkage with the regression model
Differential equations can be used to approximate the
Simplied procedures have been developed to esti- inuence of transmission losses on runoff volumes.
mate transmission losses in ephemeral streams. These Because the solutions to these equations can be ex-
procedures include simple regression equations to pressed in the same form as the regression equations,
estimate outow volumes (Lane, Diskin, and Renard least-squares analysis can be used to estimate param-
1971) and simplied differential equations for loss rate eters in the transmission loss equations.
as a function of channel length (Jordan 1977). Other,
more complicated methods have also been used (Lane Unit channel
1972; Wu 1972; Smith 1972; Peebles 1975). The rate of change in volume, Q (as a function of arbi-
trary distance), with changing inow volume, P, can be
Lane, Ferreira, and Shirley (1980) developed a proce- approximated as
dure to relate parameters of the linear regression equa-
tions (Lane, Diskin, and Renard 1971) to a differential
= c k Q( x )
dQ
equation coefcient and the decay factor proposed by (eq. 1919)
dx
Jordan (1977). This linkage between the regression
and differential equations provides the basis of the ap- Substituting the initial condition and dening
plications described in this chapter. P = Q(x = 0), the solution of equation 1919 is

c(
Empirical basis of the regression Q( x ) = 1 e kx ) + Pe kx (eq. 1920)
k
equation
For a unit channel, equation 1920 becomes
When observed inow-outow data for a channel
reach of an ephemeral stream with no lateral inow c
are plotted on rectangular coordinate paper, the re- Q= (1 e k ) + Pe k (eq. 1921)
sult is often no outow for small inow events, with k
outow increasing as inow increases. When data are
which corresponds to the regression equation
tted with a straight-line relationship, the intercept on
the X axis represents an initial abstraction. Graphs of Q = a + bP (eq. 1922)
this type suggest equations of the form
Equating equations 1921 and 1922, it follows that
0

(
P Po x, w )
Q( x , w ) = b = e k (eq. 1911)

( ) ( )
a x, w + b x, w P (
P > Po x, w ) and
(eq. 191)
c(
1 e k ) = (1 b) (eq. 1923)
c
a=
k k
are the linkage equations.

(210VINEH, August 2007) 19A1

Chapter 19 Transmission Losses Part 630
National Engineering Handbook

Equation 1923 can be solved for c as and through the linkage, the outow volume equation
for upstream inow augmented by uniform lateral
a inow is
c = k
1b
Q( x, w ) = a( x, w ) + b( x, w )P +
QL
kw
[
1 b( x, w ) ]
Channel of arbitrary length and width
For a channel of width w and length x,

Approximations for peak discharge

= wc wkQ( x, w )
dQ
dx
The basic assumption for peak discharge, q(x,w), is
that the outow peak, once an average loss rate has
a
where: c = k so that the differential equation is been subtracted, is equal to b(x,w) times the peak of
1b the inow hydrographs, p. That is, assume that

P Q( x , w )
dQ a
= wk wkQ ( x , w ) +
dx 1 b q( x, w ) = + b( x, w ) p
D

Dening P as Q(x = 0) and substituting this initial con- where:

P Q ( x, w ) = a ( x, w ) + 1 b ( x, w ) P
dition, the solution is

Q ( x, w ) =
a
1 b
( )
1 e kxw + Pe kxw so that

From the linkage q ( x, w ) =

12.1
D
{ }
a ( x, w ) 1 b ( x, w ) P + b ( x, w ) p

b( x, w ) = e kxw (eq. 1913) where:

