115

Annex 1

Estimating soil hydrological

characteristics from soil
texture and structure
It is possible to derive rough estimates of the hydraulic conductivity (K) and the
drainable pore space () from observations of the soil profile. This is because these soil
hydraulic qualities depend on soil texture and structure. Table A1.1 average presents
values, compiled by FAO (1980) and based on data from the USBR (1984), together
with K values estimated from the /K relationship. For soils with distinct horizontal
layers, the vertical K may be taken as being at least 10 and on average 16 times lower
than the horizontal one.
As these estimates may be imprecise, more realistic K values are obtained through
field measurements, as described in Annex 3.
However, interpreting the soil structures mentioned in Table A1.1 may not be easy.
It should be done through observations of soil profiles, but shallow groundwater levels
often prevent excavation of soil pits. Moreover, soil texture and structure should be
evaluated when the soil is moist throughout.
However, in special cases, it is possible to estimate drain spacings directly from the
visual aspects of the soil profile, as was done by people with detailed local experience in
the Zuiderzee polders, the Netherlands, where it was the only possible method drain
spacings of 8, 12, 16, 24, 36 and 48 m were distinguished and the choice between
possibilities was possible.
For pure sands (almost without clay and silt), an estimate is:

where:
K = permeability (m/d).
m50 = median size of grains above 50 m. Half of the weight is above this size, half
below.

TABLE A1.1

K and values according to the soil texture and structure

Texture (USDA)1

Structure

K
(m/d)

C, heavy CL
C, CL, SC, sCL

Massive, very fine or fine columnar

0.010.02

With permanent wide cracks

0.100.20

0.010.05
> 10

Very fine or fine prismatic, angular blocky or platy

0.010.03

0.010.1

C, SC, sC, CL, sCL, SL, S, sCL

Fine and medium prismatic, angular blocky and platy

0.030.08

0.10.4

Light CL, S, SL, very fine sL, L

Medium prismatic and subangular blocky

0.060.12

0.31.0

Fine sandy loam, sandy loam

Coarse subangular block and granular, fine crumb

0.120.18

1.03.0

Loamy sand

Medium crumb

0.150.22

1.66.0

Fine sand

Single grain

0.150.22

1.66.0

Medium sand

Single grain

0.220.26

>6

Coarse sand and gravel

Single grain

0.260.35

>6

C: clay; L: loam; S: silt; s: sand.

Source: Adapted from FAO, 1980, with further elaboration.

116

Guidelines and computer programs for the planning and design of land drainage systemss

The presence of silt (< 50 m) and especially clay (< 2 m) will lower this value
considerably. Therefore, this formula should not be used for such soils.
REFERENCES
FAO. 1980. Drainage design factors. FAO Irrigation and Drainage Paper No. 38. Rome. 52 pp.
USBR. 1984. Drainage manual. A Water Resources Tech. Publication. Second printing. Denver,
USA, US Dept. of Interior, Bureau of Reclamation. 286 pp.

117

Annex 2

Statistical analysis of extremes

GUMBELS METHOD
The Gumbel distribution can be used for extrapolating from a limited number of
extreme values (Gumbel, 1954 and 1958). The basic data appear in groups, such as
the daily rainfall in August (31 days per year), or the water levels in a river per year
(365/366 days). The highest value in such a group is the extreme. The groups should
contain at least ten elements, and the minimum number of extremes (often years) is at
least ten.
The method assumes that the underlying process remains constant. This supposition
is doubtful because of recent climate changes, which also influence data such as river
flows. These changes are especially noticeable in the extreme values. Therefore, the
method should be used with care.
Extreme values are obtained as follows:
Select the highest (sometimes lowest) value in a group, e.g. the highest autumn
rainfall or the highest river discharge in a year. Each group should contain at least
ten values.
These extremes are sorted according to their magnitude in order to prepare for
further analysis.
The probability that a certain value x does not exceed a limit x0 is:
with

(1)

where:
P = probability;
n = number of extremes;
u = constant (shift);
x = values of the extremes. The average is the standard deviation is sx;
x0 = limiting value;
y = reduced Gumbel variable, with average c and standard deviation sy. For y and
for a very large number of observations, c = 0.57722 = Eulers constant;
= constant (slope).
The probability that x exceeds x0 is:
(2)
The return period T is the number of groups in which the limit x0 is exceeded. If
there is one group per year, T is in years (as in the above examples). T is defined as:
(3)
For the x values, the procedure is:

118

Guidelines and computer programs for the planning and design of land drainage systems

standard deviation of y.
Table A2.1 shows the values derived by Kendall for a smaller number of
observations.
The line y = (x - u) has two parameters: the slope , and the shift u. They can be
found by plotting on Gumbel probability paper, usually with the return period T on
the horizontal axis, the value of the extremes on the vertical. The line may be drawn
visually through the points to allow extrapolation. In this way, the once-per-century
rainfall or the river discharge can be estimated. This is even possible for much longer
return periods.
The program GUMBEL calculates the parameters automatically and provides
estimates for the extremes to be expected with a certain return period.
For agricultural drainage design, a return period of 210 years is often taken, 2
5 years for field drainage and even 10 years for crop systems with high planting costs,
and 510 years for the main system where it does not affect inhabited places.
By extrapolation, a prediction can be given over much longer periods of time in order
to obtain estimates for values to be expected once in 100 years (the once-per-century
value) and even for much longer times. However, the uncertainty of the estimates
becomes very large for such longer return periods. Moreover, for such periods (and
even for a century), the basic data series cannot be considered as constant, owing to
human and geological influences.
Nevertheless, such a prediction is valuable for engineering purposes, e.g. the height
of a river embankment able to withstand a 100-year flood. This will almost certainly
not occur 100 years later, but it has a
TABLE A2.1
chance of 1 percent of occurring next
Values of c and sy as a function of n
year.
n
c
s
The influence of climate changes
10
0.495
0.950
can be analysed by comparing data
15
0.513
1.021
from the last 1020 years with earlier
20
0.524
1.063
ones (where available), and it is wise
25
0.531
1.092
to employ the worst prediction.
30
0.536
1.112
40
0.544
1.141
Where not different, the basic data
50
0.548
1.161
include recent changes already.
y

REFERENCES
Gumbel, E.J. 1954. Statistical theory of extreme values and some practical applications. Applied
mathematics series 33. U.S. Dept. of Commerce, National Bureau of Standards.
Gumbel, E. J. 1958. Statistics of extremes. New York, USA, Columbia Press.

119

Annex 3

Field methods for measuring

hydraulic conductivity
INTRODUCTION
The K value can be measured directly in the soil layers situated below the groundwater
level using the methods described below. Less reliable methods are used to estimate
the saturated hydraulic conductivity above this level. For well-moistened granular
soils, the soil permeability for saturated flow can be estimated from the capillary
hydraulic conductivity of the unsaturated zone. However, this is not the case in wellstructured soils where this permeability is caused by cracks, holes or other macropores.
Infiltrometer or inverse auger-hole methods are often used as a compromise. They
measure conductivity under almost saturated conditions.
The field methods for determining K are based on a basic principle: water flows
through a volume of soil, whose boundary conditions are known, and the discharge is
measured; the K value is calculated by applying an equation derived from Darcys Law
applied to the specific geometry of the soil volume.
The following paragraphs review the suitability of the field methods most commonly
used to measure the soil hydraulic conductivity (auger-hole, piezometer, and inverse
auger-hole). The methods are different according to the groundwater depth at the time
of measurement. Details on these methods can be found in the bibliographic references
(Van Hoorn, 1979; USBR, 1984; Oosterbaan and Nijland, 1994; Amoozegar and Wilson,
1999).
AUGER-HOLE METHOD FOR DETERMINING SOIL PERMEABILITY
The auger-hole method (Van Beers, 1983) is the most suitable way of measuring the K
value of saturated homogeneous soils down to a depth of about 3 m. It is based on the
relationship between the K value of the soil surrounding a hole and the rate at which
the water level rises after pumping. The method measures the saturated permeability in
a rather large volume, which is an advantage in view of the large variability in natural
soils.
Method
This method for determining the soil
hydraulic conductivity (Figure A3.1)
consists of the following steps:
1. Make a hole of known depth
with a soil auger of known
diameter to a depth of at least
50 cm below the water table.
In unstable soils (e.g. sand), a
perforated filter may be needed
to support the walls.
2. Find or estimate the depth of
any impermeable soil layer. If
more than 100 cm below the
bottom of the hole, assume an
infinite depth.

FIGURE A3.1

Definition sketch: A: auger-hole method, B: inverse method

Groundwater

Initial level y 1
Final level y 2

y
H

Final level y 2
Initial level y 1
h1 h2

D > H/2 : eq. (2a)

120

Guidelines and computer programs for the planning and design of land drainage systems

3. Pump water out (e.g. with a bailer) several times and let that water flow back into
the hole.
4. Let the groundwater (where present) fill the hole until equilibrium. For
impermeable soils, return the next day; for permeable soils, a few hours are
sufficient (sometimes even a few minutes).
5. Measure the groundwater depth below soil surface.
6. Pump water out.
7. Measure the rise of the water level over time. Time intervals should be short
initially.
Example
The following data can be considered:
Depth of 8-cm diameter hole: 150 cm;
Groundwater at equilibrium: 50 cm;
Water level, first measurement: 8583 cm, t = 20 s;
Water level, second measurement: 8078 cm, t = 24 s;
Water level, third measurement: 7068 cm, t = 31s;
Impermeable base: deep (300 cm).
From these data (all distances below soil surface), the average permeability K
follows. This value is the mean value (mainly horizontal) between the groundwater
table and a few centimetres below the bottom of the hole.
It should be noted that:
The permeability of different layers can be found from measurements in holes of
different depths, but this is not very reliable; the piezometer method is better.
The first measurement may deviate because water is still running off the wall; in
this case, it should be discarded.
Measurements soon after lowering by pumping the water out are preferred.
The above methods cannot be used without an existing groundwater table at the time
of measurement. The following methods can be used in such cases. However, they are
less reliable.
The inverse method, also known as the Porchet method, may be also applied to
determine the saturated hydraulic conductivity above the groundwater level. In this case,
water is poured into an augered hole and the rate of lowering of the water level inside
the hole is measured (Figure A3.1). The measurements are taken after water has been
infiltrating for a long time until the surrounding soil is sufficiently saturated (in order
to diminish the effect of unsaturated soil on the rate of drawdown). The equation used
to calculate the K value has been derived from the balance between the water flowing
through the side walls and bottom of the hole, and the rate of lowering of the water
level in the hole. The basic assumption is that the flow gradients are unity. Although less
reliable than the measurements using an existing water table, it is often necessary where
measurements must be made outside a wet period in dry soils. However, many dry soils
swell so slowly that their permeability can only be reliably measured by the auger-hole
method during the wet season.
Van Hoorn (1979) made a comparison between normal and inverse methods and
found reasonably corresponding values for K, thus confirming the assumption about the
gradient.
Theory
According to Ernst and Westerhof (1950), Van Beers (1983) and Oosterbaan and
Nijland (1994), for the auger-hole method, the saturated soil permeability is calculated
using:

Annex 3 Field methods for measuring hydraulic conductivity

K =C

dy

(1)

dt

in which:
4000
C=

121

r
y

(2a)

y
H

+ 20 2
H
r

where the bottom of the hole is far above the impermeable base (D > H/2), or:
3600
C=

r
y

(2b)

y
H

+ 10 2
H
r

where the bottom of the hole reaches the impermeable base (D = 0). In these
formulae:
C
= constant, depending on hole geometry;
dy/dt
= rate of rise in water level (cm/s);
D
= depth of impermeable layer below bottom (cm);
h = H - y = height of water column (cm);
= initial and final water column in hole (cm);
h1,h2
H
= depth of borehole below groundwater (cm);
K
= average soil permeability (m/d);
r
= radius of borehole (cm);
t
= time (s);
y
= depth of water level below groundwater (cm);
= average value of y in the interval where y > 3/4y0 (cm);
dy y1 y2

dt
t2 t1

y1 > y2; t2 > t1

(3)

Where the impermeable base is close to the bottom of the hole, an interpolation
between Equations 2a and 2b is used.
For the inverse method, Oosterbaan and Nijland (1994) recommend:
r
r
2
ln
K=
r
2(t2 t1 )
h2 +
2
h1 +

h1 > h2; t2 > t1

(4)

which was derived analytically by integration of the following differential

equation:
dh
2K
dt
=
r
r
h+
2

(5)

In Equation 4, the value of K is expressed in centimetres per second. To convert K

from centimetres per second to metres per day, it should be multiplied by the factor
864.
The results within the same auger hole are usually quite consistent, but between
different holes, even nearby ones, differences may be considerable owing to local
soil variations. However, in predicting drain spacings, these differences become less

Guidelines and computer programs for the planning and design of land drainage systems

122

FIGURE A3.2

Definition sketch: piezometer method

r

y2
y1

important because the calculated

spacings are proportional to the
square root of K.
The program AUGHOLE makes
the necessary calculations according
to the above formulae.
The resulting K values can be used
as input in programs for calculating
drain spacings.

PIEZOMETER METHOD
FOR DETERMINING SOIL
PERMEABILITY
D
The piezometer method is more
d
Impermeable layer
convenient than the auger-hole
method for measurements of the K
value in stratified soils and in layers
deeper than 3 m. In these cases, water is pumped out of a piezometer, of which only
the lowest part is open, while the upper part of the hole is protected by a pipe. The
rate of rise in the water level inside the tube is measured immediately after pumping.
Therefore, the K value of the small layer of soil near the open part is determined.
L

Method
The piezometer method (Luthin and Kirkham, 1949) differs from the auger-hole method
in that the upper part of the hole is covered by a non-perforated pipe (Figure A3.2).
The lower part of the borehole is open and collects the water from a specific layer. In
this way, the permeability of separate layers can be found easily.
The procedure is as follows:
1. Make an auger hole and cover the upper end with a tightly fitting pipe, while the
remaining open part acts as the water-collecting cavity, or cover the entire hole
and make a narrower cavity below the pipe with a smaller auger.
2. Measure the groundwater depth at equilibrium.
3. Pump some water out and measure the rise in water level at different times.
It is most convenient to take all measurements with reference to the top of the
protecting pipe. The computer program PIEZOM is based on Kirkhams formula.
It calculates the permeability K (in metres per day) from these observations and the
geometric factors.
Theory
The basic formula is:
(6)

where A is a factor depending on the geometry of the piezometer and the hole below
the end of the piezometer and 864 a constant for converting centimetres per second
(for K) to metres per day. Various authors (Luthin and Kirkham, 1949; Smiles and
Youngs, 1965; Al-Dhahir and Morgenstern, 1969; Youngs, 1968) have provided graphs
or tables for A. Except for very small distances between the top of the piezometer and
groundwater (and within certain limits), the tables for A/d given by Youngs (1968)
(with the necessary corrections for diameter rather than radius) may be approximated
by empirical formulae for the two limiting cases and for the standard value H = 8d:

Annex 3 Field methods for measuring hydraulic conductivity

L
= 4.40
d
d

A8

123

0.661

+ 2.6

(7a)

where the bottom of the cavity hole is at the impermeable base, and:
A
L
= 4.40
d
d

0.661

L
+ 0.2 0.06 1
d

(7b)

where the bottom of the cavity hole is far above the impermeable base (more than
four times the cavity diameter). For H/d less than eight, rather complicated corrections
are made to obtain A/d.
For H/d greater than ten, no values are tabulated. As an approximation, it is
supposed that for H/d > 8 the cylindrical cavity may be represented by a sphere and
that the remaining flow is radial. For this part of the flow, the inner radius is r8 = 8d
+ L/2, whereas the outer radius is taken as the depth of the cavity centre below the
groundwater level, H + L/2. These approximations are used in the program PIEZOM;
the corrections are small because most of the resistance to flow occurs immediately
around the cavity. They are:
A A8 (1 / ro 1 / r8 )
=
d d 1 / ro 1 / r *

(8)

where:
ro =

(9a)

4 / A8 + 1 / r8

r8 = 8d + L / 2
r* H + L / 2

(9b)

for H>8d

In these formulae (see Figure A3.2):

A
= factor depending on shape (cm);
= same, for H = 8d;
A8
d
= diameter of cavity (cm);
H
= depth of top cavity below groundwater (cm);
K
= permeability (m/d);
L
= length of cavity (cm);
r
= radius of protecting pipe (cm);
= radius of sphere equivalent to cavity (for H > 8D) (cm);
ro
= radius 8d beyond which flow is supposed to be radial (cm);
r8
r*
= distance centre of cavity to surface, to be used if H/D > 8 (cm);
D
= distance to impermeable layer from cavity bottom (cm);
t
= time (s);
y
= water level below groundwater (cm);
y1, y2 = initial and final value of y (cm);

= 3.14

(9c)

124

Guidelines and computer programs for the planning and design of land drainage systems

REFERENCES
Al-Dhahir, Z.A. & Morgenstern, N.R. 1969. Intake factors for cylindrical piezometer tips. Soil
Sci., 107: 1721.
Amoozegar, A. & Wilson, G.V. 1999. Methods for measuring hydraulic conductivity and
drainable porosity. In R.W. Skaggs & J. Van Schilfgaarde, eds. Agricultural drainage, pp. 1149
1205. Agronomy Series 38. Madison, USA.
Ernst, L.F. & Westerhof, J.J. 1950. A new formula for the calculation of the permeability factor
with the auger hole method. Translated from the Dutch by H. Bouwer (1955). Ithaca, USA,
Cornell University.
Luthin, J.N. & Kirkham, D. 1949. A piezometer method for measuring permeability of soil in
situ below a water table. Soil Sci., 68: 349358.
Oosterbaan, R.J. & Nijland, H.J. 1994. Determining the saturated hydraulic conductivity.
In H.P. Ritzema, ed. Drainage principles and applications, pp. 435475. 2nd edition. ILRI
Publication 16. Wageningen, The Netherlands, ILRI.
Smiles, D.E. & Youngs, E.G. 1965. Hydraulic conductivity determinations by several field
methods in a sand tank. Soil Sci., 99: 8387.
USBR. 1984. Drainage manual. A Water Resources Tech. Publication. Second printing. Denver,
USA, US Dept. of Interior, Bureau of Reclamation. 286 pp.
Van Beers, W.F.J. 1983. The auger-hole method; a field measurement of the hydraulic
conductivity of soil below the watertable. 6th edition. ILRI Bulletin 1. Wageningen, The
Netherlands, ILRI.
Van Hoorn, J.W. 1979. Determining hydraulic conductivity with the inversed auger hole
and infiltrometer methods. In J. Wesseling, ed. Proceedings of the International Drainage
Workshop, pp. 150154. ILRI Publication 25. Wageningen, The Netherlands, ILRI.
Youngs, E.G. 1968. Shape factors for Kirkhams piezometer method for soils overlying an
impermeable bed or infinitely permeable stratum. Soil Sci., 106: 235237.

125

Annex 4

Determining drainable soil porosity

ESTIMATIONS FROM A PF CURVE

One option is to estimate the value
on a pF curve as the difference in the
water content by volume at saturation
and at field capacity. This procedure
has an important drawback because
of the differences between a small
undisturbed soil sample and the
actual field conditions. However, an
estimated average value of can be
obtained where several laboratory
measurements are taken for the same
soil layer.

Figure A4.1
Relationships between the values of K and

0.3

O
-FA
BR
US

0.2

0.1
0.09
0.08
0.07
0.06

ers

n
Va

Be

1
2
3

0.05
0.04
1

ESTIMATIONS FROM
0.02
2
3
PERMEABILITY
Another option is to estimate the
0.5 0.7 0.9
0.04 0.06 0.08
0.3
K (m/d)
0.01
value from empirical relationships
2 3 4 5 6 7 8 9 10
0.05 0.07 0.1
0.2
0.4 0.6 0.8 1
between the macroporosity and the
Note: 1. all clay content; 2. less than 15% clay; 3. 15 < clay < 30%.
hydraulic conductivity. Figure A4.1
Source: Adapted from Chossat and Saugnac, 1985.
shows the relationships developed
by Van Beers (ILRI, 1972) and the
USBR (1984) and those obtained by Chossat and Saugnac (1985) for soils with different
clay contents.
However, as there are large variations, the field methods described below may be
preferable.
OBSERVATIONS OF GROUNDWATER-LEVEL VARIATIONS
A better method is to measure the rise in groundwater level at short intervals, for
example, before and soon after a heavy rain of short duration. The rainfall is divided
by the observed rise, both expressed in the same units. If a sudden rain of 20 mm and
no runoff causes a rise of 40 cm = 400 mm, = 20/400 = 0.05 (5 percent).
In drained lands, the fall in a rainless period can also be used, in combination with
drain outflow measurements, as described in Annex 8.
LARGE CYLINDER
A more laborious method uses a large cylinder of undisturbed soil, carefully dug out.
An oil drum (without its bottom) pushed tightly over the remaining column of soil
is suitable for the purpose. After taking out, a new bottom is made by sealing the
container to a plastic plate or welding it to a steel one. Water is added, and the water
table rise inside is measured.
REFERENCES
Chossat, J.C. & Saugnac, A.M. 1985. Relation entre conductivite hydraulique et porosite de
drainage mesurees par la methode du puits et des piezometres. Sci. du Sol, 1985/3.

