Color profile: Generic CMYK printer profile

Composite Default screen

539

Modified rational method for sizing infiltration

structures
A. Osman Akan

Abstract: A simple method is presented to size infiltration structures like infiltration basins and trenches to control
storm water runoff. The runoff hydrograph is assumed to be trapezoidal in shape with a peak runoff rate calculated using the rational formula. Given the watershed time of concentration and the allowable runoff rate, the method determines the required size of the infiltration structure. A practical application section is included to demonstrate the use of
the method.
Key words: rational method, infiltration basin, infiltration trench, capture volume, storage time.
Rsum : Une mthode simple est prsente afin de dimensionner des structures dinfiltration telles que des bassins ou
des tranches dinfiltration pour le contrle des coulements deau de pluies. Lhydrogramme dcoulement est assum
tre de forme trapzodale avec un taux dcoulement de pointe calcul en utilisant la formule rationnelle. Pour un
temps de concentration du bassin et un taux dcoulement permis donns, la mthode dtermine les dimensions requises pour la structure dinfiltration. Une section avec une application pratique est incluse dans larticle afin de dmontrer lutilisation de cette mthode.
Mots cls : mthode rationnelle, bassin dinfiltration, tranche dinfiltration, volume demmagasinement, temps de rtention.
[Traduit par la Rdaction]

Akan

542

Introduction
Infiltration structures, such as infiltration basins and
trenches, are constructed to capture storm water runoff and
let it percolate into the underlying soil. These structures are
feasible where the soil has adequate permeability and the
maximum water table (and (or) bedrock) elevation is sufficiently low. Unlike detention basins, there do not exist
widely accepted design standards and procedures for infiltration practices. Generally, an infiltration structure is designed
to store a capture volume of runoff for a period of storage time.
Like all storm water control structures, the design of infiltration structures involves the calculation of a runoff hydrograph corresponding to a design storm. A number of
computer-based rainfall runoff models, including the TR-20
(U.S. Department of Agriculture 1982) and HEC-1 (U.S.
Army Corps of Engineers 1990), have been popular in the
past. A much more versatile and user-friendly model, HECHMS (U.S. Army Corps of Engineers 2001), is now available for rainfall-runoff calculations. HEC-HMS includes a
variety of options for design storm selection, separation of
losses from rainfall, and runoff calculations.
Received 15 November 2001. Revision accepted 23 April
2002. Published on the NRC Research Press Web site at
http://cjce.nrc.ca on 18 June 2002.
A.O. Akan. Department of Civil Engineering, Old Dominion
University, Norfolk, VA 23529, U.S.A.
(e-mail: [email protected]).
Written discussion of this article is welcomed and will be
received by the Editor until 31 December 2002.
Can. J. Civ. Eng. 29: 539542 (2002)

I:\cjce\cjce_29\cjce-04\L02-038.vp
Thursday, June 13, 2002 10:25:14 AM

For watersheds not exceeding 812 ha in size, the rational

method, which can be viewed as a simple rainfall-runoff
model, has also found widespread use in drainage design.
This method lumps all the watershed characteristics into a
runoff coefficient and a time of concentration. In addition,
the method assumes that the design storm intensity remains
constant over the duration. As a result of these oversimplifications, the method should be used cautiously and only
where a more complex model is not warranted in terms of
the required accuracy, availability of data, and the project
budget.
The rational method was originally meant for sizing hydraulic structures such as storm sewers and culverts that are
designed to accommodate a peak discharge. The method was
not meant for design problems involving runoff volumes.
However, different variations of the rational method have
been reported in the recent years for sizing detention basins
successfully to serve small areas. One of these modified rational methods, originally presented by Aron and Kibler
(1990), was also included in the Hydrology Handbook of the
American Society of Civil Engineers (1996) and is expected
to find widespread use. The method presented herein extends
the Aron and Kibler (1990) procedure to sizing infiltration
structures.

Capture volume
As suggested by Aron and Kibler (1990), the postdevelopment runoff hydrograph is approximated by a trapezoid in the proposed method as shown in Fig. 1. The equilibrium is reached at t = Tc, where t is time and Tc is the time
of concentration of the urban watershed. The equilibrium

DOI: 10.1139/L02-038

2002 NRC Canada

Color profile: Generic CMYK printer profile

Composite Default screen

540

Can. J. Civ. Eng. Vol. 29, 2002

Fig. 1. Capture volume and filling time.

discharge is set equal to CiA, where C is the rational method

runoff coefficient, i is the rate of design rainfall, and A is the
watershed area. The state of equilibrium continues until t =
td, where td is the design storm duration. After the rain stops
at t = td, the runoff decreases linearly to zero at t = td + Tc.
Most storm water management programs require that the
post-development peak discharge be reduced to the predevelopment peak for the design return period by a best
management practice. Therefore, an infiltration practice
needs to capture that part of the runoff occurring before the
runoff rate falls to a value equal to the pre-development discharge (Maryland Department of Natural Resources 1984).
The shaded area shown in Fig. 1 represents the capture volume, Vc. From the geometry,
[1]