D = mean duration of ow, and
and 12.1 converts acre-feet per hour to cubic feet per
a second
a ( x, w ) = 1 b ( x, w )
1b
(eq. 1914)
For a peak lateral inow rate of qL (ft3/s/ft), uniform
a ( x, w ) =
a
[1 e kxw ] along the reach, the peak discharge equation becomes
1b
where:
a and b = unit channel parameters
k = decay factor
q ( x, w ) =
12.1
D
{a ( x, w ) 1 b ( x, w ) P }
q (5, 280 )
+ b ( x, w ) p + L 1 b ( x, w )
Inuence of uniform lateral inow kw
If QL is the uniform lateral inow (acre-feet/mile), this
inow becomes an additional term in the differential
equation where 5,280 converts cubic feet per second per foot to
cubic feet per second per mile.
dQ a
= wk wkQ ( x , w ) + Q L For small inows where the volume of transmission
dx 1 b losses is about equal to the volume of inow, the peak
discharge equation, equation 193, overestimates the
The solution is
peak rate of outow. The relation between peak rate
a (
Q( x , w ) = 1 e kxw ) + Pe kxw + L (1 e kxw )
Q of outow observed and that computed from equation
1 b kw 193 is shown in gure 19A1. The bias shown in g-
ure 19A1 is for small events and tends to overpredict,

19A2 (210VINEH, August 2007)

Chapter 19 Transmission Losses Part 630
National Engineering Handbook

but the equation does well for the larger events. The Figure 19A1 Observed versus computed peak discharge
computed values shown in gure 19A1 were based of the outow hydrograph
on the mean duration of ow for each channel reach.
10,000
Better agreement of predicted and observed peak
rates of outow might be obtained by using actual ow

Computed peak discharge (ft3/s)

durations.

1,000

100

Walnut Gulch, AZ reach 11-8

Queen Creek, AZ
Elm Fork of the Trinity River, TX
10
10 100 1,000 10,000
Observed peak discharge (ft3/s)

(210VINEH, August 2007) 19A3

Appendix 19B Analysis of Selected Data Used to
Develop the Procedure for Estimating
Transmission Losses
Selected data had to be analyzed so that param- The Trinity River data represent pumping diversions
eters of the prediction equations could be related to entirely within the channel banks. Data for the Kansas-
hydrograph characteristics and to effective hydraulic Nebraska streams represent oods of unknown size
conductivity. Events involving little or no lateral inow and may include out-of-bank ow.
were selected from channel reaches in Arizona, Kan-
sas, Nebraska, and Texas (table 19B1). Data summarized in table 19B1 were subjected to
linear regression analysis to estimate the parameters
The data shown in table 19B1 are not entirely con- a(x,w), b(x,w), Po(x,w), and kxw. These parameters
sistent because the events were oods of different are summarized in table 19B2. Parameters for the
magnitudes. The Walnut Gulch data are from a series unit channels were computed for 10 channel reaches
of small to moderate events representing inbank ow, and are shown in table 19B3.
whereas the Queen Creek data are for relatively larger
oods and no doubt include some out-of-bank ow.

Table 19B1 Hydrologic data used in analyzing transmission losses (Lane, Ferreira, and Shirley 1980)

Location Reach Length, x Average Number - - - Inow volume - - - - - - Outow volume - - -

identication width, w of events
Mean Standard Mean Standard
deviation deviation
mi ft acre-ft acre-ft acre-ft acre-ft

Walnut Gulch, AZ 1/ 11-8 4.1 38 11 16.5 14.4 8.7 11.4

8-6 0.9 -- 3 13.7 -- 11.4 --
8-1 7.8 -- 3 16.3 -- 1.6 --
6-2 2.7 107 30 75.1 121.6 59.9 101.0
6-1 6.9 121 19 48.3 51.7 17.1 26.5
2-1 4.2 132 32 49.3 42.7 24.4 31.4
Queen Creek, AZ 2/ Upper to lower 20.0 277 10 4,283 5,150 2,658 3,368
gaging station
Elm Fork of Trinity Elm Fork-1 9.6 -- 3 454 -- 441 --
River, TX 3/ Elm Fork-2 21.3 -- 3 441 -- 424 --
Elm Fork-3 30.9 120 3 454 -- 424 --
Kansas-Nebraska 4/ Prairie Dog 26.0 17 5 1,890 1,325 1,340 1,218
Beaver 39.0 14 7 2,201 2,187 1,265 1,422
Sappa 35.0 23 6 6,189 8,897 3,851 7,144
Smokey Hills 47.0 72 4 1,217 663 648 451

1/ Data le at USDA-ARS, Southwest Rangeland Water Research Center, 442 E. 7th Street, Tucson, AZ 85705.
2/ Data from Babcock and Cushing (1942).
3/ Data from the Texas Board of Water Engineers (1960).
4/ Data from Jordan (1977).