126

Guidelines and computer programs for the planning and design of land drainage systems

ILRI. 1972. Fieldbook for land and water management experts. Wageningen, The Netherlands.
USBR. 1984. Drainage manual. A Water Resources Tech. Publication. Second printing. Denver,
USA, US Dept. of Interior, Bureau of Reclamation. 286 pp.

127

Annex 5

Determining other soil hydrological

characteristics

DEPTH TO IMPERVIOUS BARRIER

The position of an impermeable base (bedrock or tight clay) can be found from
borings or soundings, or by geophysical methods. The existence of an impervious or
slowly permeable soil layer can be commonly identified by observations in an auger
hole where the barrier occurs within the depth of the hole, for example, when a net
change in the soil texture or a sharp increase in the soil compactness is observed and,
specifically, where a relatively dry material is found below a layer saturated with water.
However, it is not always easy to distinguish an impervious layer. In this case, a layer
can be considered as such if its hydraulic conductivity (K) is less than one-tenth of the
permeability of the overlying layer.
Where the impervious layer is not within the depth range of the auger hole, deep
borings must be carried out. Although cumbersome, hand augerings to 810 m are
possible in moist soils. Where this is not possible or does not give a result, the depth
can be estimated from soil maps or geological maps. Existing deep-water wells, or logs
from drilled wells, may provide indications of the depth. Other solutions can be found
in rough estimates of the aquifer transmissivity as described below.
THICKNESS OF THE FLOW REGION
In very deep homogeneous soils or aquifers, the lateral flow of groundwater tends to
be concentrated in the upper part, to a depth about one-third of the distance between
source and sink. In anisotropic aquifers (Kv < Kh), the active flow depth is even less.
Thus, the flow in a drained field with 20-m drain spacing, would be concentrated in the
upper 7 m, whereas flow from a hill to a valley, over a distance of 1 km mostly takes
place in the upper 300 m (although aquifers are seldom so thick). Such figures form the
upper limit of the equivalent layer (Hooghoudt, 1940).
The presence of an impermeable soil layer at a greater depth will not have a
significant effect on the flow. On the other hand, at shallower depth, the influence
becomes noticeable. The difference between real thickness and equivalent thickness is
large at first for wide drain spacings, but it becomes less as the aquifer becomes thinner,
until finally both become almost equal.
However, in drained fields, aquifers may be much thicker than one-third of the
distance between drains. Here, the equivalent thickness (d) is taken. This adjustment
is necessary because of the change from an almost horizontal flow through the aquifer
to a radial flow near the drain. Consequently, the streamlines are concentrated there,
leading to extra radial resistance and, thus, a smaller equivalent layer thickness,
with one-third of the spacing as a maximum. Deeper parts of the aquifer hardly
contribute to the flow entering the drain.
However, in thin aquifers, the water flow above the drain level is also relevant and it
cannot be ignored. Then, D = D1 + d, D1 being the average thickness of the flow region
above drain level. In some cases, as in many flat deltaic areas at or slightly above sea level
with unripened clay subsoils (e.g. the Guadalquivir Marshes in Spain, the lower part of
the Nile Delta in Egypt, and the Zuiderzee polders in the Netherlands), drains are laid
on the impervious layer and, consequently, water flows only above drain level.

128

Guidelines and computer programs for the planning and design of land drainage systems

AQUIFER TRANSMISSIVITY
The transmissivity of an aquifer is the product of permeability and thickness (KD). In
regional groundwater flow, the distances are so large (mostly several kilometres) that
the entire thickness of the aquifer can be taken. In almost all cases, it will be thin in
comparison with one-third of this distance, so that the real thickness can be taken for
D.
Estimations of the average value of KD may be made by means of a regional
approach, by applying Darcys Law to the flow area:
(1)
The hydraulic gradient, s (dimensionless), is determined on the isohypses map.
The discharge Q (cubic metres per day) over a length L (perpendicular to the flow) is
measured or derived from a water balance.
Therefore, if Q is 2 m3/d over a length of 50 m, and s = 2/1 000, KD = 20 (square
metres per day). If the layer has a thickness of 5 m, K = 4 (metres per day).
For drained fields, the KD values can be determined by field observations if the
impervious layer is not deeper than 35 m from the rise in water level in between
existing open drains and the water level in the drains and the estimate of outflow to the
drainage system at the moment of measuring. Additional details on measurement of KD
can be consulted in Annex 8. From the KD value and the measured K, it is possible to
derive the D value. Where the thickness of the aquifer is greater, pumping tests in drilled
wells are required, or regional methods can be applied (described above).
VERTICAL RESISTANCE
Another parameter, useful for estimating regional flow, is the vertical resistance (c).
Many aquifers are covered by a less permeable (but not impermeable) layer. They
are semi-confined. In many river valleys, there is a clay layer on top of a thick
sandy aquifer, the top layer formed in the Holocene, the lower one in the Pleistocene.
Groundwater has to pass through the top layer twice: first, as downwards leakage; at
the end, as upward seepage.
Such resistive layers are characterized by their thickness (D) and their
vertical permeability (Kv), and c is their proportionality quotient for vertical flow
contribution:
(2)
For a clay with Kv = 0.001 m/d and D = 2 m, the vertical resistance is c = 2 000 days.
This value is expressed in days, as electrical resistance is in Ohms. A head difference
of 1 m between bottom and top will cause upward seepage of 1/2 000 m/d or about
180 mm/year. If this groundwater contains diluted seawater, with 11 kg/m3 of salts,
the annual salt load will be about 20 tonnes/ha. Even if the water seeping upward
through the clay cap is less salty, it will cause heavy topsoil salinization in the long
run, especially in arid and semi-arid regions.
CHARACTERISTIC LENGTH
The combination of transmissivity and resistance determines the properties of the
system. Thus, the characteristic length () is a measure for the extent of seepage zones
and is roughly equal to their width. It is found from:
(3)

Annex 5 Determining other soil hydrological characteristics

where:
c = vertical resistance of covering layer (d);
d = equivalent thickness of aquifer (m);
K = permeability of the aquifer (m/d);
= characteristic length (m).
Values for c are found from pumping tests, estimated directly from experience or
derived form the thickness D and the (measured or estimated) vertical permeability
Kv of the upper layer. Pumping tests are the most reliable method (and supply values
for KD at the same time). Methods for pumping tests are described in the bibliographic
references (Boonstra and De Ridder, 1994; Kruseman and De Ridder, 1994).
Models for such regional flow, such as SAHYSMOD (ILRI, 2005), are also
available.
REFERENCES
Boonstra, J. & De Ridder, N.A. 1994. Single-well and aquifer tests. In H.P. Ritzema, ed. Drainage
principles and applications, pp. 341375. 2nd edition. ILRI Publication 16. Wageningen, The
Netherlands, ILRI.
Hooghoudt, S.B. 1940. Algemeene beschouwing van het probleem van de detailontwatering en
de infiltratie door middel van parallel loopende drains, greppels, slooten en kanalen. Versl.
Landbouwk. Onderz., 46(14).
ILRI. 2005. SAHYSMOD, spatial agro-hydro-salinity model. Version 1.7. Description of
principles, user manual and case studies. SAHYSMOD working group of ILRI. Wageningen,
The Netherlands. 134 pp.
Kruseman, G.P. & De Ridder, N.A. 1994. Analysis and evaluation of pumping tests data. ILRI
Publication 47, 2nd. Wageningen, The Netherlands, ILRI. 377 pp.

129

131

Annex 6

Estimating recharge due to

irrigation

DETERMINING DEEP PERCOLATION IN IRRIGATED FIELDS

Where drainage projects are planned and designed for irrigated lands, actual figures
of deep percolation can be estimated from the water balance on the soil surface and
in the rootzone. In dry periods when precipitation is negligible, the amount of deep
percolation produced by an irrigation application is:
(1)
where:
E = evaporation losses (mm);
I
= gross irrigation depth applied at the field level (mm);
In = amount of irrigation water infiltrated into the soil profile (mm);
Sr = amount of surface runoff (mm);
R = recharge (mm);
W = change (increase [+] and decrease [-] of the moisture content of the rootzone
(mm).
In Equation 1, the gross amount of water applied to a field, whose size is known, can
be calculated if the flow is measured with a flume and the time of watering is determined
with a watch. In a similar way, the amount of surface runoff can be measured. The value
of W can be estimated by determining the water content of soil samples taken before
and after the irrigation application. The calculated value should be checked with the
amount of water consumed by the crop (ETc) in the previous period, which can be
estimated by several methods (FAO, 1977 and 1998). Where relevant, precipitation
should also be considered (FAO, 1974).
However, soil sampling is a tedious procedure that can be avoided by taking the
period equal to an irrigation cycle. Just before irrigation, the soil has dried out; whereas
just after irrigation, it is at field capacity. Thus, a period from before the first to before
the second watering, or one from after the first until after the second, will have W
0, and Equation 1 reads:
(2)
where:
ETc = consumptive use during the irrigation cycle (mm).
Once ETc in that period has been estimated and irrigation and runoff losses have
been measured, R can be determined.
Example
Data from irrigation evaluations made in an pilot area of an irrigation scheme, situated
in northeast Spain, show that on average 90 mm of water is applied by basin irrigation
in the peak period, with an interval between two consecutive waterings of 12 days.
Surface runoff is negligible (levelled field with small bunds) and direct evaporation
losses during the irrigation application are about 3 mm. The consumptive use in the

Guidelines and computer programs for the planning and design of land drainage systems

132

peak period is about 66 mm (ETc 5.5 mm/d). Therefore, deep percolation is about
21 mm and the average value in the period considered is 1.75 mm/d.
PREDICTING DEEP PERCOLATION IN NEW IRRIGATION PROJECTS
Where the irrigation and drainage systems are designed jointly in new developments,
the amount of expected percolation can be determined during the calculation of
irrigation requirements from water retention data:
(3)
being:
(4)
where:
ea = ETc/I = application efficiency (0.001.00), which represents the ratio between
the amount of water consumed by crops and the gross application
depth;
= average thickness of the rootzone (m);
Zr
fc
= soil water retained at field capacity (m3/m3);
i
= minimum soil water fraction that allows for non-stress of the crop
(m3/m3).
Where the i value is unknown, the amount of water readily available to the crops
can be estimated as approximately half the interval between field capacity and the
permanent wilting point:
(5)
where:
wp = soil water retained at wilting point (m3/m3).
For this calculation, an average value of ea must be assumed (see below).
ESTIMATIONS WHERE NO FIELD DATA ARE AVAILABLE
In the planning phase, field data for the project area are usually scarce or non-existent.
In these cases, tentative values for ea and R can be used from literature.
In 1980, FAO provided information on water management from irrigated lands
of arid zones (FAO, 1980). These guidelines considered only readily obtainable data,
such as soil texture and irrigation method and some qualitative information on water
management at the field level (Table A6.1).
TABLE A6.1

FAO guidelines to estimate the values of ea and R

Irrigation method

Application practices

Soil texture
Fine

Coarse

Fine

ea (%)
Sprinkler

Coarse

R (%I)

Daytime application; moderately strong

wind

60

60

30

30

Night application

70

70

25

25

80

80

15

15

60

45

30

40

Trickle
Basin

Poorly levelled and shaped

Well levelled and shaped

75

60

20

30

Furrow & border

Poorly graded and sized

55

40

30

40

Well graded and sized

65

50

25

35

Source: Adapted from FAO, 1980.

Annex 6 Estimating recharge due to irrigation

133

TABLE A6.2

Estimated values for deep percolation

Application method

Distribution
uniformity

Water application efficiency

Tanji & Hanson, 1990

SJVDIP, 1999

Estimated deep
percolation

(%)
Sprinkler
Periodic move

7080

6580

7080

1525

Continuous move

7090

7585

8090

1015

Solid set

9095

8590

7080

510

Drip/trickle

8090

7590

8090

520

Furrow

8090

6090

7085

525

Border

7085

6580

7085

1020

Basin

9095

7590

Surface irrigation

520

Note: Estimates for deep percolation were made on the basis of the following assumptions: no surface runoff under drip and
sprinkler irrigation; daytime evaporation losses can be up to 10 percent sprinkling and 5 percent during night irrigation; tailwater
in furrow and border irrigation can be up to 10 percent and evaporation losses up to 5 percent; no runoff is expected in basin
irrigation and evaporation losses up to 5 percent (FAO, 2002).
Sources: Tanji and Hanson, 1990; SJVDIP, 1999.

In the past 20 years, considerable efforts have been made to improve irrigation
application efficiencies in order to save water. Table A6.2 shows data from well-designed
and well-managed irrigation systems in California, the United States of America, and
potential maximum values for application efficiencies determined in irrigation evaluations
in the San Joaqun Valley Drainage Implementation Program as mentioned in FAO
(2002).
Tables A6.1 and A6.2 contain data from different types of systems and management.
According to the expectations of a specific project area, the order of magnitude for
a first approach to deep percolation can be estimated with the help of these tables.
However, sensitivity analyses with various values should be performed in order to
see the consequences in case the estimates are not correct. In addition, after the first
parts of the irrigation system have been constructed, a direct verification in the field is
recommended.
REFERENCES
FAO. 1974. Effective rainfall, by N.G. Dastane. FAO Irrigation and Drainage Paper No. 25.
Rome. 68 pp.
FAO. 1977. Crop water requirements, by J. Doorenbos & W.O. Pruitt. FAO Irrigation and
Drainage Paper No. 24, revised edition (reprinted 1992). Rome. 144 pp.
FAO. 1980. Drainage design factors. FAO Irrigation and Drainage Paper No. 38. Rome. 52 pp.
FAO. 1998. Crop evapotranspiration guidelines for computing crop water requirements, by
R.G. Allen, L.S. Pereira, D. Raes & M. Smith. FAO Irrigation and Drainage Paper No. 56.
Rome. 300 pp.
FAO. 2002. Agricultural drainage water management in arid and semi-arid areas, by K.K. Tanji
& N.C. Kielen. FAO Irrigation and Drainage Paper No. 61. Rome. 188 pp.
San Joaquin Valley Drainage Implementation Program (SJVDIP). 1999. Source reduction
technical committee report. Sacramento, USA, Department of Water Resources 33 pp.
Tanji, K.K. & B.R. Hanson. 1990. Drainage and return flows in relation to irrigation management.
In B.A. Stewart & D.N. Nielsen, eds. Irrigation of agricultural crops. Agronomy Monograph
No. 30. Madison, USA, American Society of Agronomy.

135

Annex 7

Leaching for salinity control

THE WATER AND SALT BALANCES
During rainfall, snowmelt or
FIGURE A7.1
irrigation, part of the water is lost
Water
balance
of
an
irrigated soil with macropores
by runoff and evaporation, but a
considerable part enters the soil
ET
P
In
and is stored there. This storage
EvapoPrecipitation
Irrigation
is partly taken up by plant roots,
transpiration
to soil
while any excess drains below the
I"= f i I n
(1-f i )I n
rootzone. On the other hand, in
Rootzone
dry periods, the rootzone may
Flow through
receive water from deeper layers by
macropores
FL=R"/I n "
capillary rise, especially where the
water table is shallow and drainage
(1-f r )R
R"=frR
Capillary
poor. Monthly water balances are
rise
generally sufficiently revealing for
G
water table control, while annual
Leaching R
soil salinity balances usually provide
Subsoil or water table
enough information for soil salinity
control.
Coupled to this water balance, a balance can be made for soluble salts. They enter
in tiny amounts through rain or snow, and in much larger quantities in irrigation
water, even where this is considered as being of good quality. In the soil, these salts are
concentrated by drying out, whereas plant roots take up water, but exclude the entry
of salts. This increase in concentration should not be allowed to reach harmful levels
for crop growth. This requires:
adequate leaching: the inflow of water during a year must generate enough
leaching to keep the salinity levels down;
adequate natural or artificial drainage to allow removal of the leacheate, and a safe
depth of the water table to prevent harmful capillary rise of saline water;
irrigation water of good quality, or, where poor, an extra amount to provide an
increased leaching.
Therefore, a first estimate can be made by estimating the annual balances.
However, a complication is that not all water entering or leaving the soil is effective
in leaching. Especially in many clay soils under surface irrigation (basin, furrow or
border), part of the water passes downward through cracks and other macropores
without contributing much to the removal of salts.
LEACHING FRACTION OF AN IRRIGATED FIELD
This is expressed by a leaching efficiency: the part of the water that is effective. There
are two such coefficients: for the surface (fraction of the entering water, fi); and at the
bottom of the rootzone (fraction of the percolating water, fr).
For irrigated lands, where water conservation and salinity control are required, it is
necessary to compare the actual amounts of deep percolation produced by irrigation
with the leaching required to ensure soil salinity control. The first step is to determine
the actual value of the leaching fraction, which can be taken as a first approximation
as:

136

Guidelines and computer programs for the planning and design of land drainage systems

(1)
However, to allow for flow through macropores it is better defined as:
(2)
This flow usually goes directly to the subsoil. In this case (Figure A7.1):
or

(3)

Therefore, one of the two coefficients is sufficient.

In these equations:
fi = leaching efficiency coefficient as a fraction of the irrigation water applied;
fr = leaching efficiency coefficient as a function of the percolation water;
In = net amount of irrigation water (amount infiltrating into soil) (mm);
LF = required leaching fraction;
R = amount of percolation water (mm).
As I is usually much larger than R, so fi is considerably larger than fr. The leaching
efficiency coefficient fr was defined by Boumans in Iraq (Dieleman, 1963), and later fi
was introduced by Van Hoorn in Tunisia (Van Hoorn and Van Alphen, 1994). In the
literature, both values are used. The fi coefficient is commonly used. This coefficient
depends on soil texture and structure as well as on the irrigation method. It is higher
(0.951.0) in well-structured loamy soils than in heavy clay cracking soils (< 0.85). It is
also higher with sprinkler irrigation than with surface irrigation, and close to 1 under
drip irrigation. Where needed, fr can be found from Equation 3.
Therefore, the actual value of the LF depends on soil characteristics, the irrigation
method and the specific water management practised by farmers.
Example
The data in the example in Annex 6 show that farmers apply a net irrigation of about
87 mm during the peak irrigation season, and that about 21 mm of this amount
percolates below the rootzone. It was also determined that about 6 percent of the
infiltrated water flows directly through cracks without mixing with the soil solution (fi
0.94 and fr 0.75). This means that during this irrigation cycle farmers are irrigating
with an LF of about 0.2. Following a similar approach, the average LF during the
irrigation season can be obtained where the total values of In and R are available.
LEACHING REQUIREMENTS IN TERMS OF A MINIMUM LF
In order to control soil salinity in irrigated lands, a minimum LF is required. This can
be calculated where the value of the electrical conductivity of the irrigation water (ECi)
and the salt tolerance of the crop are known. One option is to apply the approach
developed by Van Hoorn and Van Alphen (1994) based on the water and salt balances
in equilibrium status. In this approach, it was considered that water extraction by
crops decreases within the rootzone from 40 percent of the total in the top quarter to
10 percent in the deepest quarter (FAO, 1985). Following this approach, a relationship
between the ECi and the average soil salinity in the rootzone (expressed in terms of the
electrical conductivity of the saturated paste [ECe]) can be obtained for several values
of the LF (Figure A7.2). Similar graphs can be obtained from water and salt balances
derived considering other water extraction models adapted to specific local conditions,
as crop root distribution is affected severely by soil properties and by irrigation water
management.
By means of Figure A7.2, the minimum LF to control soil salinization (caused by
the salts applied with irrigation water with certain ECi) can be determined once the

Annex 7 Leaching for salinity control

threshold value of ECe that must not

be exceeded in the rootzone has been
established from crop salt tolerance
data. Data provided by Maas and
Grattan (1999) about crop salt
tolerance can be used (FAO, 2002).