Vc = CiAt d

2
Qpre
Tc

2CiA

where Qpre is the pre-development peak discharge or the allowable peak discharge. The rate of rainfall, i, and the duration, td, are related through the intensitydurationreturn
period relationships for the project location. For a given design return period, an infinite number of combinations of i
and td are possible. The i and td pair leading to the largest Vc
is the most critical. As demonstrated in the practical application section below, a trial-and-error procedure is needed to
identify the critical values of i and td.
Mathematical expressions are sometimes available to describe the intensitydurationreturn period relationships. A
commonly used form for a given return period is
[2]

i=

a
b + td

where a and b are fitting parameters. If such a relationship is

available, we can eliminate the trial-and-error procedure to
determine the critical storm duration and intensity. In that
event, we substitute eq. [2] into eq. [1], differentiate the resulting equation with respect to td and set it equal to zero,
and simplify to obtain the critical storm duration as
[3]

td =

CAa
Qpre

2b
b
Tc

Filling time
Infiltration starts as soon as the runoff reaches the infiltration structure. In the current design practice, it is assumed
that the infiltration takes place at a constant rate (Maryland
Department of Natural Resources 1984; Federal Highway
Administration 1996). The structure starts filling when the
runoff rate reaching the structure exceeds fAf, where f is the
infiltration rate and Af is the effective horizontal area over
which infiltration occurs. The structure continues to fill until
the runoff rate falls to a value equal to Qpre, as shown in
Fig. 1. The filling time, Tf, can be expressed as
[4]

Tf = t d + Tc

Qpre Tc
CiA

Equation [4] should also include an additional term,

( fAf Tc )/(CiA), on the right-hand side, but this term is usually negligible.

Sizing an infiltration structure

Once the capture volume, Vc, and the filling time, Tf, are
determined, an infiltration structure can be sized employing
the procedures used in the current engineering practice. Details of these procedures are available in the literature (e.g.,
Maryland Department of Natural Resources 1984; Federal
Highway Administration 1996). The general criteria are
(i) the infiltration structure should drain completely within a
specified storage time, (ii) the bottom of the infiltration
structure should be at least a specified distance above the
high water table level, and (iii) the infiltration structure
should have enough storage capacity to accommodate the
capture volume plus the rain falling directly over the structure minus infiltration during the filling time.
The criteria (i) and (ii) for infiltration structures to be excavated into the ground define an upper limit for the depth
of the basin, respectively, as
[5]

dmax =

f Ts
n

and
[6]

dmax = GW hreq
2002 NRC Canada

I:\cjce\cjce_29\cjce-04\L02-038.vp
Thursday, June 13, 2002 10:25:15 AM

Color profile: Generic CMYK printer profile

Composite Default screen

Akan

541
Table 1. Practical application.
td
(min)
(1)

i
(cmh1)
(2)

td
(s)
(3)

i
(ms1)
(4)

30
40
50
60
70
80
90
100
110
120

5.54
4.74
4.14
3.67
3.30
3.00
2.75
2.53
2.35
2.20

1800
2400
3000
3600
4200
4800
5400
6000
6600
7200

1.5389
1.3167
1.1500
1.0194
9.1667
8.3333
7.6389
7.0278
6.5278
6.1111

where dmax is the maximum allowable depth for the infiltration practice, f is the infiltration capacity of the underlying
soil, which is assumed to remain constant, Ts is the storage
time, n is the porosity of the aggregate material filling the
infiltration structure such as infiltration trenches, GW is the
high water table elevation, and hreq is the minimum required
distance between the bottom of the infiltration practice and
high water table. The smaller of the two dmax obtained from
eqs. [5] and [6] will govern. For infiltration basins, n = 1.0.
For a trapezoidal structure having a rectangular bottom
area of W times L and side slopes of z:1 (horizontal over vertical), criterion (iii) yields
[7]

4 2 3
z d ]
3
= Vc + (L + 2zd)(W + 2zd)P (L + zd)(W + zd) fTf

n [LWd + (L + W ) zd 2 +

where d is the depth of the infiltration structure and P is the

rain falling directly over the structure. Equation [7] assumes
that the effective infiltration area is equal to the horizontal
area of the structure at mid-depth. Again, for infiltration basins, n = 1.0.