(210VINEH, August 2007) 19B1

Chapter 19 Transmission Losses Part 630
National Engineering Handbook

Table 19B2 Parameters for regression model and differential equation model for selected channel reaches (Lane, Ferreira,
and Shirley 1980)

Location Reach Reach Length, Average Regression Model Threshold Decay R2

identication no. x width, w intercept, slope, volume, factor,
a(x,w) b(x,w) Po(x,w) kxw
mi ft acre-ft acre-ft

Walnut Gulch, AZ 11-8 1 4.1 38 4.27 0.789 5.41 0.2370 .98

8-6 2 0.9 -- 0.34 0.860 0.40 0.1508 .99
8-1 3 7.8 -- 2.38 0.245 9.71 1.4065 .84
6-2 4 2.7 107 4.92 0.823 5.98 0.1948 .98
6-1 5 6.9 121 5.56 0.469 11.86 0.7572 .84
2-1 6 4.2 132 8.77 0.673 13.03 0.3960 .84
Queen Creek, AZ Upper to 7 20.0 277 117.2 0.648 180.90 0.4339 .98
lower station

Elm Fork of Elm Fork-1 8 9.6 -- 15.0 1.004 1/ -- -- .99

Trinity River, TX Elm Fork-2 9 21.3 -- +7.6 1/ 0.944 -- -- .99
Elm Fork-3 10 30.9 120 8.7 0.952 9.14 0.0492 .99
Kansas-Nebraska Prairie Dog 11 26.0 17 353.1 0.896 394.10 0.1098 .95
Beaver 12 39.0 14 157.3 0.646 243.50 0.4370 .99
Sappa 13 35.0 23 1,076.3 0.796 1,352.10 0.2282 .98
Smokey Hills 14 47.0 72 99.1 0.614 161.40 0.4878 .81

1/ Channel reaches where derived regression parameters did not satisfy the constraints.

19B2 (210VINEH, August 2007)

 Table 19B3 Unit length, unit width, and unit length and width parameters for selected channel reaches (Lane, Ferreira, and Shirley 1980)
Chapter 19

Location Identication - - - - - - Unit length parameters - - - - - - - - - - - - Unit width parameters - - - - - - - - - - - - - - - Unit length and width parameter - - - - - - - -
a(w) b(w) Po(w) a(x) b(x) Po(x) a b Po k
Walnut 118 1.13657 0.94384 1.2042 0.12587 0.99378 0.1267 0.03076 0.998480 0.0308 0.001521
Gulch, AZ 62 1.93484 0.93039 2.0796 0.05059 0.99818 0.0507 0.01874 0.999326 0.0187 0.000674
61 1.08819 0.89607 1.2144 0.06541 0.99376 0.0658 0.00950 0.999094 0.0095 0.000907
21 2.41320 0.91002 2.6518 0.08046 0.99700 0.0807 0.01915 0.999286 0.0192 0.000714
Queen Upper to 7.14508 0.97854 7.3018 0.52273 0.99843 0.5236 0.02597 0.999922 0.0260 0.0000783
Creek, AZ lower station
Trinity Elm Fork-3 -0.28825 0.99841 0.2887 -0.07427 0.99959 0.0743 0.002404 0.999987 0.0024 0.0000133
River, TX
Kansas- Prairie Dog 14.30986 0.99579 14.3705 21.86124 0.99356 22.0029 0.842008 0.999752 0.8422 0.000248
Nebraska Beaver 4.95071 0.98886 5.0065 13.65447 0.96927 14.0874 0.355480 0.999200 0.3558 0.000800
Sappa 34.28091 0.99350 34.5052 52.07808 0.99013 52.5972 1.493102 0.999717 1.4935 0.000283
Smokey Hills 2.65060 0.98968 2.6782 1.73337 0.99325 1.7451 0.036970 0.999856 0.0370 0.000144
Transmission Losses

(210VINEH, August 2007)