137

FIGURE A7.2

Relationship between ECe and ECi for different leaching

fractions
ECe (dS / m)

10

LF
=0
.05

Example
9
Following the example of the
8
previous section, it is possible to
7
10
calculate the minimum LF required
0.
F=
6
L
to control the salt buildup caused by
.15
=0
0
LF =0.2
5
the salts applied with the irrigation
LF
0
water, whose salinity content in
4
0.4
LF=
terms of ECi is 0.6 dS/m. If maize
3
40 %
is the most salt-sensitive crop of the
30
20
2
10
cropping pattern, and its tolerance
1
threshold in terms of ECe is 1.7 dS/
EC i (dS / m)
m, then a minimum LF of 0.05 is
1
2
3
4
5
required to control soil salinity
(Figure A7.2).
Assuming that the average LF during the irrigation season is 0.2 and the minimum
LF is 0.05, it can be concluded that no salt buildup should be expected in the rootzone,
and even the irrigation application efficiency might be increased while keeping soil
salinity under control.
In irrigated lands, it is possible to check whether the actual value of the LF satisfies the
minimum LF necessary to control soil salinity. Therefore, if the amount of percolation
water is enough to cover the leaching requirements, water might be saved by improving
the application efficiency. If not, the leaching requirements must be calculated.
LEACHING REQUIREMENTS
Once the minimum LF is known, the long-term leaching requirements, for example,
during the irrigation season, can be calculated by means of the salt equilibrium
equation developed by Dieleman (1963) and later modified by Van Hoorn and Van
Alphen (1994):
R * = (ETc Pe )

1 fi (1 LF )
fi (1 LF )

(4)

where:
ETc = actual crop evapotranspiration (mm);
Pe = effective precipitation (mm);
R* = long-term leaching requirement (mm).
Therefore, the net irrigation requirement (I) is:
I = (ETc Pe ) + R *

(5)

Example
This example uses the case of the irrigated lands mentioned in the previous example (in
which fi = 0.94) and assumes that farmers need to irrigate with groundwater with an ECi
of 1.5 dS/m. If they still wish to grow maize in the soil of the previous example, they
will need to irrigate with an LF of 0.3 (Figure A7.2). If the net irrigation requirement
(ETc - Pe) during the irrigation season is about 560 mm, at least 290 mm will be required
to leach the salts accumulated in the rootzone. The net irrigation requirement will be

Guidelines and computer programs for the planning and design of land drainage systems

138

850 mm. If the actual LF is 0.2, about 185 mm of leaching can be obtained during the
irrigation season (Equation 4). Therefore, the leaching deficit will be about 105 mm
(290 - 185).
Where slightly soluble salts (e.g. gypsum, and magnesium and calcium carbonates)
are present in the irrigation water, the leaching requirement is calculated first for the
soluble salts. Then, the small contribution of the slightly soluble salts to the total soil
salinity is added (Van der Molen, 1973). For average salt contents, the total solubility of
gypsum and carbonates is about 40 meq/litre, which is equivalent to an EC of 3.3 dS/m.
Where bicarbonates predominate in the irrigation water, it is advisable to decrease the
sodium adsorption ratio (SAR) by increasing the calcium content of the soil solution
by applying gypsum (520 tonnes/ha).
Once long-term soil salinity increases are no longer expected, a check should be
made on the short term in order to be certain that the salt content of the soil solution
does not exceed the threshold value of the crop salt tolerance. For this purpose, the
salt storage equation derived for predicting the buildup of soil salinity on a weekly or
monthly basis can be used (Van Hoorn and Van Alphen, 1994). The variation of salinity
in the short term (z) can be calculated thus:
z = z2 z1 =

fi In ECi fr REC1
f R
1+ r
2Wfc

(6)

where:
EC1 =
Wfc
z1
z2

z1
Wfc

= initial soil electric conductivity (deciSiemens per metre);

= moisture content at field capacity (mm);

= salt content in the rootzone at the start of the period (mm.dS/m);
= salt content in the rootzone at the end of the period (mm.dS/m).

OPTIONS TO COVER THE LEACHING REQUIREMENTS

Where the actual value of the LF does not satisfy the minimum LF, options should be
considered to cover the leaching deficit.
In monsoon and temperate regions, the salt content in the rootzone may increase
during the irrigation season. However, excess rainfall after the irrigation period will
supply enough percolation water to leach out the salts accumulated in the rootzone.
In this way, the salt content at the beginning of the next irrigation season will be
sufficiently low to prevent secondary salinization.
Example
In the case described in the previous example, 100 mm of excess rainfall in winter might
provide the percolation required to cover the leaching deficit. Therefore, even when
irrigating with water with an ECi of 1.5 dS/m, the soil salinity might be controlled on
an annual basis under actual irrigation management.
However, where no effective precipitation is available for leaching, as is usually the
case in arid and semi-arid zones, the leaching deficit must be covered by increasing
the annual allocation of irrigation water. To cover uniformity deficiencies in water
distribution over the irrigated field, the amount of percolation water should exceed the
leaching requirements:
I = (ETc Pe ) + aR

(7)

The a coefficient may vary from 1.15 to 1.20 if irrigation uniformity is fairly
appropriate.

Annex 7 Leaching for salinity control

If, under the current irrigation management, the leaching requirements are not
satisfied (R aR*), there are two options: grow crops that are more tolerant of salinity
and in this way reduce the minimum LF; or find out how to cover the leaching deficit.
In the latter case there are two possibilities: remove the accumulated salts before
sowing the next crop by applying irrigation water; or split up the leaching requirement
during the irrigation period by increasing each irrigation application.
EFFECTS OF LEACHING FOR SALINITY CONTROL ON SUBSURFACE DRAINAGE
DESIGN
Where the leaching requirements are covered by the actual irrigation management
or after the cropping season by rainfall or out-of season leaching irrigation, salinity
control does not affect the drainage coefficient used for subsurface drainage design.
However, if more water has to be added with each application in order to increase the
LF, salinity control affects subsurface drainage design because the drainage coefficient
must also be increased.
The option of increasing the irrigation allocation depends on the availability of
water resources during or at the end of the growing season. It also depends on the
internal drainage capacity of the soils. Coarse-textured soils permit leaching fractions
of 0.150.25, while in fine-textured soils with low permeability the LF should be lower
than 0.10 because of their limited internal drainage (unless rice is grown). In addition,
the environmental effect of increasing the volume of drainage water on drainage
disposal should be considered. Thus, growing more salt-tolerant crops is frequently a
better option than using more water and increasing field and disposal drainage needs.
Controlling soil salinity caused by capillary rise generally does not increase the
drainage coefficient. This is because it is dependent on adopting a suitable depth of the
groundwater table and maintaining a downward flow of water during the irrigation
season. Where leaching is required in order to remove the accumulated salts in the
rootzone, water is generally applied before the start of the cropping season.
REFERENCES
Dieleman, P., ed. 1963. Reclamation of salt affected soils in Iraq. ILRI Publication 11.
Wageningen, The Netherlands. 174 pp.
FAO. 1985. Water quality for agriculture, by R.S. Ayers & D.W. Westcot. FAO Irrigation and
Drainage Paper No. 29 Rev. 1. Rome. 174 pp.
FAO. 2002. Agricultural drainage water management in arid and semi-arid areas, by K.K. Tanji
& N.C. Kielen. FAO Irrigation and Drainage Paper No. 61. Rome. 188 pp.
Maas, E.V. & Grattan, S.R. 1999. Crop yields as affected by salinity. In R.W. Skaggs & J. Van
Schilfgaarde, eds. Agricultural drainage. Agronomy Series 38. Madison, USA.
Van der Molen, W.H. 1973. Salt balance and leaching requirement. In: Drainage principles and
applications, pp. 59100. ILRI Publication 16, Vol. 2. Wageningen, The Netherlands, ILRI.
Van Hoorn, J.W. & Van Alphen, J.G. 1994. Salinity control. In H.P. Ritzema, ed. Drainage
principles and applications, pp. 533600. 2nd edition. ILRI Publication 16. Wageningen, The
Netherlands, ILRI.

139

141

Annex 8

Procedures for determining soil

hydrological characteristics in
drained lands

PROCEDURE FOR DETERMINING HYDRAULIC CONDUCTIVITY

Steady-state flow
Where water flows toward the drains under steady-state conditions, an average value
of the hydraulic conductivity can be obtained from:
(1)
where:
B = drain length (m);
D = average thickness of the horizontal flow region (m);
hh = hydraulic head for horizontal flow (m);
K = hydraulic conductivity (m/d);
L = drain spacing (m);
Q = outflow (m3/d);
= specific discharge (m/d).

hh=h1-h2

hap=h3-h4

hr=h2-h3

In Equation 1, L is a design parameter that is known; q is calculated from the value

of Q measured at the drain outlet; hh is measured by difference in piezometer readings
in tubes laid midway between two drains (h1) and at some distance from the drain
(h2), outside the zone where radial flow is important, as shown in Figure A8.1. The
radial flow in the vicinity of the
drain has been excluded from the
FIGURE A8.1
measurements.
Piezometer line for determining the components of the total
For shallow aquifers (D < L/4),
hydraulic head in a drained soil
D approaches the real thickness of
1'
2'
3' 4 3
2
1
the permeable layer. However, for
deeper ones, the maximum value for
D is L/3. Where the D value has been
determined by augering, an average
value of K can be calculated with
Equation 1.
h1
Table A8.1 shows an example of
h2
h3
the calculation of KD values from
x=0
x= 1 L
x= 1 L
8
2
groundwater-level observations in
piezometers laid midway between
two drains (z25) and in the vicinity of
the drain (z6.5), for drains laid at 50-m
spacings and 1.8 m deep in a pilot field
of peat soils with a sandy substratum
severely recharged by seepage.

Guidelines and computer programs for the planning and design of land drainage systems

142

TABLE A8.1

Determination of KD values from groundwater-level observations in a drained soil with a sandy substratum
Drain
no.

Period of
observations (1984)

(m)

(m)

(m)

(m)

13

JanuaryMarch

0.95

0.97

1.07

0.85

AprilJune

1.03

1.04

1.14

0.77

JulyOctober

1.08

1.09

1.17

January March

0.86

0.89

0.97

AprilJune

0.95

0.97

January March

0.52

AprilMay

0.57

14
16

z25

z12.5

z6.5

h1 = 1.8-z25

hh

KD

(m)

(m)

(mm/d)

(m2/d)

0.73

0.12

22.3

58.1

0.66

0.11

19.5

55.4

0.72

0.63

0.09

17.0

59.0

0.94

0.83

0.11

22.6

64.2

1.05

0.85

0.75

0.10

18.0

56.3

0.56

0.62

1.28

1.18

0.10

21.1

65.9

0.60

0.66

1.23

1.14

0.09

18.0

62.5

FIGURE A8.2

Example of qt/ht relationships during the drawdown of the

water table in non-steady flow towards drains laid on an
impervious layer in a silt-clay soil
q/h (d -1x 10-3 )
7

July 1975

h2 = 1.8-z6.5

The average KD value calculated

from observations made in three
drains over ten months was 60 m2/
d. If the sandy layer in which
the drains are laid has an average
thickness of about 8 m, the average
value for the hydraulic conductivity
of the sandy layer is 7.5 m/d.

March 1975
5

February 1975

h (m)

0
0

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2

Source: Adapted from Martnez Beltrn, 1978.

qt =

3.46 K
2

ht

Non-steady-state flow
In drained lands where laterals are laid
on the impervious layer, water flow
is generally non-steady, especially
after an irrigation application or
heavy rainfall. However, the average
value of the hydraulic conductivity
of the permeable layer can be
calculated from observations of the
drawdown of the water table, where
the phreatic level has an elliptic
shape. Under these conditions, the
Boussinesq equation for the specific
discharge reads:

(2)

where:
qt = specific discharge at time t (m/d);
ht = hydraulic head midway between drains at time t (m).
Therefore, if the function qt/ht = f(ht) is represented graphically, with data from
observations made during several drainage periods, straight lines can be obtained, as
those represented as an example in Figure A8.2.
The slope of the qt/ht = f(ht) function is equal to:
tg =

3.46 K
L2

From Equation 3, K values can be obtained, as shown in Table A8.2.

(3)

Annex 8 Procedures for determining soil hydrological characteristics in drained lands

143

TABLE A8.2

Calculation of hydraulic conductivity with the Boussinesq equation

Drawdown of the
groundwater level
(m)

Correlation coefficient
qt/ht = f(ht)

tg 10-3

K
(m/d)

February 1976

0.301.10

0.96

4.05

0.47

JulyAugust 1976

0.101.10

0.91

8.67

1.00

JanuaryFebruary 1977

0.601.10

0.97

3.81

0.44

JuneJuly 1977

0.501.00

0.94

4.80

0.55

Period of observations

Source: Martnez Beltrn, 1978.

Results from Table A8.2 show K values of about 0.5 m/d where the groundwater level
is below the top layer (030 cm). A higher value of 1 m/d was obtained when the water
level was close to the ground surface. However, in this case, the correlation coefficient
was lower than in the previous cases (probably because of an almost flat shape of the
water table and because of the high hydraulic conductivity of the top layer).
DETERMINING RADIAL RESISTANCE
Resistance to steady-state radial flow towards drains installed above the impervious
layer can also be determined from observations in drained lands:
Wr =

hr

(4)

qL

where:
hr = hydraulic head for radial flow (m);
Wr = radial resistance (d/m).
In Equation 4, hr is measured by the difference in piezometer readings in tubes laid
at some distance from the drain (h2) and close to the drain trench (h3), as shown in
Figure A8.1.
Table A8.3 shows an example of calculation of Wr values from water-level
observations in piezometers laid in the vicinity of the drain (z6,5) and close to the drain
(z0), for drains laid at 50-m spacings and 1.8 m deep in a sand layer.
Results from three drains observed during different periods show an average radial
resistance of 0.24 d/m.
PROCEDURE FOR DETERMINING THE DRAINABLE PORE SPACE
For drained lands, the value of the layer above drain level can be measured from the
drawdown of the water table (determined by piezometer recording) and the amount
of water drained in the period considered (calculated from measurements of the drain
discharge). The restrictions are that evaporation and seepage to or from deeper layers
must be low and can be ignored relative to the drain discharge.
TABLE A8.3

Determination of Wr from observations in a drained soil with a sandy substratum

z6,5

z0

h2 = 1.8z6.5

hr

Wr

(m)

(m)

(mm/d)

(d/m)

0.42

0.31

22.3

0.28

0.42

0.24

19.5

0.25

0.63

0.47

0.16

17.0

0.19

Drain
no.

Period of observations
(1984)

(m)

(m)

(m)

13

JanuaryMarch

1.07

1.38

0.73

AprilJune

1.14

1.38

0.66

JulyOctober

1.17

1.33

14
16

h3 = 1.8-z0

January March

0.97

1.26

0.83

0.54

0.29

22.6

0.26

AprilJune

1.05

1.26

0.75

0.54

0.21

18.0

0.23

January March

0.62

0.87

1.18

0.93

0.25

21.1

0.24

AprilMay

0.66

0.87

1.14

0.93

0.21

18.0

0.23

144

Guidelines and computer programs for the planning and design of land drainage systems

TABLE A8.4

Calculation of the value from the water balance in drained lands

Period of
observations

Drawdown of the
water level

Dr

(m)

(mm)

(mm)

(%)

(%)

January 1975

0.550.80

11.2

219

5.1

4.3

0.800.95

5.3

156

3.4

0.951.10

4.7

125

3.8

0.951.10

4.8

97

4.9

1.101.20

2.1

46

4.6

0.751.10

7.1

169

4.2

0.851.20

10.2

288

3.5

February 1976
January 1977

4.7
3.9

Source: Martnez Beltrn, 1978.

Therefore, if the recharge to the water table and natural drainage are negligible and
there is no depletion of the water table from plant roots in the time interval selected, the
drainable pore space can be found from:
(5)
where:
Dr = amount of drainage water converted to an equivalent surface depth (mm);
= drainable pore space;
h = average drawdown of the water table in the time considered (mm).
Dr and h must be expressed in the same units.
To determine the average value, it is only necessary to measure, during the interval of
time selected, the average drawdown of the water table from piezometer readings and the
amount of water drained in the same period. The drainable pore space is a dimensionless
fraction, often expressed as a percentage, as in Table A8.4. Table A8.4 shows an example
calculation of the average value of a silty-clay soil, with data from observations made
during three consecutive winters.
The results of this table show the tendency of to decrease with soil depth. For
example, the 1975 observations show a value of 5.1 percent for a soil layer with a
prismatic structure and about 3.9 for the deeper, less-structured soil layer. However,
for drain spacing calculations an average value of 4.3 percent can be considered.
The average value calculated with the results of the following years was of the same
magnitude.
REFERENCES
Martnez Beltrn, J. 1978. Drainage and reclamation of salt affected soils in the Bardenas area,
Spain. ILRI Publication 24. Wageningen, The Netherlands, ILRI. 321 pp.

145

Annex 9

Procedure for deriving drainage

design criteria from drained lands
SUBSURFACE DRAINAGE
COEFFICIENTS
From observations of the groundwater level and measurements of
drain discharge, hydrographs such as
those in Figure A9.1 can be drawn.
This example (from a flat coastal
area in eastern Spain) shows that
during dry periods (from mid-June
to late September), in the absence of
irrigation, the subsurface drainage
flow towards the observed drain
was steady, with a drain discharge
of about 17 mm/d, due to seepage.
However, in winter and spring, the
drainage system was also recharged
by percolation of rainfall, and then
the water flow was non-steady.
With this information, sound
drainage criteria can be formulated
for steady-state flow drainage design.
If in addition to seepage, during the
irrigation season, there is a recharge
of about 1 mm/d from irrigation
losses, a drainage coefficient of
18 mm/d will be required in order
to control the water table during
the dry period. However, if after
heavy rainfall, high water tables are
affecting winter crops or hampering
soil trafficability, the drain spacing
calculated for steady flow should be
checked for non-steady conditions.
In irrigated lands without such
high seepage, water flow towards
drains is generally non-steady, as
Figure A9.2 shows. Information
from drainage periods such as those
shown in Figure A9.2 is useful for
determining the magnitude of the
rise of the water table after irrigation
and further drawdown during the
interval between two consecutive
irrigation applications.

FIGURE A9.1

Water depth and drain discharge hydrographs determined by

observations in a drainage experimental field (drained peat
soil under considerable seepage)
z (m)
0.1
0.2
0.3

q (mm/d)
34
32
30

0.4
0.5
0.6

28
26
24

0.7
0.8
0.9

22
20

1.0
1.1
1.2
1.3
1.4
1.5

18
16
14
12
10
8
6

1.6
1.7
1.8

4
2
0
J

Groundwater level
Discharge

FIGURE A9.2

Drawdown of a water table after irrigation to reclaim saline

soils
h (m)

Ground surface

q (mm/d)

1.2

12

1.0

10

0.8

0.6

0.4

4
2

0.2
0

Drain level
10 15 20

25

30

January 1975

10

15

20

25

February 1975
Groundwater level
Discharge

Source: Adapted from Martnez Beltrn, 1978.

0
t (d)

Guidelines and computer programs for the planning and design of land drainage systems

146

FIGURE A9.3

Example of drainage discharge after irrigation

q (mm/d)
3.00
2.50
2.00
1.50
1.00
0.50
0.00
08-03-03

09-03-03

10-03-03

11-03-03

12-03-03

13-03-03

14-03-03

Source: FAO/IMTA, 2004.

However, for irrigated lands,

the actual non-steady drainage
criteria can be translated into more
or less equivalent steady-state
drainage criteria. For example, the
hydrograph in Figure A9.3 shows
that after an irrigation application,
discharge decreases from a maximum
value of about 2.5 mm/d to zero (just
before the next irrigation). However,
the average discharge during the
drainage period was about 1 mm/d.
Therefore, this latter discharge can
be used as the drainage coefficient
for drain spacing calculations using
steady-state equations.