Practical application
An urban watershed has a drainage area of A = 80 000 m2,
a runoff coefficient of C = 0.60, and a time of concentration
of Tc = 30 min = 1800 s. An infiltration basin is being
planned to reduce the peak discharge resulting from a 10year design storm to Qpre = 0.5 m3s1. The intensityduration relationship for the 10-year events is tabulated in the
first two columns of Table 1. The infiltration basin will be
excavated into a sandy loam that has an infiltration capacity
of f = 2.5 cmh1. It is required that the bottom of the basin
be at least hreq = 1.2 m above the water table. The water table elevation is GW = 4 m below the ground surface. The
side slope of the basin will be 3 horizontal over 1 vertical,
that is z = 3, and the infiltration basin is to drain completely
within Ts = 72 h.
To determine the design capture volume, Vc values are
calculated for the different td and i pairs tabulated in columns 1 and 2 of Table 1. In order to use consistent units, the
td and i values are first converted to seconds and metres per
second, respectively, and listed in columns 3 and 4. The corresponding values of Vc, tabulated in column 7, are calcu-

105
105
105
105
106
106
106
106
106
106

CiAtd
(m3)
(5)

Q2preTc/(2CiA)
(m3)
(6)

Vc
(m3)
(7)

1330
1517
1656
1762
1848
1920
1980
2024
2068
2112

305
356
408
460
511
563
614
667
718
767

1025
1161
1248
1302
1337
1358
1366
1357
1350
1345

lated using eq. [1]. The largest Vc (1366 m3) results from a
storm of 90-min duration and 2.75 cmh1 intensity. Therefore, the critical storm duration is 90 min. The corresponding filling time is determined by using eq. [4] as being
Tf = 4745 s = 79 min. Also, the depth of rain falling directly
over the infiltration basin is P = itd = (7.639
10 6 ms 1)(5400 s) = 0.041 m.
We should note that we can fit an expression in the form of
eq. [2] to the intensityduration data given in columns 1 and 2
of Table 1 with a = 327 cmminh1 = 327/[(100)(60)] =
0.0546 m and b = 29 min = 1740 s. Then by using eq. [3],
we find td = 5548 s = 92.5 min. The corresponding rainfall
intensity is i = 2.69 cmh1 = 7.476 106 ms1 and then the
capture volume is Vc = 1357 m3. This result is very close to
the one obtained by trial-and-error. The small discrepancy is
due to the use of discrete values of td and i.
To size the infiltration basin, we calculate the upper limit
for the basin depth using eqs. [5] and [6], respectively,
as dmax = (0.025 mh1)(72 h)/(1.0) = 1.8 m and dmax = 4.0
1.2 = 2.8 m. Therefore, the depth of the infiltration basin
should not exceed 1.8 m. Let us set d = 1.8 m. Then any L
and W combination satisfying eq. [7] with n = 1.0, d =
1.8 m, z = 3.0, Vc = 1366 m3, P = 0.041 m, f = 2.5 cmh1 =
6.944 106 ms1, and Tf = 79 min = 4740 s is acceptable.
For instance, if we pick W = 20 m, eq. [7] will yield L =
24.55 m. Let us choose L = 25 ft for practicality. Thus, the
infiltration basin will be 1.8 m deep, and it will have side
lengths of 20 and 25 m at the bottom. At the ground surface,
the side lengths will be [20 + (2)(3)(1.8)] = 30.8 m and
[25 + (2)(3)(1.8)] = 35.8 m.

Concluding remarks
Most infiltration structures are designed to serve areas that
are smaller in size than the upper limit of 812 hectares established for the general applicability of the rational method.
Therefore, the rational method appears to be suitable for infiltration structures. However, it should not be used where
more complex procedures are warranted.

References
American Society of Civil Engineers. 1996. Hydrology handbook.
ASCE Manuals and Reports on Engineering Practice No. 28,
New York, N.Y.
2002 NRC Canada

I:\cjce\cjce_29\cjce-04\L02-038.vp
Thursday, June 13, 2002 10:25:16 AM

Color profile: Generic CMYK printer profile

Composite Default screen

542
Aron, G., and Kibler, D.F. 1990. Pond sizing for rational formula
hydrographs. American Water Resources Association Water Resources Bulletin, 26: 255258.
Federal Highway Administration. 1996. Urban drainage design
manual. Hydraulic Engineering Circular No. 22, Office of Engineering, Washington, D.C.
Maryland Department of Natural Resources. 1984. Standards and
specifications for infiltration practices. Water Resources Administration, Stormwater Management Division, Annapolis, Md.

Can. J. Civ. Eng. Vol. 29, 2002

U.S. Army Corps of Engineers. 1990. HEC-1 flood hydrograph
package users manual. Hydrologic Engineering Center, Davis,
Calif.
U.S. Army Corps of Engineers. 2001. Hydrologic modeling system
HEC-HMS version 2.1 users manual. Hydrologic Engineering
Center, Davis, Calif.
U.S. Department of Agriculture. 1982. TR-20 project formulation
hydrology. Technical Release 20, Soil Conservation Service,
Washington, D.C.

2002 NRC Canada

I:\cjce\cjce_29\cjce-04\L02-038.vp
Thursday, June 13, 2002 10:25:16 AM