Part 630
National Engineering Handbook

19B3
Blank
Appendix 19C Estimating Transmission Losses When
No Observed Data are Available

Estimating transmission losses when observed inow- However, because a and (1 b)P are in acre-feet and
outow data are not available requires a technique KD, the product of conductivity and duration, is in
for using effective hydraulic conductivity to develop inches, the dimensionally correct equation is
parameters for the regression analysis.
( )
a + 1 b P = 0.0101KD
Estimating effective hydraulic
conductivity where 0.0101 converts inches over a unit channel to
acre-feet. Because this equation is in two unknowns (a
The total volume of losses for a channel reach is KD, and b), an additional relationship is required to solve
where K is the effective hydraulic conductivity and D it. As a rst approximation, the total losses are parti-
is the duration of ow. Also, the total losses are tioned between the two terms in the equation.
PQ(x,w), so that: That is, let

a = ( 0.0101KD )
KD = 0.0275 [ P Q ( x , w )]
and
(1 b ) = (1 ) 0.0101 KD
where 0.0275 converts acre-feet per foot-mile-hour to
inches per hour. Or, solving for K: P

K=
[
0.0275 P Q( x, w )] Solving for b,
D KD
b = 1 (1 ) 0.0101
But P

where 0 1 is a weighting factor. Solve for k by

[ ]
P Q( x, w ) = a( x, w ) + 1 b( x, w ) P substituting b = e-k and taking the negative natural log
of both sides; i.e.:
so that
KD
k = ln 1 (1 ) 0.0101
P
K=
0.0275
D { [
a( x, w ) + 1 b( x, w ) P ]}
The selected data were analyzed to determine by
(eq. 1924) least-squares tting as shown in table 19C1. For the
data shown in table 19C1, the estimate of was 0.46.
is an expression for effective hydraulic conductivity. If Figures 19C1 and 19C2 show the data in table 19C1
mean values for D and P are used, then equation 1924 plotted according to the equations
estimates the mean value of the effective hydraulic
conductivity. a ( D) = 0.00465KD (eq. 1915)
and
Effective hydraulic conductivity versus KD
k ( D, P ) = 1.09 ln 1.0 0.00545 (eq. 1916)
model parameters P
For a unit channel, outow is the difference between where for each channel reach, mean values were used
inow and transmission losses: for K, D, and P. These relationships were used to calcu-
late the values shown in table 191 of the main text.
Q = P KD
Auxiliary data compiled in a report by Wilson, De-
Because Q = a + bP, Cook, and Neuman (1980) are shown in table 19C2.
Although the estimates of inltration rates were ob-
( )
a + 1 b P = KD tained by a variety of methods, most rates were based
on streamow data. Because these estimates generally
involved longer periods of ow than in the smaller

(210VINEH, August 2007) 19C1

Chapter 19 Transmission Losses Part 630
National Engineering Handbook

Table 19C1 Data for analysis of relations between effective hydraulic conductivity and model parameters (Lane, Ferreira,
and Shirley 1980)

Location Unit channel Decay K KD KD/P ln[10.00545(KD/P)] Comments

intercept, a factor, k
acre-ft (ft-mi)-1 in/h in in/acre-ft

Walnut Gulch
118 0.03076 0.001521 1.55 4.96 0.3010 0.001643 Inbank ow
62 0.01874 0.000674 1.36 6.26 0.0834 0.000455
61 0.00950 0.000907 1.03 3.71 0.0768 0.000419
21 0.01915 0.000714 1.11 4.44 0.0901 0.000492
Queen Creek 0.02597 0.0000783 0.54 29.16 0.0068 0.0000371 Mixed ow
Elm Fork 0.00240 0.0000133 0.01 0.84 0.0019 0.0000104 Inbank ow
Kansas-Nebraska
Prairie Dog 0.84201 0.000248 1.28 122.9 0.0650 0.000355 Mixed ow:
Beaver 0.35548 0.000800 1.38 169.7 0.0771 0.000421 average widths
Sappa 1.49310 0.000283 2.57 287.8 0.0465 0.000254 may be under-
Smokey Hills 0.03697 0.000144 0.17 16.3 0.0134 0.000073 estimated