FIGURE A9.4

DESIGN DEPTH TO THE HIGHEST

WATER TABLE
The relationship between the
average depth to the water table and
+
crop yields and trafficability or the
duration and intensity with which
groundwater levels exceed a cropspecific critical depth during the
+
growing season can also be estimated
from observations in drained lands.
Table A9.1 shows groundwater
Alfalfa
depth
data from four plots with
Maize
+ + +
different drainage conditions and
+
Estimated
their impact on yields of irrigated
+
maize and alfalfa.
Table A9.1 also includes the
Z (cm)
SDW
value, as used in the Dutch
50
70
80
90
100
60
polders. It is the sum of days with
waterlogging during the period
considered (Sieben, 1964). In this case, the SDW50 (sum of days with less than 50 cm
depth) is also a good measure for crop damage. In the Dutch polders, SDW30 (less than
30 cm depth) is usually taken for field crops.

Relationship between the average depth of the water table

and maize and alfalfa yields
Y
1.0

0.8

0.6

0.4

0.2

TABLE A9.1

Maize and alfalfa yields compared with data of the groundwater table
Period (1977)

Consecutive days in which the groundwater level was above the depth indicated (cm)
25

50

75

100

25

50

75

100

25

50

75

100

25

50

75

100

June

20

10

30

22

30

20

30

30

July

10

10

31

10

25

31

19

31

31

August

16

10

31

14

28

31

24

30

31

September

23

17

30

14

30

16

19

41

71

Alfalfa yield (kg/

ha) and relative
yield

12 195

7 600

5 780

5 415

1.00

0.62

0.47

0.44

Maize yield (kg/

ha) and relative
yield

5 800

4 000

1 730

1 180

1.00

0.69

0.30

0.20

SDW50

Source: Adapted from Martnez Beltrn, 1978.

Annex 9 Procedure for deriving drainage design criteria from drained lands

Although under irrigation the water level varies with time, the average depth of
the water table is a good indicator concerning crop yields. Figure A9.4 shows the
relationship between the relative crop yield (Y) and the average depth of the water
table ( ) during the irrigation season, as per the data in Table A9.1.
Although data from only one irrigation season are not sufficient to obtain a
statistically sound relationship, these results are useful for providing practical guidance
to be confirmed later with further information. It seems that an average depth of
85 cm is critical for maize and alfalfa, which were the most relevant irrigated crops in
the study area. In this case, the groundwater depth criterion is dominant because no
long dry fallow periods or periods with frequent shortages of irrigation water occur.
Where this is not the case, especially where the groundwater is rather salty, deeper
groundwater levels during such extended dry periods are required in order to avert soil
salinization by capillary rise.
The data in Table A9.1 also show that short periods of high water tables are not
harmful for the above-mentioned crops.
In the Dutch polders, with a humid climate, no appreciable damage to crops
was found where during heavy rains in winter the groundwater did not rise above
0.30 m depth below the surface, provided that it receded within a few days. Higher
groundwater levels led to slaking of the ploughed layer, causing more permanent
anaerobic conditions and damage to field crops. These silty-clay soils needed a drainage
depth of 1.20 m in order to keep the average levels low enough.
REFERENCES
FAO/IMTA. 2004. Evaluacin de sistemas de drenaje en el Distrito de Riego 041, Ro Yaqui,
Son., y en el Distrito de Riego 076, Valle del Carrizo, Sin., Mxico.
Martnez Beltrn, J. 1978. Drainage and reclamation of salt affected soils in the Bardenas area,
Spain. ILRI Publication 24. Wageningen, The Netherlands, ILRI. 321 pp.
Sieben, W.H. 1964. De invloed vande ontwateringstoestand op stikstofhuishouding en
opbrengst. Landbouwkundig Tijds., 76: 784802.

147

149

Annex 10

Calculations regarding elements of

the main drainage system
OPEN CHANNELS AND THEIR CROSS-SECTIONS
For open channels, Mannings formula is widely used:
(1)
being:
;
;
;
= average flow velocity over the cross-section A;
where (see Figure A10.1):
A = cross-sectional area of flow (m2);
b = bottom width (m);
= 1/n = roughness coefficient (m1/3/s);
Km
n = 1/Km roughness coefficient (s/m1/3);
Q = discharge (m3/s);
R = hydraulic radius (m);
s = hydraulic gradient (-);
u = wetted perimeter (m);
v = average flow velocity (m/s);
y = water depth (m);
= coefficient in side slope (v:h) 1:.
The roughness coefficient Km
depends on factors such as the
irregularities of the drain bed and side
FIGURE A10.1
slopes, amount of vegetation, irregular
Cross-sections of an open ditch (A) and a large channel (B)
alignment and hydraulic radius of the
A
open drain. Values range from 50 for
large channels in bare earth, to 20 for
Level at design discharge
open drains two-thirds choked with
y
Slope (v:h) 1:1.5
vegetation, to less than 10 for entirely
b
choked ones. Table A10.1 lists design
B
values for normally maintained
Level at design discharge
channels. For the coefficient Km, the
Slope (v:h) 1:1.5
Level at frequent discharge
following equations for such open
waterways (with some vegetation) are
y y
Water
1 2
Slope (v:h) 1:1.5
used, in which it is supposed that the
channels have been cleaned before the
b
onset of the wet season (so that they
are in a reasonable condition).

Guidelines and computer programs for the planning and design of land drainage systems

150

TABLE A10.1

If y < 1 then

(2a)

else

(2b)

Design parameters for open drains

Drain size

Water
depth y
(m)
< 0.75

Ratio b:y Soil

texture

Mannings
Km (m1/3s-1)

The ratio of bottom width (b)

to water depth (y) should remain
Medium
0.751.5
23
preferably within certain limits
clayey
20
0.050
(Table A10.1). Where this ratio is
Large
> 1.5
34
4050
0.0200.025
Sources: Adapted from ILRI, 1964; and from Smedema, Vlotman and Rycroft,
known, the required cross-section
2004.
can be calculated with the above
formulae.
TABLE A10.2
The average flow velocity v over
Maximum average water velocity and bank slopes for open
ditches
the cross-section should not be so
Soil type
v
Bank v:h
high that erosion of the bottom or
(m/s)
banks occurs. Table A10.2 gives
Heavy clay
0.600.80
1:0.75 to 1:2
some values for the maximum
Loam
0.300.60
1:1.5 to 1:2.5
average flow velocities and also
Fine sand
0.150.30
1:2 to 1:3
Coarse sand
0.200.50
1:1.5 to 1.3
the recommended side slopes for
Tight peat
0.300.60
1:1 to 1:2
trapezoidal cross-sections.
Loose peat
0.150.30
1:2 to 1:4
For safety, it is advisable to check
Source: Adapted from ILRI, 1964.
the behaviour of the system at a
larger discharge. At 1.52 times design discharge, some inundation may be allowed to
occur in low places, but disasters and extensive inundation should not occur.
Small

12

sandy

20

0.050

clayey
sandy

15
30

0.067
0.033

Depth and freeboard

The depth of a drainage channel equals:
(3)
where:
F = freeboard (m);
y = water depth (m);
Zc = collector depth below soil surface (m).
The freeboard F must be such that at design discharge the outlets of any subsurface
drains, including pipe collectors, are just above or equal to the drainage-channel water
level, although a slightly higher water level can be tolerated temporarily. This usually
leads to water levels of 12 m below the land surface at design discharge. In arid
regions, drain outlets should remain above the water level, although they may become
temporarily submerged after an infrequent rainfall has caused large surface runoff
volumes to the open drain.
Wind effects
Similar to shallow seas, long canals (> 10 km) may be subject to storm surges when
strong winds blow in the direction of the waterway. However, in most situations, such
wind effects are negligible.
An estimate for storm surges at sea, but also for all kinds of waterways, is:
or

where:
B = length of waterway, in wind direction (km);
v = wind velocity (m/s);

(4)

Annex 10 Calculations regarding elements of the main drainage system

151

g = acceleration of gravity (m/s2);

h = head (m);
x = distance (m);
y = water depth (m);
h = head difference along canal, caused by wind (m);
4.10-6 = coefficient;
0.0004 = coefficient.
For seas and estuaries, the calculation must be numerical, using sections of the same,
or almost the same, depth.
Normal flow and inundation
Where the water level downstream is lower than the upstream water level of an outflow,
channel flow occurs. Depending on the conditions, this channel flow may be streaming
or shooting. This is governed by the FroudeBoussinesq number:
(5)

where:
Fr = FroudeBoussinesq number;
g = 9.81 = acceleration gravity (m/s2);
y = water depth (m);
v = flow velocity (m/s).
For streaming water, it is required that Fr < 1; while for Fr > 1, shooting occurs.
Streaming water is supposed to obey Mannings formula (Equation 1).
If the water level downstream becomes higher than the land surface, overflow and
inundation occur.
Backwater effects
Backwater curves occur near the downstream end of a channel, where it joins other
watercourses with a higher water level or within the reach with a backwater curve effect
upstream of weirs. Upstream, the water will reach a constant equilibrium depth in
accordance with a given flow. However, near the downstream end, the water level will
come under the influence of the fixed downstream level and form a curve upwards
or downwards (Figure A10.2)
depending on whether this level is
higher or lower than the water level
FIGURE A10.2
Convex and concave backwater curves
corresponding with the upstream
equilibrium depth. Complications
Land surface
arise when the land is inundated or
when the channel overflows.
The program BACKWAT is
based on these considerations. This
Water level
program calculates the equilibrium
River
depth by iteration. The calculations
start at the downstream end, where
High
the water level is given. They are
Low
Channel bottom
numerical, with steps in water
depth of a given size. The water
depth diminishes inland if the curve
is convex, and increases inland if
Bottom
concave (Figure A10.2). In the latter
case, overflow may occur upstream.

Guidelines and computer programs for the planning and design of land drainage systems

152

If shooting occurs, the program

terminates.

FIGURE A10.3

Bridge (non-obstructing)

CULVERTS AND BRIDGES

For culverts, there are two types of
head losses, caused by:
Watercourse
Watercourse
Bottom
convergence of streamlines at the
entrance these losses are not
recovered at the exit;
friction losses, occurring at the walls of culverts.
For the former, laws for flow through openings apply. The hydraulic section of a
culvert can be calculated using:
Road

Bridge

Road

Water level

(6)
where:
A = area of the hydraulic section (m2);
g = 9.8 m/s2 is the gravity acceleration;
Q = design discharge (m3/s), preferably increased by a safety factor;
= coefficient that depends on the shape of the entrance and at the exit;
h = head loss along the culvert (m).
The design discharge is often taken some 2550 percent higher than for the upstream
drainage channel. This is because the flexibility of culverts to accommodate for higher
flows without causing structural damage is less than for open waterways. The values
of are about 0.7 for long culverts (2030 m) and 0.8 for short culverts (< 10 m) (ILRI,
1964). Head losses of 5 cm for small structures and 10 cm for large ones are generally
taken (Smedema, Vlotman and Rycroft, 2004). In order to calculate the cross-section of
the structure, in addition to the wet section A, a minimum of 10 cm of clearance should
be added.
The friction losses in culverts are of minor importance for the usual short passages
under rural roads. For longer culverts, the head losses for friction must be added.
Mannings formula is often used, with a Km of 6070 for smooth and 3040 for
corrugated walls.
Bridges are often constructed in such a way that the watercourse passes freely
underneath, in which case they have no influence (Figure A10.3). If the channel is
narrowed by the bridge, Equation 6 may be used, with = 0.80.9 (Smedema, Vlotman
and Rycroft, 2004). Friction losses can be ignored as the influence of the short length of
the narrow passage is small.
WEIRS AND DROP STRUCTURES
The width of freely discharging rectangular weirs and drop structures is calculated with
the formula:
(7)

where:
b = crest width (m);
g = 9.8 m/s2 is the gravity acceleration;
h = head above the crest level (m);
Q = discharge (m3/s);
= contraction coefficient.

Annex 10 Calculations regarding elements of the main drainage system

For submerged discharge the

following equation may be used:

153

FIGURE A10.4

Crest form of weirs

(8)
where:
h1 = upstream water head (m);
h2 = downstream head (m);
h = h1 - h2 = available head (m).
Broad
The values of the coefficients
in Equation 7 and 8 are mostly
determined by the width/shape of
the weir crest (broad or sharp, as
Sharp
shown in Figure A10.4) and by the
nature of the approach flow (degree
of streamline contraction and entry
turbulence). For similar weirs, the
values are in principle the same
for both equations. Values for semi-sharp crested weirs commonly used in drainage
channels (e.g. stop-log weirs) are generally in the order of 1.01.1 (Smedema, Vlotman
and Rycroft, 2004). For sharp-crested weirs, the higher values of should be used.
OUTLET STRUCTURES
Sluices and flap gates
The discharge rate through a sluice or flap gate can be calculated with Equation 6,
being in this case b the width of the sluice and a coefficient from 0.9 to 1.1. The water
depth h2 should be increased by 3.5 percent if the sluice discharges directly into the
sea, because of the heavier saltwater outside (Smedema, Vlotman and Rycroft, 2004).
The outside water heights vary with tides or floods, so that at high levels discharge
is not possible and water must be stored inside. Therefore, the calculations must be
numerical, in time steps, for water level and storage conditions that are typical for the
location involved.
Pumping stations
The capacity of a pumping station is determined by the total discharge from all sources:
rainfall, irrigation excess, seepage, municipal and industrial wastewaters, etc. However,
it is not simple to estimate the simultaneous occurrence of all these events. In contrast
to open watercourses, pumps have a rather inflexible capacity, so that some reserve is
usually added.
A pumping station often has to run at full capacity for short periods only. Most of the
time it has to remove the base flow from more permanent sources, of which seepage and
tail-end losses from irrigation systems are the main ones. More than the strongly variable
inputs from rainfall, these flows determine the number of pumping hours per year and,
consequently, the costs of operation.
In order to cope with the variable capacity needed in different periods, more than one
pump is usually installed, of which one to remove the base flow and one or more to cope
with larger discharges and the design discharge at critical periods.
In order to select the most appropriate capacity arrangement and type of pump, some
design parameters should be calculated, namely: the base, usual and maximum discharge,
the lift and the dynamic head, and the power requirement.
The lift equals the static difference between inside and outside water. The dynamic
head may be calculated using:

Guidelines and computer programs for the planning and design of land drainage systems

154

h = hs + h +

vd 2
2g

(9)

where:
g 9.8 m/s2;
h = total head (m);
hs = lift or static head (m);
vd = flow velocity at the outlet of the delivery pipe (m/s);
h = total head loss in the suction and delivery pipes (m).
Consideration should be given to the head-increasing effect of choking of trashracks
that usually protect the inlet section of drainage pumping stations from the entrance of
floating debris such as mown aquatic weed, plastic, and branches, if timely cleaning of these
racks is not secured.
The power requirement may be calculated using:
P=

gQh
t p

(10)

where:
h = total head (m);
P = power required (kW);
Q = discharge rate (m3/s);
and p are the transmission (0.900.95) and pump efficiencies, respectively;
p = density of water 1 000 kg/m3.
The p values can vary for axial pumps from 0.65 for 1-m lift to 0.80 for 2.53.0m lift; for radial pumps from 0.6 for 1-m lift to 0.800.85 for lifts of more than 4.0 m
(Smedema, Vlotman and Rycroft, 2004); and Archimedes screws may have an efficiency
of 6575 percent (Wijdieks and Bos, 1994).
Some correction factors may be also considered in Equation 10 in order to take account
of the elevation of the site and safe load (Smedema, Vlotman and Rycroft, 2004).
REFERENCES
ILRI. 1964. Code of practice for the design of open watercourses and ancillary structures.
Bulletin 7. Wageningen, The Netherlands. 82 pp.
Smedema, L.K., Vlotman, W.F. & Rycroft, D.W. 2004. Modern land drainage. Planning, design
and management of agricultural drainage systems. Leiden, The Netherlands, A.A. Balkema
Publishers, Taylor&Francis. 446 pp.
Wijdieks, J. & Bos, M.G. 1994. Pumps and pumping stations. In H.P. Ritzema, ed. Drainage
principles and applications, pp. 965998. 2nd edition. ILRI Publication 16. Wageningen, The
Netherlands, ILRI.

155

Annex 11

Example of the batch method

for flat lands
The batch method for flat lands is described by means of an example for water
distribution from an extreme rainfall, with data from the Ebro Delta in northeast Spain,
where the climate is Mediterranean and extreme rainfalls are common in autumn.
Although the rainfall period extends for several consecutive days, an exceptional rain
of about 100 mm may fall in one day for a return period of 5 years. The following
days are rainy but the amount of precipitation decreases progressively. These autumn
rainfalls may affect irrigated rice fields during harvesting operations. On the left bank
of the Ebro Delta, flat areas of 2 2003 000 ha are served by drainage pumping stations
managed by the local water users association. The farm in this example is served by
a station with four Archimedes screws, each able to remove 9.5 mm/d, so that the
maximum total capacity of the pumping station is 38 mm/d. During the irrigation
period, only one of the pumps usually discharges about 5 mm/d, mainly surface
drainage water from the rice fields. Table A11.1 shows the results of calculations based
on the above data.
Although the rice fields are drained before harvesting by the existing surface
drainage systems, the soil is almost saturated and storage can be considered negligible.
However, about 25 mm can be stored in the channel system. On rainy days in autumn,
evaporation can remove about 3 mm/d from the area.
It is assumed that, on the first day, the full pumping capacity of the station has to
be started, evaporation is negligible and, therefore, only about 25 mm can be removed.
The excess 75 mm cannot be stored in the soil and in the channels, so inundation occurs
in the rice fields. In the following days, the four available screws work day and night.
Subsequently, the inundation storage and the water in the channels are drained. These
conditions are suitable for the rice field requirements.
However, in some areas of the Ebro Delta, vegetables are grown in fields with
surface and subsurface drainage facilities. Heavy autumn rainfalls may affect crops such
as tomato and lettuce severely. Table A11.2 shows the water distribution of extreme
rainfalls for a 10-year return period with the existing shared pumping facilities. It is
assumed that in these irrigated lands where the groundwater table is controlled by a
subsurface drainage system, the soil becomes completely saturated after storing about
50 mm.
Even with all four pumps working fully, inundation cannot be avoided on two days.
In addition to this, pumping should continue to lower the water level in the channels
in order to allow the subsurface drainage system to drawdown the water level, at least
TABLE A11.1

Water balance of a rice field in a flat area

Day

Rainfall

Evaporation

Pumped
water

Excess
rainfall

Storage in:
Soil

Channels

Inundation

Total

(mm/d)
1

100

25

75

25

50

75

21

38

55

25

30

55

38

18

18

18

15

Guidelines and computer programs for the planning and design of land drainage systems

156

TABLE A11.2

Water balance of a vegetable field in a flat area

Day

Rainfall

Evaporation

Pumped
water

Excess
rainfall

Storage in:
Soil

Channels

Inundation

Total

(mm/d)
1

125

25

100

50

25

25

100

29

38

88

50

25

13

88

38

51

50

51

30

18

18

18

15

Excess
rainfall

Soil

TABLE A11.3

Example of water balance for a 6-hour period

Hour

Rainfall

Evaporation

Pumped
water

Storage in:
Channels

Inundation

Total

(mm/h)
1

53

52.0

20

15

17.0

52.0

27

1.6

77.4

35

20

22.4

77.4

14

1.6

89.8

45

25

19.8

89.8

1.6

94.2

50

25

19.2

94.2

1.6

95.6

50

25

20.6

95.6

1.6

95.0

50

25

20.0

95.0

25 cm in one day. Inundation for two days could be tolerated by tomato and lettuce
in the Ebro Delta, providing that they are grown on beds between surface drainage
furrows. However, as the pumping requirements are higher than for standard rice field
needs, individual pumping stations may be needed in farms with surface and subsurface
drainage systems where vegetables are grown jointly with rice (as the actual shared
pumping facilities were designed mainly for covering rice field requirements).
The pumping capacity should also be increased if the critical period is less than
24 hours as it is frequently needed to cultivate more sensitive crops. If heavy rain falls
in the first three hours, soil storage may be limited by soil infiltration, which is usually
highest at the beginning. However, it soon decreases, becoming later almost constant
until the soil is saturated completely. In the example of Table A11.3, water distribution
is shown with pumping capacity and channel storage similar to the previous example.
In this example, inundation reaches its maximum value after about 2 hours. After
this time, it decreases slightly, but stagnation occurs in the following hours. If the
critical period is about 6 hours and the excess rainfall should be removed during this
time interval, the pumping capacity should be increased substantially or less sensitive
crops should be cultivated. Consequently, in certain areas of the Ebro Delta, where
horticultural crops are grown, in addition to the pumping stations for subsurface
drainage water, independent pumping stations with a higher capacity discharge surface
drainage water during the critical periods of heavy rainfall.

157

Annex 12

Cypress Creek formula

PRINCIPLES
The Soil Conservation Service
(now called the Natural Resource
Conservation Service) of the United
States Department of Agriculture
developed a simple formula called
the Cypress Creek equation (NRCS,
1998):
Q = qA5/6

(1)

TABLE A12.1

Typical drainage coefficients for humid areas

Drainage coefficient
(m3s-1km-2)
Coastal plain cultivated

0.59

Delta cultivated lands

0.52

Cool northern cultivated

0.48

Coastal plain pasture

0.39

Cool northern pasture

0.33

Delta and coastal rice lands

0.30

Semi-humid northern cultivated

0.26

Semi-humid southern range lands

0.20
where:
Coastal
plain
woodlands
0.13
Q = design discharge (m3/s)
Source: Adapted from ASAE-EP 407.1, 1994.
not peak discharge as
some flooding can take
place;
q = 0.21 + 0.00744P24 = drainage coefficient related to the drainage area and the
magnitude of the storm (cubic metres per second per square kilometre)
(Ochs and Bishay, 1992);
P24 = 24-hour excess rainfall (mm) the excess rainfall can be calculated with the
CN graph, but considering that the CN method was developed for free
drainage conditions; for storm periods longer than a day, the total rainfall
excess is divided by the length of the storm period in days (Ochs and Bishay,
1992);
A = area served by the drain (square kilometres).
The equation was developed for the eastern portion of the United States of America.
It is basically applicable for humid flat lands covering less than 5 000 ha, with conditions
similar to the areas for which was developed.
Table A12.1 shows drainage coefficients for the east of the United States of
America.

REFERENCES
ASAE-EP 407.1. 1994. Agricultural drainage outlets - open channels. In: American Society of
Agricultural Engineers book of standards, pp. 728733. St. Joseph, USA.
Natural Resource Conservation Service (NRCS). 1998. Water management (drainage).
Chapter 14 of Part 650 Engineering Field Handbook. Washington, DC. 160 pp.
Ochs, W.J. & Bishay, B.G. 1992. Drainage guidelines. World Bank Technical Paper No. 195.
Washington, DC. 186 pp.

159

Annex 13

Statistical analysis of measured

flows

PRINCIPLES
The maximum discharge at the outlet of the main drainage system can be determined
statistically where a data series of measured flows is available covering a period of at
least 1520 years. For example, the occurrence probability can be calculated with the
following formula:
(1)
where:
P = probability;
T = 1/P = return period (years);
m = order number in the data series;
N = number of total data available.
Example
Equation 1 has been applied in the example shown in Table A13.1.
With the data of Table A13.1, the maximum discharge for a return period of up to
20 years can be determined (98.3 m3/s in this case), which is sufficient to design the
main drainage system. Where a higher return period is required in order to design
special structures, the design discharge can be estimated by extrapolation, once the
TABLE A13.1

Frequency analysis of drainage flows (for N = 19)

Year

QM

(m3/s)

QM

(m3/s)

P=

m
N +1

T = 1/P years

1967

85.1

98.3

0.05

20

1968

50.1

17

90.2

0.10

10

1969

48.2

18

85.3

0.15

1970

68.3

10

85.1

0.20

1971

60.4

13

80.7

0.25

1972

55.2

14

80.6

0.30

1973

80.7

78.4

0.35

1974

90.2

78.3

0.40

1975

85.3

76.7

0.45

1976

61.3

12

68.3

10

0.50

1977

98.3

61.5

11

0.55

1978

78.4

61.3

12

0.60

1979

80.6

60.4

13

0.65

1980

36.7

19

55.2

14

0.70

1981

50.2

15

50.2

15

0.75

1982

61.5

11

50.2

16

0.80

1983

50.2

16

50.1

17

0.85

1984

78.3

48.2

18

0.90

1985

76.7

36.7

19

0.95

Source: Adapted from Smedema, Vlotman and Rycroft, 2004.

160

Guidelines and computer programs for the planning and design of land drainage systems

available data are plotted on a probability paper, for example by using the normal
distribution. However, this type of calculation is based on historical data, and runoff
may change with changes in land use.
REFERENCES
Smedema, L.K., Vlotman, W.F. & Rycroft, D.W. 2004. Modern land drainage. Planning, design
and management of agricultural drainage systems. Leiden, The Netherlands, A.A. Balkema
Publishers, Taylor&Francis. 446 pp.

161

Annex 14

Unit hydrograph

PRINCIPLES
This method, developed by Sherman (1932), is based on the proportionality principle:
the surface runoff hydrograph produced by certain amount of rainfall (P) can be
obtained from the hydrograph of other storm of equal duration (P) by multiplying
the ordinates of the latter hydrograph by the following conversion factor:
(1)

Example
In Figure A14.2, the hydrograph
for the surface runoff produced by
a rainfall of 40 mm accumulated
in 6 hours, of which 25 mm was
accumulated in the first 3 hours, has

q (mm/h)

ym

ay

ay m

td

t (h)

t
td Lag time
t

Time base

FIGURE A14.2

Example of the unit hydrograph method

P (mm)
25
15
q (mm/h)

where:
a = conversion factor;
Sr = amount of surface runoff
produced by precipitation
P (mm);
Sr = amount of surface runoff
produced by precipitation
P (mm).
This method is also based on the
concept that the base length (t) of a
hydrograph depends on the duration
of the storm, but is independent of
the amount of rainfall and surface
runoff, as shown in Figure A14.1.
The recession time (t - td) is
almost constant. This is because
it only depends on the physical
characteristics of the basin.
For practical applications, it is
advisable to convert the available
hydrographs to unit hydrographs,
namely, hydrographs for precipitations of 1 or 10 mm. Thus, for
the project basin, a series of unit
hydrographs can be obtained for
different rainfall durations. In order
to determine the hydrograph for the
design rainfall, the unit hydrograph
with a time basis similar to the
design rainfall is selected.

FIGURE A14.1

Proportionality principle of the unit hydrograph

10
0

12

t (h)

10 mm unit hydrograph for a 3 hour storm

Hydrograph for a design storm of 25 mm in 3 h
Hydrograph for a design storm of 15 mm recorded from 3 to 6 h
Final hydrograph for a design storm of 40 mm in 6 h

162

Guidelines and computer programs for the planning and design of land drainage systems

been determined from the unit hydrograph available for a rainfall of 10 mm in 3 hours.
It is assumed that all rain becomes surface runoff.
The hydrograph for the first 3 hours is obtained from the 10-mm unit hydrograph
by applying a conversion factor (a = 2.5). For the following 3-hour period, a conversion
factor (a = 1.5) is used. The final hydrograph is obtained by superimposing both
hydrographs. It can be observed that the peak discharge will be produced 5 hours after
the beginning of the storm.
REFERENCES
Sherman, L.K. 1932. Streamflow from rainfall by the unit-graph method. Eng. News Rec., 108:
501505.

163

Annex 15

Rational formula
PRINCIPLES
The rational method assumes that, in small agricultural basins, the maximum flow of
surface water in the outlet is for a rainfall with a duration equal to the concentration
time. Then, the maximum discharge depends on the rainfall intensity, the surface area
and the hydrological conditions of the basin:
(1)
where:
QM = maximum discharge for a return period equivalent to the design rainfall
(m3/s);
C = coefficient for surface runoff;
I = rainfall intensity during the concentration time (mm/h);
A = area of the basin (ha).
For the return period selected, rainfall intensity is assumed: (i) constant during the
time interval considered; and (ii) equal to the ratio between the accumulated rainfall
and the concentration time. Where only the amount of rainfall in 24 hours is known,
the value of the precipitation accumulated in the concentration factor can be estimated,
first by using an appropriate coefficient for the 6-hour rainfall (P6/P24 = 0.50.7), and
then with the coefficients of the rainfall distribution model described in Chapter 6 of
the main text.
SURFACE RUNOFF COEFFICIENT
The runoff coefficient can be estimated directly through the indicative values of the Soil
Conservation Service (SCS, 1972) shown in Table A15.1.

250 m

Example
The rational method has been applied to estimate the maximum discharge of surface
water at the outlet (point D) of a farm of 85 ha shown in Figure A15.1.
In order to estimate the concentration time at point D, three sections
FIGURE A15.1
have been considered from the most
Example of drained farm with a system of furrows and open
distant point from the outlet (point
ditches
A): section AB (furrows), section BC
1 700 m
(open collector drain), and section
B
C
CD (the main drain).
Assuming values of the water
velocity of 0.15 and 0.35 m/s along
the furrows and the open ditches,
A
respectively, Table A15.2 shows the
concentration time tc for each section
as calculated using:
D

(2)
where:

Furrows
Collector

Guidelines and computer programs for the planning and design of land drainage systems

164

TABLE A15.1

Indicative values of the surface runoff coefficient for agricultural land

Land use
Arable land

Pasture

Forest

Slope
(%)

High

Soil infiltrability
Medium

<5

0.30

0.50

0.60

510

0.40

0.60

0.70

1030

0.50

0.70

0.80

<5

0.10

0.30

0.40

510

0.15

0.35

0.55

1030

0.20

0.40

0.60

<5

0.10

0.30

0.40

510

0.25

0.35

0.50

1030

0.30

0.50

0.60

Low

Source: Adapted from Smedema, Vlotman and Rycroft, 2004.

TABLE A15.2

Estimates of the concentration time

Section

Length

Slope

Difference of
elevation

Water velocity

tc

(m)

(%)

(m)

(m/s)

(h)

AB

250

0.10

0.25

0.15

0.46

BC

1 700

0.15

2.55

0.35

1.35

CD

500

0.35

0.40

AD

2 450

2.80

2.21

tc = concentration time (s);

li = distance of section i (m);
vi = average water velocity in section i (m/s).
The concentration time can also be estimated using the Kirpich formula:
(3)
In this case:
l = distance AD = 2 450 m;
h = difference of elevation between A and D = 2.8 m;
s = h/l = average slope between A and D = 0.00114;
= constant = 72 471.98 (m);
tc = concentration time = 1.79 h.
The values obtained for tc are around an average value of 2 h, which can be used for
further calculations. If during this time the accumulated rainfall for a return period of
5 years is 64 mm, the rainfall intensity is about 32 mm/h.
The runoff coefficient according to Table A15.1 is about 0.3. Then, the maximum
flow at point D is about 2.3 m3/s, as calculated with Equation 1.
REFERENCES
Smedema, L.K., Vlotman, W.F. & Rycroft, D.W. 2004. Modern land drainage. Planning, design
and management of agricultural drainage systems. Leiden, The Netherlands, A.A. Balkema
Publishers, Taylor&Francis. 446 pp.
Soil Conservation Service (SCS). 1972. Hydrology. National Engineering Handbook Section 4.
Washington, DC, USDA.

165

Annex 16

Curve Number method

PRINCIPLES
The Curve Number (CN) method
is based on the conceptual
interpretation of the hydrological
process during a rainfall period.
Initially, no surface runoff (Sr) is
produced while rainfall is intercepted
by vegetation and water infiltrates
into the soil (Ia). When rainfall
exceeds this initial interception,
overland flow begins while soil
infiltration continues (Inf). Once
the soil is saturated, any amount of
excess rainfall (P) produces surface
runoff (Figure A16.1).
Figure A16.2 shows the relationship between the precipitation
accumulated and surface runoff
during a rainfall period.
The amount of Sr is zero if the
accumulated rainfall is lower than
the Ia value. Once this threshold
value has been exceeded, the Sr
function takes a curve shape up
to the saturation point where Sr is
equal to P. From this point, the Sr
function becomes a straight line
with unit slope (a = 45 ). If this
line is extended to cut the x-axis, a
point is achieved that represents the
maximum retention potential (S).
The S value depends on the physical
characteristics of the basin and on
the soil moisture content before the
rainfall period.
Once overland flow starts, the
water balance on the soil surface is:
(1)
where:
Inf = actual infiltration while
surface runoff is produced
(mm);

FIGURE A16.1

Surface runoff during a rainfall period

Sr

I nf
t

Ia
Source: Adapted from Boonstra, 1994.

FIGURE A16.2

Relationship between precipitation and surface runoff

Sr(mm)

Sr = P

Sr = f(P)

la

P(mm)
S
S

Guidelines and computer programs for the planning and design of land drainage systems

166

Ia

= amount of water intercepted and infiltrated into the soil before overland
flow occurs (mm);
P
= amount of accumulated rainfall (mm);
P - Ia = maximum potential of surface runoff (mm);
= accumulated surface runoff (mm).
Sr
This method, developed by the Soil Conservation Service (SCS), assumes that the
relationship between the actual surface runoff and its maximum potential value is equal
to the rate between the actual infiltration and the maximum potential retention. The
latter is approximately equal to the accumulated infiltration after runoff has started
(Figure A16.2):
Sr
P Ia

(P I ) S
a

(2)

where:
S = maximum potential retention (mm).
Surface runoff can be then expressed as:

(P I )
(P I ) + S
2

Sr =

(3)

Equation 3 has been simplified by assuming that the value of the potential retention is
constant during a storm and the initial interception is about 20 percent of the maximum
potential retention (Ia = 0.2S). Thus, surface runoff depends only on precipitation and
the maximum potential retention:
Sr =

(P 0.2 S)

for

P + 0 .8 S

(4)

The SCS formulated a new undimensional parameter, named the Curve Number
(CN), to assess the capacity of a basin to produce surface runoff after certain
precipitation. This parameter is a hydrological characteristic of the basin, which
depends on the maximum potential retention:
CN =

25400

(5)

254 + S
FIGURE A16.3

Function surface runoff precipitation for different Curve

Numbers
Sr (mm)
200
175
00
=1
CN 95
90 5
8

150

80
75

125

70

100

65

60
55
50

75

45
40

50

By combining Equations 4 and

5, one expression can be obtained
to calculate the accumulated surface
runoff from the amount of rainfall
and the CN. Figure A16.3 shows the
function Sr/P in the graph developed
by the SCS (1972) for different CN
values.
Thus, in a basin characterized
by a certain CN, the amount of
surface runoff produced by a design
rainfall can be estimated by means of
Figure A16.3 or through Equations 4
and 5.

35

30
25
20

25
0
25

50

75

100

125

150

175

200

225

250

275

300

P (mm)
Source: Adapted from Boonstra, 1994.

ESTIMATION OF THE CURVE

NUMBER
The CN value depends on:
the natural vegetation and the
current land use;

Annex 16 Curve Number method

167

the hydrological soil characteristics, especially the infiltration;

the agricultural practices;
the previous soil moisture content.

This method does not consider land slope because lands with gradients of more
than 5 percent are not cultivated in the United States of America. However, classes for
different slopes can be considered in a specific project (Boonstra, 1994).
The CN value increases progressively as retention decreases, the maximum value
being 100 where retention in negligible. Table A16.1 shows the CN values established
by the SCS (1972) for average soil moisture conditions before the design storm,
considered as Class II.
In Table A16.1, the term straight rows means rows along the land slope. The
hydrological condition essentially depends on the vegetation density. Condition is
poor where meadows are intensively used or the grass quality is low, or where field
crops are in the initial stage of growing. Otherwise, condition is good for densely
vegetated meadows and for field crops covering the soil surface well.
In addition to the average soil moisture conditions considered in Table A16.1 for
Class II, the SCS defined two additional classes (I and III), taking into account the
amount of precipitation in the five-day period before the design storm (Table A16.2).
If the antecedent soil moisture condition differs from Class II, the equivalent
CN values for Class I or Class III can be estimated by using the conversion factors
developed by the SCS (1972) and shown in Table A16.3, once the CN value has been
determined for Class II.
TABLE A16.1

CN values Class II
Land use

Practice

Hydrological
condition

Soil infiltrability
Low

Very low

Fallow

Straight row

Poor

77

86

91

94

Row crops

Straight row

Poor

72

81

88

91

Good

67

78

85

89

Contoured

Poor

70

79

81

88

Good

65

75

82

86

Contoured/terraced

Poor

66

74

80

82

Good

62

71

78

81

Poor

65

76

84

88

Good

63

75

83

87

Contoured

Poor

63

74

82

85

Good

61

73

81

84

Contoured/terraced

Poor

61

72

79

82

Good

59

70

78

81

Poor

66

77

85

89

Good

58

72

81

85

Contoured

Poor

64

75

83

85

Good

55

69

78

83

Contoured/terraced

Poor

63

73

80

83

Good

51

67

76

80

Poor

68

79

86

89

Fair

49

69

79

84

Small grain

Close-seeded legumes
or rotational meadow

Straight row

Straight row

Pasture range

High

Medium

Good

39

61

74

80

Poor

47

67

81

88

Fair

25

59

75

83

Good

35

70

79

Meadow (permanent)

Good

30

58

71

78

Woodland

Poor

45

66

77

83

Fair

36

60

73

79

Good

25

55

70

77

Contoured

Source: Adapted from Boonstra, 1994.

Guidelines and computer programs for the planning and design of land drainage systems

168

TABLE A16.2

Classes for previous soil moisture conditions

Class

P in the previous 5-day period

Dormant season

Growing season
(mm)

< 13

< 36

II

1328

3653

III

> 28

> 53

Source: Adapted from Boonstra, 1994.

TABLE A16.3

Equivalent CN according to the antecedent soil moisture classes

Class

CN

100

78

63

51

40

31

22

15

II

100

90

80

70

60

50

40

30

20

10

III

100

96

91

85

78

70

60

50

37

22

Source: Adapted from Boonstra, 1994.

In order to estimate the average CN value of a basin, all the sections with different
hydrological conditions, land use and agricultural practices should first be mapped.
Then, the respective CN is assigned to each independent section. Last, the weighted
average is calculated according to the surface area of each section.
Example
In this example, the CN method has been applied to estimate the amount of surface
runoff produced by an extreme rainfall of 125 mm in 24 hours, determined for a
return period of 10 years, in a basin of 4 740 ha, where the current land use is rainfed
agriculture and forest. This was the previous stage to calculate later the maximum water
flow at the outlet of the main watercourse draining the basin.
The first step for this calculation was to estimate the concentration time of the basin
with the Kirpich formula (although this formula was developed for small agricultural
basins). For a watercourse with a length of 15.5 km and a difference in elevation
between the most distant point from the outlet and the outlet itself of 299.4 m, the tc
value is 2.5 hours.
The second step was to assess the rainfall distribution during the first 6 hours of
the storm. This period of 6 hours was selected, because the concentration time is less
than 6 hours. It was assumed that during the first 6 hours, 60 percent of the one-day
precipitation occurred, i.e. 75 mm. The rainfall distribution during this period can be
estimated by the WMO model for time intervals of 0.5 hours, as shown in Table A16.4.
In order to estimate the weighted average CN for the whole basin, the area was
split into six sections with homogeneous land use and hydrological conditions by
superimposing the land-use map and the soil map. The physical characteristics of these
sections are described in Table A16.5, where the individual CN, estimated for Class II,
were assigned to each section.
The weighted average CN for the basin as a whole is 69 for Class II (Table A16.5).
However, the previous soil moisture conditions are more similar to those of Class III
as in the area studied extreme rainfalls are frequent in autumn. Therefore, it is more
adequate to use the equivalent CN for Class III, i.e. 85 according to Table A16.3.
TABLE A16.4

Distribution of the total precipitation in a period of 6 hours

Time (h)
Rainfall distribution (%)
Accumulated rainfall (mm)

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

6.0

15

22

60

70

78

84

88

92

96

100

1.5

6.0

11.3

16.5

45.0

52.5

58.5

63.0

66.0

69.0

72.0

75.0

Annex 16 Curve Number method

169

TABLE 16.5

Physical characteristics and CN values of the hydrologically homogeneous sections

Section

Surface area

Soil type

Land use

Shallow soils on shale rock

Agricultural practice

Infiltrability

CN

Pasture

Low

79

Woodland & pasture

Medium

69

Medium

71

High

59

(ha)
1

762

1 566

1 161

Terraced deep soils

Vineyard

990

Terraced deep soils

Field crops

30

231

Basin

Straight rows

Terraced soils

Dense field crops

Low

76

Moderately shallow soils

with slopes > 2%

Pasture

Low

74

4 740

69

TABLE A16.6

Estimation of the amount of surface runoff for CN = 85

Time (h)

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

6.0

Accumulated rainfall (mm)

1.5

6.0

11.3

16.5

45.0

52.5

58.5

63.0

66.0

69.0

72.0

75.0
39.3

Accumulated runoff (mm)

0.1

1.1

16.1

21.5

26.0

29.5

31.9

34.4

36.8

Sr (mm)

1.0

15.0

5.4

4.5

3.5

2.4

2.5

2.4

2.5

The maximum potential retention

for this CN is 44.8 mm (Equation 5).
With this value, the surface runoff
produced for the design rainfall can
be calculated with Equation 4 or
estimated by means of Figure A16.3.
Table A16.6 shows the results.
HYDROGRAPH OF THE SPECIFIC
DISCHARGE
The dimensionless unit hydrograph
developed by the SCS can be used
to calculate the maximum specific
discharge of surface runoff and
the maximum water flow. In this
hydrograph, time is expressed as a
function of the elevation time, and
discharge is related to its maximum
value. Figure A16.4 shows this
hydrograph and a table with average
values.
From numeric integration of
this hydrograph, the following
expression can be obtained for the
maximum specific discharge:

FIGURE A16.4

SCS dimensionless unit hydrograph

q/q M

t/t e

1.0

0.8

0.6

0.4

0.2

t/t e

q/q M

0
0.25

0
0.12

0.50

0.43

0.75

0.83

1.00

1.00

1.25

0.88

1.50

0.66

1.75

0.45

2.00

0.32

2.25

0.22

2.50

0.15

2.75

0.105

3.00

0.075

3.25

0.053

3.50

0.036

3.75

0.026

4.00

0.018

4.25

0.012

4.50

0.009

4.75

0.006

5.00

0.004

Source: Adapted from Boonstra, 1994.