Least-squares t:

a ( D) = 0.00465KD

k ( D, P ) = 1.09 ln 1.0 0.00545
KD
P

Figure 19C1 Relation between KD and regression Figure 19C2 Relation between KD/P and decay factor
intercept
-10 .01
a unit channel regression intercept (acre/ft)

Walnut Gulch
Queen Creek Walnut Gulch
Elm Fork Queen Creek
Kansas-Nebraska Elm Fork
Kansas-Nebraska
-1.0
k decay factor (ft/mi)-1

.001
a (D) = 0 . 00465KD
-.10

.0001
-.01

k  D, P  1.09 ln 1.0 0.00545
KD
P
-.001 .00001
0.1 1.0 10 100 1,000 .001 .01 .10 1.0
KD effective conductivity-duration product (in) KD
P Conductivity-duration product normalized
by inflow volume (in/AF)

19C2 (210VINEH, August 2007)

Chapter 19 Transmission Losses Part 630
National Engineering Handbook

ephemeral streams, they should be representative of For comparison, seepage loss rates for unlined canals
what is called effective hydraulic conductivity. The are shown in table 19C3. Though these data are not
data show the range of estimates of hydraulic conduc- strictly comparable with loss rates in natural channels,
tivity for various streams within a river basin as esti- they do show the variation in inltration rates with dif-
mated by several investigators. For this reason, they ferent soil characteristics. Inltration rates varied by a
should be viewed as qualitative estimates. Improved factor of over 20 (0.123.0 in/h) from a clay loam soil
estimates based on site-specic conditions were used to a very gravelly soil.
in developing the prediction equations.

Table 19C2 Auxiliary transmission loss data for selected ephemeral streams in southern Arizona (Wilson, DeCook, and
Neuman 1980)

River basin Stream reach Estimation method Effective Source of

hydraulic estimates
conductivity
(in/h)

Santa Cruz Santa Cruz River, Tucson to Continental Streamow data 1/ 1.5 3.4 Matlock (1965)
Santa Cruz River, Tucson to Cortero Streamow data 3.2 3.7 Matlock (1965)
Rillito Creek, Tucson Streamow data 0.5 3.3 Matlock (1965)
Rillito Creek, Cortero Streamow data 2.2 5.5 Matlock (1965)
Pantano Wash, Tucson Streamow data 1.6 2.0 Matlock (1965)
Average for Tucson area 1.65 Matlock (1965)
Gila Queen Creek Streamow data:
Summer ows 0.07 0.52 Babcock and
Cushing (1942)
Winter ows 0.37 1.05 Babcock and
Cushing (1942)
Average for 0.54 Babcock and
all events Cushing (1942)
Seepage losses >2.0 Babcock and
in pools 2/ Cushing (1942)
Salt River, Granite Reef Dam to 7th Ave. Streamow data 0.75 1.25 Briggs and
Werho (1966)
San Pedro Walnut Gulch Streamow data 1.1 4.5 Keppel (1960)
Keppel and
Renard (1962)
Walnut Gulch Streamow data 2.4 Peebles (1975)
San Simon San Simon Creek 0.18 Peterson (1962)

1/ Transmission losses estimated from streamow data.

2/ Measurement of loss rates from seepage in isolated pools.

(210VINEH, August 2007) 19C3

Chapter 19 Transmission Losses Part 630
National Engineering Handbook

Table 19C3 Range of seepage rates in unlined canals 1/

Effective hydraulic Description of materials 2/

conductivity (in/h)

0.120.18 Clay-loam, described as impervious

0.250.38 Ordinary clay loam
0.380.50 Sandy loam or gravelly clay-loam
with sand and clay
0.500.75 Sandy loam
0.750.88 Loose sandy soil
1.01.25 Gravelly sandy soils
1.53.0 Very gravelly soils

1/ Data from Wilson, DeCook, and Neuman (1980) after Kraatz

(1977).
2/ Does not reect the ashy, sediment-laden character of many
ephemeral streams.

19C4 (210VINEH, August 2007)