(6)
where:
qM = maximum specific discharge (litres per second per hectare);
Sr = amount of surface runoff (mm);
te = elevation time (h).

Guidelines and computer programs for the planning and design of land drainage systems

170

FIGURE A16.5

Example of hydrograph for the total specific discharge

q (l s-1 ha-1)
30
25
20
15
10
5
0

The te value can be estimated from

the concentration time (te 0.7tc).

Example
The elevation time (te) in the basin
of the previous example is about
1.75 hours. With this value, in
Table A16.7 the maximum specific
discharge (qM) for each increment
of surface runoff (Sr) has been
calculated with Equation 6. In
Table A16.7, the distribution of
the specific discharge has also been
determined by applying the tabulated
t (h)
values of the undimensional hydro2
3
4
5
6
7
8
9
10 11 12
graph represented in Figure A16.4 to
the qM values.
The hydrograph for the total specific discharge (Figure A16.5) was obtained by
superimposing the partial hydrographs obtained with the results of Table A16.7.

TABLE A16.7

Calculation of the partial specific discharges qM and the total discharge qt

t (h)

0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
9.0
9.5
10.0
10.5
11.0
11.5
12.0
12.5
13.0
13.5
14.0

Undimensional
hydrograph
t/te
q/qM
0.00
0.00
0.29
0.17
0.57
0.54
0.86
0.91
1.14
0.93
1.43
0.72
1.71
0.48
2.00
0.32
2.29
0.21
2.57
0.14
2.86
0.09
3.14
0.06
3.43
0.04
3.71
0.03
4.00
0.02
4.29
0.01
4.57
.008
4.86
.005
5.14
.003
5.43
5.71
6.00
6.29
6.57
6.86
7.14
7.43
7.71
8.00

1.5

2.0

2.5

1.0

15.0

5.4

1.2

17.9

6.4

0.20
0.65
1.09
1.12
0.86
0.58
0.38
0.25
0.17
0.11
0.07
0.05
0.04
0.02
0.01
0.01
0.01

3.04
9.67
16.29
16.65
12.89
8.59
5.73
3.76
2.51
1.61
1.07
0.72
0.54
0.36
0.18
0.14
0.09
0.05

1.09
3.46
5.82
5.95
4.61
3.07
2.05
1.34
0.90
0.58
0.38
0.26
0.19
0.13
0.06
0.05
0.03
0.02

t (h)
3.0
3.5
4.0
Sr (mm) (see Table A16.6)
4.5
3.5
2.4
qM = 2.08 Sr/te = 1.19Sr
5.4

0.92
2.92
4.91
5.02
3.89
2.59
1.73
1.13
0.76
0.49
0.32
0.22
0.16
0.11
0.05
0.04
0.03
0.02

4.2

0.71
2.27
3.82
3.91
3.02
2.02
1.34
0.88
0.59
0.38
0.25
0.17
0.13
0.08
0.04
0.03
0.02
0.01

2.9

0.49
1.57
2.64
2.70
2.09
1.39
0.93
0.61
0.41
0.26
0.17
0.12
0.09
0.06
0.03
0.02
0.01
0.01

qt (l s-1 ha-1)
4.5

5.0

5.5

2.5

2.4

2.5

3.0

2.9

3.0

0.51
1.62
2.73
2.79
2.16
1.44
0.96
0.63
0.42
0.27
0.18
0.12
0.09
0.06
0.03
0.02
0.02
0.01

0.49
1.57
2.64
2.70
2.09
1.39
0.93
0.61
0.41
0.26
0.17
0.12
0.09
0.06
0.03
0.02
0.01
0.01

0.51
1.62
2.73
2.79
2.16
1.44
0.96
0.63
0.42
0.27
0.18
0.12
0.09
0.06
0.03
0.02
0.02

0.20
3.69
11.85
21.79
26.96
27.09
24.50
21.60
19.10
16.85
14.03
10.59
7.34
4.93
3.28
2.13
1.43
0.92
0.61
0.38
0.24
0.13
0.08
0.04
0.03

Annex 16 Curve Number method

Figure A16.5 shows that about 4 hours after of the beginning of the design storm
the maximum specific discharge is expected, its value then being about 27 l s-1 ha-1.
With this surface drainage coefficient, each section of the main drainage system can be
dimensioned. At the outlet of this basin of 4 740 ha, the maximum estimated flow will
be about 128 m3/s.
REFERENCES
Boonstra, J. 1994. Estimating peak runoff rates. In H.P. Ritzema, ed. Drainage principles and
applications, pp. 111143. 2nd edition. ILRI Publication 16. Wageningen, The Netherlands,
ILRI.
Soil Conservation Service (SCS). 1972. Hydrology. National Engineering Handbook Section 4.
Washington, DC, USDA.

171

173

Annex 17

Formulae for steady-state flow to

drains
This annex gives formulae for the
calculation of open or covered
parallel drain spacings for use for
different soil profiles.
FLOW ABOVE DRAIN LEVEL; THE
ELLIPSE EQUATION
The ellipse equation (Figure A17.1)
is valid for a single layer above drain
level (Van der Ploeg, Marquardt and
Kirkham, 1997).
Where an impermeable layer is
present at drain level, the phreatic
groundwater table between two
drains has an elliptic shape. The
resulting formula for the drain
spacing then equals:
(1)
where:
h = groundwater elevation midway drains (m);
K = permeability above drain
level (m/d);
L = drain spacing (m);
q = design discharge (m/d).
The ellipse formula is used in the
programs for the flow above drain
level, either as the only discharge or
in combination with flow through
deeper layers.

FIGURE A17.1

Steady-state flow above drain level the ellipse equation

Recharge q

x=0

Impermeable base

x=L

FIGURE A17.2

Steady-state flow above and below drain level the

Hooghoudt equation

Discharge q1
above drains

Recharge q
K1
Drain level

Discharge q2
below drains

D1

K2 D

Equivalent layer d

K2
x=0

Impermeable base

x=L

FLOW ABOVE AND BELOW DRAIN LEVEL; THE HOOGHOUDT EQUATION

The Hooghoudt approach (Hooghoudt, 1940) considers a soil that is either
homogeneous above and below the drain level or consists of two layers with different
properties above and below drain level (Figure A17.2). Hooghoudts formula for
calculating drain spacings under steady-state flow assumptions is:
(2)
where:
d = f(D2,L,r) = effective thickness of lower layer (m);
D1 = thickness of the layer above drain level (m) mentioned in Figure A17.2;

Guidelines and computer programs for the planning and design of land drainage systems

174

D2 = real thickness of the layer below drain level, down to the impermeable subsoil
(m);
K1 = permeability above drain level (m/d);
K2 = permeability below drain level (m/d);
r = effective drain radius (m).
Inputs for Equation 2 are D2, h, K1, K2, q and r, of which D2 may be infinite. Because
d depends on the required distance L, iteration is necessary.
Hooghoudts method for calculating drain spacings is valid for a two-layered soil
profile: one layer above and one below drain level. The latter not only offers resistance
to horizontal flow, but also radial resistance that occurs near the drain, where the
streamlines are converging.
In this approach, the flow pattern is replaced by horizontal flow through a thinner
layer; the actual thickness D2 of the layer below the drains is replaced by the equivalent
layer d without radial resistance (Figure A17.2). For steady-state flow, this is allowed,
but errors may occur in non-steady cases.
The equivalent layer d, which is a complicated function, is used as a substitute
correction for the radial resistance caused by the convergence of streamlines near the
drain. It is smaller than the real thickness D2 of the lower layer and was tabulated by
Hooghoudt. Subsequently, nomographs were based on these tables (Van Beers, 1979).
However, for computer applications a series solution is more effective. The following
series solution may be used to find d:
d=

L / 8
L
ln + G(x)
r

x=

2D2
L

(3)

(4)

which converges rapidly for x > 0.5.

For smaller values of x, Dagans formula results in the expression:
(5)
These formulae are well-suited for computer application.
ERNST EQUATION
The Ernst method (Ernst, 1956) for calculating drain spacings allows two-layered
profiles with a horizontal boundary at arbitrary level but not necessarily at drain
depth (Figure A17.3). If homogeneous, layers 1 and 2 are supposed to be of equal
composition (K2 = K1 and an2 = an1).
In this method, the flow is divided into three parts, each of which is calculated:
a vertical flow to the aquifer, with a vertical head loss hv;
a horizontal flow to the vicinity of the drain, with horizontal head loss hh;
a radial flow towards the drain, with radial head loss hr.
The total head loss in the soil h is:
h = h v + hh + h r
The theory gives rise to a quadratic equation in L.

(6)

Annex 17 Formulae for steady-state flow to drains

175

THE TOKSZKIRKHAM
Figure A17.3
ALGORITHM
Drainage of a multilayered anisotropic soil profile the
Toksz and Kirkham (1971a
TokszKirkham and Ernst methods
and 1971b) devised a general
(for Ernst, K1 = K2 and D4 = 0)
theory for determining drain
spacings in multilayered soils with
Recharge q
arbitrary horizontal boundaries
D1
(Figure A17.3). It consists of a set
of complicated hyperbolic functions Discharge q
h
K v1
1
that depend on the number and above drains
K h1
thickness of layers considered.
x=L
The method calculates the flow Discharge q x = 0
2
D2
K h2
through 13 different layers below below drains
drain level (Figure A17.3). It uses
K h3
D3
the following definitions:
The layer above drain level
K h4
D4
Impermeable base
has permeability K1. It is not
considered in the theory, but the
resulting flow can be calculated
by Equation 1.
The first layer below drain level has permeability K2 and thickness D2.
The second layer below drain level has permeability K3 and thickness D3.
The third layer below drain level has permeability K4 and thickness D4.
The drain spacing is L, the drain radius r, the recharge intensity q, and the head
midway h.
Distances a, b, c and s are defined as:
a = D2

c = D2 + D3 + D4

b = D2 + D3

s=

(7)

The following auxiliary quantities are calculated:

K4

m =

K3

m (b a)
cosh
s

m (b a)
m = tanh
s
cosh
m =

m a
s

m b
sinh
s

m =

K4
K3

coth

sinh
m =

m b
s

m a
s

m b
sinh
s

m =

K3

1
m a
K2
sinh
s

m = tanh

m =

K3
K2

m (c b)
s

coth

m a
s

(8.a)

(8.b)

(8.c)

Furthermore:
Sm = me
Tm =

m a
s

m
1
m (1 + m m ) (1 + m m ) + m m ( m m m )

Um = 1 Sm Tm m Sm Tm m ( m m + m m )

(9)

(10)

(11)

Guidelines and computer programs for the planning and design of land drainage systems

176

The head h midway between drains is determined from:

(12)

Combination with the ellipse equation for flow above drains requires an iterative
solution.
These formulae are suited for computer applications.
INFLUENCE OF ANISOTROPY
In many soils, permeability depends on the direction of flow. Considerations here are
confined to horizontal layering and vertical cracks. The former results in a permeability
that is larger in the horizontal than in the vertical direction, the latter in the reverse.
In such cases, where the axes of the anisotropy coincide with the horizontal and
vertical x and z axes, the following rules may be used (Boumans, 1963):
An anisotropy factor ani is defined for each layer i as:
ani =

Khi

(13)

Kvi

with Kh horizontal and Kv vertical permeability of layer i.

Hydraulic heads and discharges remain the same.
Horizontal distances remain the same.
Vertical distances zi in layer i (especially thickness Di) are transformed to:

i = zi ani

(14)

The permeability is transformed to:

i =

Khi

(15)

ani

In this transformed isotropic system (Figure A17.4), all formulae for steady-state
flow are valid. The resulting spacing L is horizontal and, consequently, it remains
unchanged.
For flow above drains, a different
approach
is used. Here, the vertical
FIGURE A17.4
permeability
Kv1 of the first layer is
Anisotropic and transformed isotropic system
used to find the head loss between
ANISOTROPIC CASE
TRANSFORMED ISOTROPIC CASE
maximum head h and drain level
Recharge q
Recharge q
and, consequently, the corrected
head hc (the head at drain level) as:
K h1 K v1
x=0

K h2

an2

K h3

an3

K h4

an4

Impermeable base

x=L

D2

(16)

1
x=0

D3
D4

Recharge q 2

2
3
4
Impermeable base

x=L

With this corrected head,

all subsequent calculations are
executed.
The program SPACING is based
on the above theory. However, the
Ernst equation is not included. In
cases where it is applicable, it gives
practically the same results as the more
general TokszKirkham algorithm.

Annex 17 Formulae for steady-state flow to drains

REFERENCES
Boumans, J.H. 1963. Een algemene nomografische oplossing voor het stationaire
ontwateringsvraagstuk met toepassingen voor isotroop en anisotroop doorlatende gronden.
Polytechn. Tijdschr., 18(14B): 545551.
Ernst, L.F. 1956. Calculation of the steady flow of groundwater in vertical cross-sections. Neth.
J. of Agr. Sci., 4: 126131.
Hooghoudt, S.B. 1940. Algemeene beschouwing van het probleem van de detailontwatering en
de infiltratie door middel van parallel loopende drains, greppels, slooten en kanalen. Versl.
Landbouwk. Onderz., 46(14).
Toksz, S. & Kirkham, D. 1971a. Steady drainage of layered soils: I, Theory. J. Irr. Dr. Div.
ASCE, 97, IR1, 118. Erratum in J. Irr. Dr. Div. ASCE, 104, IR3, 333.
Toksz, S. & Kirkham, D. 1971b. Steady drainage of layered soils: II, Nomographs. J. Irr. Dr.
Div. ASCE, 97, IR1, 1937.
Van Beers, W.F.J. 1979. Some nomographs for the calculation of drain spacings. ILRI Bulletin 8.
Wageningen, The Netherlands.
Van der Ploeg, R.P., Marquardt, M. & Kirkham, D. 1997. On the history of the ellipse
equation for soil drainage. Soil Sci. Soc. Am. J., 61: 16041606.

177

179

Annex 18

Drainage under vertical seepage

INFLUENCE OF VERTICAL SEEPAGE
Artesian seepage (upward flow from deeper layers) is caused by groundwater flow
from higher areas. The sources may be nearby (e.g. irrigated lands on higher grounds)
or far away (through aquifers under pressure recharged in hills or mountains). Water
escaping from such aquifers causes upward flow to the rootzone. Drainage of such
seepage areas is often difficult. In many cases, temporary or even permanent wetness
and salinization occur.
Two main methods have been proposed for drain spacing design under these
conditions:
Vertical drainage is a good solution under special hydrological conditions.
Therefore, where there is no previous experience in the region, a careful
hydrogeological survey is needed.
Relief wells are another possibility where the aquifer is under pressure.
Where neither of these solutions is applicable, drains need to be laid at a narrower
spacing than normal. In this case, a formula developed by Bruggeman (Van Drecht,
1983; Bruggeman, 1999) can be used. However, in severe cases, where the drain spacing
must be greatly reduced, it is often better to leave the area as a wetland.
BRUGGEMANS FORMULA FOR ARTESIAN CONDITIONS
For horizontal drainage under artesian conditions, Bruggemans method may be
used. This calculates flow below drain level under the following circumstances
(Figure A18.1):
a moderately permeable top layer, in which the drains are located, overlies a highly
(infinitely) permeable aquifer;
between the top layer and the aquifer a semi-confining layer (aquitard) occurs;
the artesian head in the aquifer may be above drain level as well as below (in the
latter case, natural downward drainage will occur);
the artesian head is not influenced by the drainage system.
The final condition is seldom
respected in large projects. Such
FIGURE A18.1
works usually exert a profound
Drainage under artesian conditions the Bruggeman
influence on the underlying aquifer.
method
This limits the applicability of the
method to rather small areas. In Piezometric head
in aquifer h
large projects, combination with
Recharge R
a geohydrological model of the
aquifer is indispensable. The model
z
Topsoil above drains
D1
SAHYSMOD (ILRI, 2005) can
hor. permeability K
h
anisotropy an
be used for this combination. It
also allows an analysis of the salt
x=0
x = L/2
x = L Topsoil below drains
D2
balance.
hor. permeability K
anisotropy an
Because flow above drain level
is not considered in the Bruggeman
Aquitard, c
formulae, the ellipse equation can
Aquifer
Seepage S
infinite permeability
be used to calculate this part of the
flow.

Guidelines and computer programs for the planning and design of land drainage systems

180

Spacings are to be calculated for two cases:

high recharge by heavy rain or irrigation, in combination with a criterion for
groundwater table depth under such wet conditions;
zero recharge, with a criterion for a design groundwater depth under dry
conditions, deep enough to avoid permanent wetness in humid climates and
salinization in arid regions.
For the latter, groundwater should remain below a critical depth.
Bruggeman derived the following algorithm for two-dimensional flow below drain
level under artesian conditions (Figure A18.1):

D2
u
1 u 1 (P q1 )+ h
c +

K2v
L

Q2 =

D
c + 2

K2v
cb

+
+ 1
u
L
1 =

aB L2

u K2v
3

F (n,z) =

n =1

1
n u
sin 2
F (n,0)
3
n
L

(n + 1) e + (n
(n + 1) e (n
n

aB =

K2v

1 =

K2h

(1)

1
1

1) e -n

1) e -n

2 K2v c

(3)
3

2 =

aB L

(2)

2 (D2 y )
aB L

3 =

2 D2
aB L

(4)

At drain level, where y = 0 and 2 = 3:

F (n,0) =

(n
(n

1
1

+ 1)+ (n 1 1) e -2n

+ 1) (n 1 1) e -2n

(5)
3

The flux density is:

q2 =

Q2
L u

where:
c = resistance of semi-confining layer (d);
cb = entry resistance of drain (cb = 0) (d);
D2 = thickness of layer below drain level (m);
h = head midway, at drain level (m);
ha = head in artesian aquifer, above drain level (m) (in Figure A18.1);
K2h = horizontal permeability below drains (m/d);
K2v = vertical permeability below drains (m/d);
L = drain spacing (m);
q2 = flux density below drain level (m/d);
Q2 = flux below drain level, per metre of drain (m2/d);
R = recharge by precipitation or irrigation excess (m/d) (in Figure A18.1);
u = wet circumference of drain (m);
y = vertical coordinate, positive downward (m).

(6)

Annex 18 Drainage under vertical seepage

For artesian conditions and a two-layer profile (one of which is below drain level),
the design program ARTES was developed. It is based on Bruggemans algorithm, in
combination with flow above drain level according to the ellipse equation.
It also requires general design criteria. These are followed by the soil properties,
which now include the hydraulic head in the underlying artesian aquifer and the
vertical resistance of a semi-confining layer between the aquifer and the two top layers
mentioned.
An approximation is to use Hooghoudts formula with the expected seepage from
below added to the recharge from above. In most cases, the difference in spacing is
negligible in practice (less than 510 percent). However, there are exceptions, especially
where the resistance of the semi-confining layer is low and part of the drainage water
passes through the aquifer.
ARTES uses the Bruggemans method except in the rare cases where this procedure
is not convergent or is otherwise doubtful. Then, the Hooghoudt approximation is
given, together with a warning.
REFERENCES
Bruggeman, G.A. 1999. Analytical solutions of geohydrological problems. Dev. Wat. Sci., 46:
331.
ILRI. 2005. SAHYSMOD, spatial agro-hydro-salinity model. Version 1.7. Description of
principles, user manual and case studies. SAHYSMOD working group of ILRI. Wageningen,
The Netherlands. 134 pp.
Van Drecht, G. 1983. Calculation of steady groundwater flow to ditches. RID-Mededelingen 833. In Dutch with English abstract. The Netherlands, National Institute for Water Supply.

181

183

Annex 19

Formulae for non-steady-state

flow to drains

FLOW ABOVE DRAINS THE

FIGURE A19.1
BOUSSINESQ SOLUTION
Non-steady-state flow above drain level the Boussinesq
In 1904, Boussinesq found a
solution
solution for non-steady-state
(transient) flow to drains lying on
t = 0 ponded h 0
an impermeable subsoil layer (K2
t=
= 0), as occurring after heavy rain
or irrigation. Boussinesqs equation
(Boussinesq, 1904; Guyon, 1966;
t = + 1
Moody, 1967) describes the fall
t = + 2
in the water table after recharge.
Where the initial shape of the
t = + 4 days
groundwater between the drains
follows a special curve (nearly an
h(t) after t days
K
ellipse), it retains this shape during
x=0
x=L
the drainage process because the
head diminishes proportionally
everywhere. It can be shown that,
soon after the end of the recharge event, the shape of the groundwater table becomes
almost elliptical, and during its lowering, the curve becomes flatter, but retains its
shape (Figure A19.1).
If the soil surface is ponded and the soil profile is completely saturated at the
beginning, the theory is not valid for short times. The lowering of the water table
reaches the mid-point between drains only after some lag time , being the time to
approach Boussinesqs pseudo-ellipse, after which a phreatic surface of constant shape
is approached. The lag time is approximately:
(1)
where:
C = 38, this is an empirical constant derived from numerical experiments;
Z = drain depth (m);
h0 = initial head midway the drains, equal to drain depth (m);
K = permeability above drain level (m/d);
L = drain spacing (m);
= storage coefficient;
= lag time (d).
Boussinesqs formula is a solution of the non-linear differential equation:
(2)

184

Guidelines and computer programs for the planning and design of land drainage systems

Based on this solution, Guyon proposed the following formula for calculating drain
spacing (with = 0), valid for Boussinesqs pseudo-ellipse:
(3)
where:
h = hydraulic head midway, at time t (m);
h0 = initial head midway between drains (at time t = 0) (m);
K = soil permeability (m/d);
L = drain spacing (m);
t = time (d);
= lag time (d);
= storage coefficient.
The factor 4.5 is an approximation of an expression that yields 4.46208
If the lag time has to be considered, the L value may be calculated with the
following formula, obtained by combining Equations 1 and 3:
(4)

Equation 4 is the non-steady-state flow equivalent of the steady-state flow ellipse

equation. The program NSABOVE, which is based on this equation, describes the
flow to drains lying on an impermeable soil layer. The shapes of the water table closely
resemble semi-ellipses of decreasing height.
FLOW ABOVE AND BELOW DRAINS NUMERICAL SOLUTION
Analytical approximations (GloverDumm, and Kraijenhoff van de Leur) can be
used to calculate drain spacings where h << D2. However, these solutions do not
consider radial resistance and resistance near the drain. Therefore, numerical methods
are preferable because they are easier to handle and are accurate enough for practical
purposes. Moreover, evaporation losses, which vary with the depth of the phreatic level
and also the effect of outflow restrictions, can readily be incorporated. The latter are
caused by the radial resistance concentrated near the drain and the limited capacity of
the collecting system.
For drains lying above an impermeable soil layer (Figure A19.2),
FIGURE A19.2
the flow below the drain level
Non-steady-state flow above and below drain level
must be considered through a layer
with a transmissivity KD2. The
Surface
permeability K is the same above and
h (1)
below drain level (K1 = K2 = K) and
D1
D2 the thickness of the layer below
h 0(1)
h (2)
h p(1)
drain
level.
h p(2)
h0(2)
After a heavy rain, the water
levels in the watercourses and the
Drain level
or ditch bottom
head in the pipes will be higher than
designed. This will in turn restrict
D2
the outflow from the soil until
K
equilibrium is reached. In view of
Impermeable base
the turbulent flow in pipes, their
behaviour is supposed to follow a
(1) - (2): successive stages
square-root function at four times
the design head, the outflow will be

Annex 19 Formulae for non-steady-state flow to drains

185

twice the design discharge. It is further supposed that, at design discharge, no water is
standing above the drain (hp = 0).
The outflow is further restricted by the radial and entrance resistance near the drain.
This quantity is given as Wr in the program SPACING and here denoted as resistance
W. It causes a head loss proportional to the flow.
Evaporation aids in lowering the groundwater, but it decreases rapidly with
increasing groundwater depth. For this relationship, there are two options:
linear reduction to zero at a given groundwater depth;
exponential reduction with a given characteristic groundwater depth where E =
0.4343Eo.
These principles form the framework of the programs NSDEPTH and NSHEAD
to check calculated drain spacings under non-steady-state flow.
Principle for numerical solution
The principle for numerical solutions is that both time and (horizontal) space are
divided into discrete elements and steps. In each element, the water balance during one
time-step is:
(5)
where:
Qin = flux entering an element, per metre of length (m2/d);
Qout = flux leaving an element, per metre of length (m2/d);
x
= distance (m);
h = fall of groundwater table (m);
t = time-step (d);
x = distance step (m);
= storage coefficient.
To develop this principle into a calculation program, both explicit and implicit
methods are possible. The programs use the first approach although the risk of
instability requires small time-steps t.
Differential equation
For flow below the drain level in the area D2 (Figure A19.2) and a permeability K being
the same above and below drain level (K1 = K2 = K), Equation 2 becomes:
(6)
where:
D2 = thickness of layer below drains (m).
The explicit finite difference expression for Equation 6 is:
(7)
where:
h = hydraulic head (m);
i = index for distance step;
j = index for time step;
x = distance (m);
t = time-step (d);
x = distance step (m).
In the model based on this equation, the drain spacing L has been divided into
20 equal parts. Index i = 0 represents the left-hand boundary; and i = 10 is a plane of

Guidelines and computer programs for the planning and design of land drainage systems

186

symmetry that forms the right-hand boundary (midway between drains). Therefore,
index i = 11 is the highest used. In the drainpipe, the head is hp, near the drain it is ho.
Boundary conditions
The initial condition (j = 0) is a constant head everywhere between the drains (i.e.
groundwater at the soil surface):
hi,0 = hinit

i = 1,11

(8)

The right-hand boundary condition simulates symmetry at i = 10:

h9,j = h11,j

(9)

The left-hand boundary is more complicated. Here, two types of resistance against
flow are present:
a linear resistance W (d/m) against total flow (from both sides), being the sum of
the radial resistance (caused by convergence of streamlines near the drain) and
entry resistance for flow into the drain;
a non-linear resistance, caused by the limited capacity of the outflow system
(usually the drainpipes). Here, flow is turbulent and proportional to the square
root of the available head.
For the one-sided flow q0 (in cubic metres per day per metre of drain) converging
towards and entering into the drain:
q0 =

h0 hp

(10)

2W

where (Figure A19.2):

= head near drain (m);
h0
= head in drainpipe (m);
hp
q
= flux density to drain (m/d);
= flux to drain (absolute value), one-sided (m2/d);
|q0|
qL/2 = flux, one-sided (m2/d);
W
= total resistance near drain (radial + entry) (d/m).
For the pipe flow, the outflow system has been designed to discharge a given steady
flux density q (in metres per day) at a given head hdes (usually the slope multiplied by
the pipe length).
For larger discharges, there is a need for an extra head hp caused by insufficient pipe
capacity. Thus, for one-sided flow, originating from width L/2:

q0 =

qL hp + hdes
hdes
2

if

hp > 0

(11)

where:
hdes = design head for outflow system (m).
Finally, for horizontal flow in the first compartment:

h1 + h0
h1 h0
q0 = K
+ D2
2
x
where:
h1 = head in first compartment (m).
Equalizing Equations 1012 yields two equations in the unknown h0 and hp.

(12)

Annex 19 Formulae for non-steady-state flow to drains

187

The upper boundary receives a sudden large input at t = 0, that saturates the entire
soil profile. For t > 0, evaporation may help in lowering the water table, but it is
dependent on the groundwater depth. Two options are available in the model:
linear decrease with groundwater depth z;
exponential decrease.
The linear case is characterized by the critical depth zc:

E = E0 1

z
c

E=0

for

else

(13a)

(13b)

where:
E = actual evaporation from groundwater (m/d);
E0 = potential evaporation from groundwater (m/d);
hinit = initial head = drain depth (m);
zc = critical depth where E = 0 (linear model) (m);
z = groundwater depth (hinit - h) (m).
The exponential case is characterized by the characteristic depth zh:
E = E0 e

z
zh

(14)

where:
= depth where E = 0.4343E0 (exponential model) (m).

zh

Solution for h0 and hp (W > 0)

The relation:
h0 hp

h1 + h0
h1 h0
= K
+ D2
2W
2
x
leads to the quadratic equation:
x
2
x

2
h0 +
+ 2 D2 h0 h1 + 2D2 h1 +
hp = 0
KW
KW

(15)

(16)

The solution for the head in the drain is:

2

h0 = U + U + V

(17a)

where:
U=

x
+ D2
2KW

(17b)

V = h1 + 2 D2 h1 +
2

xhp
KW

(17c)

The relation for the head near the drain is found as follows:
h0 hp
2W

qL
2

hp
hdes

+1

(18)

Guidelines and computer programs for the planning and design of land drainage systems

188

Equation 18 leads to the quadratic equation:

2 2
2

q L W
2 2
2
2

hp 2 h0 +
hp + h0 q L W = 0

hdes

(19)

with solution:
2

h0
q L W
q L W
hp = h0 +
qLW
+
+1
2
hdes
2hdes
4hdes

(20a)

( )

hp = max hp ,0

(20b)

Iteration starts with Equation 17, with hp = 0 in (17c). The value of h0 obtained
from Equation 17 is used in Equation 20 to find a new hp value, which is inserted in
Equation 17, etc., until convergence is sufficient.
The process is repeated before each time-step. With h1,j = h0 and h2,j = h1 Equation 7
is used to find the new values for the next time-step.
The index F is used as a criterion for stability of explicit numerical calculations:
F =

KDt

(x)
where:

(21)

D = D2 + h0 = maximum initial thickness (m).

The explicit method is valid for
small time-steps and index F only.
The characteristic:

FIGURE A19.3

Stable solution at F = 0.1

% of drain spacing
50 %
30 %
20 %
10 %
5%

FIGURE A19.4

Instability at F = 0.6

% of drain spacing
50 %
30 %
20 %
10 %
5%

F =

K t
(D2 + hinit )
(x) 2

(22)

should be less than 0.5 in order to

avoid instability (Figure A19.3), and
preferably be 0.25 or less (about 0.1)
for sufficient accuracy. Figure A19.4
shows an example of instability.
The methods described, for flow
above and below drain level through
layers with the same K and values
have been used in the programs
NSDEPTH and NSHEAD. These
programs check whether the three
values for |q0| from Equations 10, 11
and 12 are indeed equal.
Finally, the water balance is
checked. Errors should not exceed
5 percent. If difficulties arise, a
smaller time-step is usually helpful.

Annex 19 Formulae for non-steady-state flow to drains

REFERENCES
Boussinesq, M.J. 1904. Recherches thoriques sur lcoulement des nappes deau infiltres dans
le sol et sur le dbit des sources. J. Math. Pures Appl., 10 : 178.
Guyon, G. 1966. Considrations sur le drainage des nappes. Thorie et Experimentation. Bull.
Tech. Gn. Rur., 79.
Moody, W.T. 1967. Nonlinear differential equation of drain spacing. J. Irrig. Drain. Div. ASCE,
92(IR2): 19.

189

191

Annex 20

Diameters of drainpipes

Elevation, m

PRINCIPLES
Drains are collecting systems. Along their length, the discharge and the flow velocity
increase gradually. Therefore, the gradient of the hydraulic head is zero at the beginning,
and will increase downstream.
Most drains are laid with a certain slope, and this slope is usually taken as a basis
for calculating the required diameter. However, not the drain slope, but the total head
loss is the basic design parameter.
At the upstream end, the hydraulic
head should remain at a certain
FIGURE A20.1
Sloping drain
depth below the soil surface, and
this depth determines the available
1
head with respect to the drain outlet,
irrespective of the pipe length. The
0.8
slope is not important, as illustrated
in the following example. A drain
0.6
200 m long with an outlet 1.50 m
below surface and a slope of
0.2 percent, without water standing
0.4
above the upper end, loses 0.40 m in
height along its length. Thus, it will
Hydraulic head
0.2
control the upstream water table at a
Drain
depth of 1.10 m. However, the same
will be the case for a horizontal drain
0
(slope zero) of the same length and
0
20
40
60
80
100
Distance, m
outlet depth if it loses 0.4 m in head
over its length owing to friction.
As an example, at the design
discharge intensity q (metres per
day for pipe flow, q is recalculated
FIGURE A20.2
and expressed in metres per second),
Horizontal drain
the drain is running full at the outlet
1
and the head at the beginning has
a design value H (m) above the
0.8
outlet. The drain itself has a slope,
and the slope is such that no water
is standing above the drain at its
0.6
beginning (Figure A20.1). If the
slope is less and also when the
0.4
drain is horizontal (Figure A20.2)
there is water above the drain at
Hydraulic head
the upper end.
0.2 H
From a hydraulic point of view,
B
the drain is functioning equally
Drain
0
well in both cases. Sometimes self0
20
40
60
80
100
Distance, m
cleaning is used as an argument for
having the drain slope. However, in
Elevation, m

120

120

Guidelines and computer programs for the planning and design of land drainage systems

192

flat lands, drain slopes are seldom more than 0.5 percent and often far less. At such low
slopes, the flow velocity is not enough to move sediments.
However, in practice, a slope for the pipe is usually prescribed. Horizontal drains
are seldom encountered, except in subirrigation projects where drains are used for
discharge in wet seasons and for recharge during droughts.
In the following, the system of Figure A20.1 is considered exclusively. Calculation
of the diameter of horizontal drains (Figure A20.2) with formulae for sloping ones
(Figure A20.1) sometimes shows small differences, but they are always on the safe
side.
The available head loss at design discharge and the amounts of water to be drained
under that condition form the basis for calculations concerning required drain
diameters. These calculations are based on the laws for pipe flow, which differ for
smooth and corrugated pipes.
Both smooth and corrugated pipe drains collect water along their length. As a
consequence, the flow is not constant, but it increases gradually from zero at the
upstream end to a maximum at the outflow. Introducing this variable Q corresponds
with integration of the expressions for laterals and collectors. In laterals, Q increases
continuously; in collectors, flow occurs stepwise, namely where the collector is joined
by another lateral. However, provisional calculations show that in practice this makes
almost no difference, provided that the laterals are of equal length.
SMOOTH PIPES
Non-perforated pipes made of glass, metals, PVC, PE and similar materials may be
considered as hydraulically smooth. Pipes that are perforated or made of ceramics
or cement are technically smooth, in which case they obey the same laws, but
with a slightly different roughness coefficient. Corrugated pipes are hydraulically
rough.
Basic equations
For smooth pipes, the DarcyWeissbach equation is valid:
(1)
where:

or

(Blasius)

(2a)

(Nikuradse)

(2b)

with:
a = coefficient;
d = pipe diameter (m);
g = acceleration of gravity = 9.81 m/s2;
h = hydraulic head (m);
Re = Reynolds number for pipes;
v = flow velocity (m/s);
x = distance along pipe (m);
= coefficient;
= kinematic viscosity ( 10-6 m2/s).
Both expressions for give comparable results (Table A20.1, for a = 0.3164). Because
Equation 2b requires iteration, Equation 2a is normally used.

Annex 20 Diameters of drainpipes

193

TABLE A20.1

Comparison between -Blasius and -Nikuradse

-Blasius
0.0473
0.0376
0.0316
0.0266
0.0212
0.0178

Reynolds number
2 000
5 000
10 000
20 000
50 000
100 000

-Nikuradse
0.0495
0.0374
0.0309
0.0259
0.0209
0.0180

% difference
4.6
-0.6
-2.4
-2.7
-1.2
1.1

TABLE A20.2

Values for the a coefficient in Blasius formula

Type of pipe

a coefficient

Remarks

Smooth, plastic, metal, glazed

0.3164

Non-perforated or well jointed

Technically smooth

0.40

Perforated, cement, ceramics

Corrugated plastic laterals

0.77

Zuidema, from field data1

Theoretically not allowed for hydraulically rough pipes, but in accordance with field data for small-diameter corrugated drains.

Completely smooth laterals and collectors do not exist. Smooth plastic pipes contain
perforations; ceramic and baked clay ones have joints and are not always aligned. For
such technically smooth drains and collectors, the a coefficient in Equation 2a was
taken as 0.40 instead of 0.3164. Table A20.2 shows values used for the a coefficient, as
found in the literature.
Smooth laterals
Drain laterals collect additional water all along their length. At any point x, measured
from their upstream end, the discharge Q and the velocity v are:
and

(3)

where:
L
= drain spacing (m);
q
= design discharge (m/s);
Q = drain discharge (m3/s).
Accordingly, the flow velocity v varies along the length and so does the Reynolds
number.
Inserting v in the basic equations (Equations 1 and 2a) leads to:
(4)
and integrating between x = Bi-1 and x = Bi:

(5)
with:
Bi-1, Bi = begin, end of a drain section (m);
Fs
= calculation coefficient for smooth pipes;
n
= 11/4;
H
= head loss in the drain (m).
In drains consisting of one pipe size only, Bi-1 = 0. However, the full expression will
be needed later for drains with increasing pipe diameters downstream (multiple drains).
The head loss H in the drain must be less than or at most equal to the design head loss
over the entire drain length, H.

Guidelines and computer programs for the planning and design of land drainage systems

194

If Bi-1 = 0, the permissible drain length B for this design head equals:

11 7/4
B=
16 2

gHd 19/4
av1/4 q 7/4 L7/4

4/11

(6)

and the minimum diameter required for a given drain length B is:

16 2
d=
11 7/4

av1/4 q 7/4 L7/4 B 11/4

gH

4/19

(7)

The maximum drain spacing allowed at a given diameter amounts to:

7/4
gHd 19/4
11
L=
16 2 av1/4 q 7/4 B 11/4

4/7

(8)

For hydraulically smooth, new, collecting pipes the required head can be calculated
with:
H=

(qL)
59 .77

7/4

7/4

19/4

4 11/4
B
11

(9)

where conversion of units, physical and mathematical parameters, and integration

have caused the numerical constants. An alternative formula for technically smooth
pipes is: Q = 89d 2.714 s0.571 (FAO, 2005), where Q = qLB and s = H/B. It gives almost the
same results as the above formulae with a = 40.
In Equations 69:
a = Blasius coefficient;
B = drain length (m);
d = inside diameter (m);
g = acceleration of gravity (m/s2);
H = head loss in drain (m);
L = drain spacing (m);
q = specific discharge (m/s).
Smooth collectors
Where the laterals are of equal length, the same formulae may be used for designing
collectors with added flows at each lateral connection. Now, Lc is the mutual distance
between collectors and Bc the length of the collector (Lc is the symbol for collector
spacing and Bc for its length. If the laterals are perpendicular to the collectors and the
laterals flow from one side only Lc equals their length B. If inflow is from both sides,
Lc = 2B). For collectors, both are substituted for L and B in the formulae for laterals.
The difference from lateral design is that the flow into collectors is discontinuous, in
contrast to laterals, where inflow may be considered as continuous along the pipe.
However, where more than five laterals are involved, the discretization error caused
by the inflow of the separate laterals may be ignored in practice.
In the case of unequal lengths of the contributing laterals, the collectors must be
calculated section-wise, in which case the discontinuous inflow is accounted for.

Annex 20 Diameters of drainpipes

195

CORRUGATED PIPES
Basic equations
Most authors calculate flow through corrugated pipes with Mannings equation:
Q = Km AR 2/3 s 1/2

(10)

where:

d
A =
2
Km
R=

= area of cross-section (m2);

= 1/n = Manning coefficient (m1/3s-1);

A d r
= hydraulic radius (m);
= =
u 4 2

= slope of H;

d
= wet circumference (m).
u = 2
2
The formula for smooth pipes is sometimes used for corrugated pipes, but with a
much larger constant a (Zuidema and Scholten, 1972), whereas other authors (e.g. Van
der Beken, 1969, Van der Beken et al., 1972) introduce an equivalent sand roughness
to account for the influence of the corrugations.
Mannings Km for corrugated pipes
In Mannings equation, the constant Km depends mostly on the spacing, depth and
shape of the corrugations S and also on the diameter d. The Km values for corrugated
pipes are compiled in Table A20.3. The narrower the corrugation spacing S, the larger
Km. According to Irwin (1984) and Boumans (1986):
Km = 70

for S < 0.01 m

Km = 18 .7 d 0.21 S 0.38

for S > 0.01 m

(10 mm)

(11a)

(10 mm)

(11b)

where:
d
= inner pipe diameter (m);
S
= spacing of individual corrugations (m).
Equations 11a and 11b for Km are used in the programs for corrugated pipes. For
safety reasons, the maximum value is taken as 65 instead of 70.
Corrugated laterals
If for full flowing pipes, Equations 3 and 10 are solved for Q:

qLx = Km

d
d2
4
2

2/3

s 1/2 = Km

d 8/3 dh

4 5/3 dx

1/2

(12)

The head loss H between points B1 and Bi-1 can be obtained by integrating
Equation 12 between these points:

dh =

4 10/3 (qL)

K d
2

2
m

16/3

x 2 dx

(13)

Guidelines and computer programs for the planning and design of land drainage systems

196

TABLE A20.3

Km values for corrugated pipes

Country

Material

Netherlands

PVC

Germany

PVC

Unite States of America

PE

United Kingdom

PP

Drain diameter d
Outer
Inner
(mm)
65
57
80
72
100
91
160
148
60
52
100
91
125
115
380
307
129
100
196
171
265
225
350
305

with n=3

Rib spacing S

Km value

(mm)
6.25
6.25
6.25
7.50
6.30
8.30
8.30
50.00
18.00
20.00
33.00
50.00

70
74
78
80
69
70
73
46
53
57
50
45

(14)

with:
Fc = calculation coefficient for corrugated pipes.
As mentioned above, in drains consisting of one pipe size only Bi-1 = 0. For
corrugated pipes, integration of Mannings equation results in:

(15)

For corrugated pipes with small corrugations an alternative formula is (FAO,

2005):
For corrugated pipes with a diameter of more than 200 mm
and large corrugations an alternative formula is (FAO, 2005):
Both give
almost the same results as those mentioned in the text.
Where the design head H is given, and Bi-1 = 0, the other values (e.g. d or L) are
readily derived from Equation 15. Thus, the permissible length B is:

(16)

Corrugated collectors
If the collectors have the same spacing Lc, the same formulae may be used for their
calculation, substituting their spacing Lc and length Bc for L and B. If they do not
have the same spacing, calculations have to be made separately for each section of the
collector. The spacing of laterals, and, thus, the distances of inflow points along the
collector, has only little effect, provided that more than five laterals are involved.
MAINTENANCE STATUS AND REDUCTION FACTORS
The problem of clogging of drainpipes
In practice, drains are seldom completely clean. This is because some siltation always
occurs, notably during and shortly after construction owing to the entrance of soil
particles from the yet unsettled soil and/or envelope around the pipe when relatively
large amounts of water enter. A layer of sediment usually forms over time. This

Annex 20 Diameters of drainpipes

sediment should be removed by maintenance, where it reduces the transport capacity

of the pipe too much. Siltation may be also caused by other materials, e.g. iron oxides.
Moreover, plant roots as well as certain animals may enter into drainpipes and hamper
their proper functioning. Detailed information about the problem of clogging of pipes
and envelopes is given in FAO (2005).
Siltation differs greatly from place to place and even in the same drainpipe. In
particular, sunks in the alignment of the pipe cause siltation problems. Therefore, drain
installation design and construction practices should take care to avoid the presence of
such vulnerable stretches.
Entry of soil and plant roots can be prevented largely by a good envelope around
the drains, by construction at sufficient depth, or by using non-perforated pipes for
the stretch that crosses under a row of trees. However, for clogging by chemical
precipitates, such as iron, this is not the case.
In addition to the effectiveness and durability of the drain envelope, the clogging
of drains is connected with cleaning operations and their frequency. Drainpipe
maintenance frequency depends on soil conditions and other circumstances. It is hardly
needed for well-constructed drains surrounded by a stable soil or by an envelope and
without iron precipitation phenomena, whereas in others deterioration is rapid. The
latter is often the case under artesian seepage, which often induces ochre deposition,
and in acid sulphate soils (cat clay soils and cat sands), where precipitation of iron
compounds is also common.
Therefore, the design usually allows for a certain amount of clogging, which
depends on the geohydrological and soil conditions at drain level and on the anticipated
frequency of inspection and cleaning.
Maintenance status
To take account of the aspects described above, the maintenance status is used as
a parameter in the programs for calculating drain diameters. As mentioned above,
maintenance status is a combination of:
local circumstances (envelope materials, soils, ochre formation, etc);
maintenance operations (frequency, intensity, availability of adequate equipment,
etc).
Maintenance status has little to do with a specified rate of cleaning, but it is an
indication of the state of cleanliness in which the drains can be kept under the given
conditions. Under certain conditions,
almost no maintenance is needed to
FIGURE A20.3
realize a good maintenance status.
Drain with sediment layer
This is the case with well-constructed
drains in stable soil layers. In other
E
conditions, much effort is required
to keep it fair, as is the case with
unstable silt soils and where iron
clogging is a severe problem.
This means that under an expected
M
poor maintenance status even
r
r
frequent cleaning is not sufficient.
h
L/2
B L/2
Hence, larger diameter pipes should
A
C
be used than under an excellent
L
status. Therefore, a reduction should
D
be applied to the described formulae,
by multiplying Km with a correction
factor f (e.g. f = 0.8).

197

198

Guidelines and computer programs for the planning and design of land drainage systems

Mannings Km for drains with sediments

Figure A20.3 shows a drain AECD, with radius r, which is partly filled with sediment
ABCD. The thickness of this layer BD is l, and the distance BM from the centre M
is h.
For a clean pipe, Mannings formula can be written as:
(17)
A correction for the sediment layer is obtained as follows.
The angle AMC is, so AMB = BMC = /2.
The thickness of the layer is:
and
The area available for water flow

(18)
is:
(19)

where the angle is expressed in radians.

Thus, the reduction factor for diminished area (A' instead of A) is:

(20)
The hydraulic radius was R = r/2 and becomes:

(21)

Thus, the reduction factor for R is:

(22)

Therefore, the drain discharge is reduced to:

(23)
The correction factor for Km is:
and

(24)

Table A20.4 shows the f values calculated for different fractions of sediment height
and area. These values are represented in Figure A20.4.
Categories according to maintenance status
For the reasons discussed above, maintenance can only be specified in a global way. From
the data in Table A20.4, the following choices were made with respect to maintenance
status by distinguishing five categories. These categories have been defined in terms of
the relative height of sediments in the drainpipes (Table A20.5). Table A20.5 shows the
influence of maintenance status on the flow in partially clogged drains.

Annex 20 Diameters of drainpipes

199

The maintenance status should be

envisaged in the design stage. As only
a rough classification is possible,
the categories in Table A20.5 have
been distinguished, for which the
corresponding f values have been
used in the programs. For these
maintenance groups, the f factors
will be used in the programs for
drainpipe design. The f values are
valid for Mannings equation. To
avoid unnecessary complications,
the programs also use these values
in the DarcyWeissbach approach
for smooth pipes. The Km values are
multiplied by f to obtain corrected
values Km, and the coefficients
a must be divided by
to obtain
corrected values for ac:

FIGURE A20.4

Correction factors for Mannings Km

f
1.2

1.2

1
0.8

Relation with height

Relation with area

0.6

0.8
0.6

0.4

0.4

0.2

0.2

0
0

0.2

0.4
0.6
0.8
1
Height and area of sediment fraction

0
1.2

TABLE A20.4

(25)

ZUIDEMAS METHOD FOR

CORRUGATED LATERALS
From numerous observations
on existing corrugated laterals,
Zuidema and Scholten (1972) found
good agreement with Blasius
formula where a larger a coefficient
was taken. They recommended
using the value a = 0.77 for these
pipes. This method is included
as an option in the programs. It
appears that the results obtained in
this way are similar to or slightly
more conservative than those for
Mannings equation with narrow
rib spacing (in the programs Km
= 65) and with a correction factor
f = 0.923, corresponding to good
maintenance.
DRAIN LINES WITH INCREASING
DIAMETERS
The above considerations refer
to drains composed of one pipe
diameter only. Long laterals and
collectors usually require pipes
of successively larger dimensions.
Because of the rapid increase in
prices with size, it often pays to
replace the upstream part of the

Correction factor f for pipes with sediment

Fraction of
sediment
height

Fraction of
sediment
area

Factors

l/2r

1 - A/A

f1

f2

.050

.019

.981

.986

.972

.100

.052

.948

.961

.923

.150

.094

.906

.930

.863

.200

.142

.858

.894

.796

.250

.196

.804

.854

.724

.300

.252

.748

.810

.650

.350

.312

.688

.764

.575

.400

.374

.626

.715

.501

.450

.436

.564

.664

.429

.500

.500

.500

.611

.360

.550

.564

.436

.556

.295

.600

.626

.374

.500

.235

.650

.688

.312

.441

.181

.700

.748

.252

.382

.133

.800

.858

.142

.259

.058

.850

.906

.094

.196

.032

.900

.948

.052

.131

.013

.950

.981

.019

.066

.003

.990

.998

.002

.013

.000

TABLE A20.5

Flow reduction in partially clogged drains

Maintenance
status

Cross-section clogged

Reduction factor for flow f

(%)
New pipe

1.000

Excellent

0.972

Good

10

0.923

Fair

20

0.796

Poor

40

0.501

Guidelines and computer programs for the planning and design of land drainage systems

200

FIGURE A20.5

Drain line with two diameters and a given slope; transition

point B at the drain

0.4

system where the flows are still

small by a section of smaller size
pipe, and use gradually larger ones
downstream. The following sections
consider drains of two sizes.

Effect of drain slope

Where a multiple drain is running full
and slopes over the entire head (full
slope), the head at the transition
A
0.2
Head
cannot fall below the top of the drain
at that point. This is illustrated by
Top
of
Figure A20.5, where the transition
B
drain
0.1
point B lies at the top of the drain.
In Figure A20.5, it may be
observed that drain AC, with given
C
Diameter 50 mm
80 mm
0
slope (0.20/150 m/m) consists of 500
50
100
150
200
mm and 80-mm pipes. B is a critical
Distance, m
point determined by the head loss in
the first section. The drain is running
full and the head is not allowed to
FIGURE A20.6
fall below the top of the drain. At
Drain line with two diameters and a given slope; transition
C, some head is still available. Thus,
point B above the drain
the system is not very efficient. The
0.4
consequence is that the drain has
excess capacity and that the available
head is not used entirely for water
0.3
transport. The outlet at C could even
be drowned to satisfy the design
head at point A.
0.2 A & A'
Figure A20.6 gives an example
Head
where this is not the case, because the
Top
transition point B lies above the drain
B
of
drain
0.1
and the full available head is used.
Drain AC, with given slope,
consists
of 50-mm and 70-mm pipes.
C
Diameter 50 mm
70 mm
0
B is not critical and the hydraulic
0
50
100
150
200
grade line lies above the drain, at the
Distance, m
intersection of the curves AB and
BC. No extra head is available at C.
In the programs, attention has been given to these aspects.

Elevation, m

Elevation, m

0.3

Given slope
It is supposed that the drain slope equals s = Ht/Bt so that at design discharge there
is no water above the upper end of the drain (Figure A20.5).
The first section AB has a length B1, governed on the one hand by Equation 5 or 14;
on the other, by the given slope. From the latter, it follows that the head loss in this
section equals:
(26)
where H is the design head.

Annex 20 Diameters of drainpipes

201

Inserting H1 in Equation 5 or
14, with B0 = 0 (first section), and
rearranging, leads to:

FIGURE A20.7

Hydraulic heads in a full drain, composed of pipes with

different diameters

(27)

0.25

0.2
For smooth pipes, F1 = Fs with d =
Head in pipe 0.08
d1 and m = 4/7; for corrugated pipes,
F1 = Fc with d = d1 and m = 1/2
0.15
If B1 exceeds the total length B,
the first section is already sufficient
0.1
Head in pipe 0.12
to meet the requirements. In this
case, a combination with narrower
0.05
pipes might be used.
The second section, with diaDiameter 0.12 mm
Diameter 0.08 mm
0
meter d2, causes a head loss H2,
0
100
200
300
for smooth drains according to
Equation 5, for corrugated pipes to
Equation 14. The factors Fs or Fc are
now calculated with d = d2 .
The total head loss Ht = H1 + H2 must be smaller than or at most equal to the
required H. If greater, a second section with larger diameter must be chosen or another
combination be tried.

Hydraulic heads along the drain

The available head along the drain depends on the distance x from the beginning. In the
programs, they are expressed as head above outlet level. Thus:
i = 1,2

(28)

For the first section, where x B1 , Hx is calculated from:

(29)
and for the second, where x > B1:
(30)
For smooth pipes:
F1 = Fs with d = d1 and n = 101/4;
F2 = F with d = d2 and n = 11/4.
For corrugated pipes:
F1 = Fc with d = d1 and n = 3;
F2 = Fc with d = d2 and n = 3.
The considerations given above form the basis of the program DRSINGLE for
the design of smooth and corrugated laterals and collectors consisting of one section.
For two or more sections with different diameters and also of different types, the
program DRMULTI can be used. Figure A20.7 shows a longitudinal profile along
such a drain. At the upstream end, no water is standing above the drain, and at the
outlet downstream there is still some head available. This indicates that the proposed
combination is sufficient to carry the design discharge.

400

202

Guidelines and computer programs for the planning and design of land drainage systems

REFERENCES
Boumans, J.H. 1986. Stromingsformules voor drainage ribbelbuis. Cultuurtech. Tijdschr., 26:
4146.
FAO. 2005. Materials for subsurface land drainage systems, by L.C.P.M. Stuyt, W. Diericks & J.
Martnez Beltrn. FAO Irrigation and Drainage Paper No. 60 Rev. 1. Rome. 183 pp.
Irwin, R.W. 1984. Corrugated pipe flow rate from pipe geometry. J. Irr. Dr. Eng., 237241.
Van der Beken, A. 1969. Bijdrage tot de hydraulica van draineerbuizen. Meded. Rijk. Landbouw.
Publ., 35/WB-5. 87 pp.
Van der Beken, A, De Troch, F., De Somer, M. & Zuidema, F.C. 1972. The lateral inflow into
submerged drains. Bull. Int. Ass. Hydrol. Sci., 17: 273289.
Zuidema, F.C. & Scholten J. 1972. De afvoercapaciteit van geribbelde plastieken draineerbuizen.
Van Zee tot Land, 49: 73102.

203

Annex 21

Interceptor drains

FLOW FROM SURROUNDINGS

Inflow from higher places and from
leaky irrigation canals can sometimes
be captured by interceptor drains,
especially where it passes through
relatively shallow aquifers. Such
drains can take the form of pipes
or open ditches. In the latter, the
stability of the side slopes is often
problematic if large amounts are
to be captured. Better solutions are
gravel-filled trenches provided with
a suitable pipe of sufficient capacity
to carry the discharge.

FIGURE A21.1

GloverDonnan analysis of an interceptor drain

Surface
D1
K
Phreatic
level
Drain

Base
(impermeable)

D0

HILLSIDES
An analysis of the interception of flow from hillsides of uniform slope was given by
Donnan (1959), as represented in Figure A21.1.
The flow from upstream, per metre of length, is:
(1)
and downstream:
(2)
The drain discharges, per metre of length, is:
(3)
where:
q1 = upstream flow per metre of length (m2/d);
q0 = downstream flow per metre of length (m2/d);
K = permeability (m/d);
D1 = upstream thickness of flow (m);
D0 = downstream thickness (m);
= angle of slope (rad).
In this analysis, the downstream flow has a thickness D0, which is entirely governed
by the distance of the drain above the impermeable base (which is governed by the
drain depth).
The upstream thickness varies from D0 near the drain to D1 far upstream. A given
thickness y appears at a distance x from the drain:
(4)

Guidelines and computer programs for the planning and design of land drainage systems

204

where:
x = distance from drain (upstream) (m).
On hill slopes, hydrological conditions are often much more complicated. Wet or
saline spots caused by seepage may sometimes be protected by an interception drain
laid at the upper end of the affected field.
This formula ignores the radial resistance encountered in the convergence of the
stream lines onto the drain. Because of this resistance, D0 has to be increased, with the
resulting head h.
In a homogeneous soil, this radial resistance can be estimated by Ernsts formula:
(5)
and
(6)

D0

where:
d = effective diameter of drain (m);
Wr = radial resistance (d/m);
hr = extra head from radial resistance (m).
In the described case of a homogeneous soil and a constant angle , this increase in
D0 will usually be slight. However in the cases described below, the consequences can
be considerable.
In most cases, an interceptor drain will be laid if: the slope decreases, the depth
of the impermeable base becomes less, or the permeability decreases. At places
where these occur, hillside flows tends to come too close to the surface and cause
waterlogging, eventually followed by soil salinization. Based on the above theory,
the program INCEP gives the required effective diameter of the drain, necessary to
diminish the radial resistance to a sufficiently low level. It is valid for a non-layered
soil (Figure A21.2), and allows jumps in thickness and permeability at the drain. The
arithmetic averages of thickness and permeability are used in order to calculate the
radial resistance.
The capacity of pipes for interceptor drains must be calculated separately from
the discharge per metre, their length and their longitudinal slope. The programs
DRSINGLE and DRMULTI can
be used for this purpose. The largest
FIGURE A21.2
value from both calculations (for
Definition sketch for program INCEP
effective diameter and for capacity)
must be taken.
Conditions become far worse
Surface
where the drain cannot reach wellpermeable subsoil and remains
D1
K1
within a less permeable top layer, a
Phreatic level
case covered by program INCEP2.
near drain: head at base
Then hr soon reaches such high
values that a single interceptor drain
h
is not sufficient, and a wide ditch or
y
Drain
even regular drainage is needed.
Base
(impermeable)
The program INCEP2 supposes
that the drain trench or open ditch
K0
has a flat bottom that is located in the
topsoil and receives the flow from the
x
permeable subsoil (Figure A21.3). In

Annex 21 Interceptor drains

205

this case, the exact solution can be found by complex transformation. An excellent
approximation for this case is obtained by calculating the parallel lines flow between
the border with the permeable subsoil and the ditch bottom with Equation 7, using a
correction factor of 0.88.
for

(7)

where:
a = distance to more permeable subsoil (K1 < 0.1K2) (m);
b = width of drain trench or ditch bottom (m);
K1 = permeability of topsoil (m/d);
K2 = permeability of subsoil (K2 > 10 K1) (m/d);
q = upward flow (m2/d);
h = difference in piezometric head above the trench bottom (m).
INCEP2 provides both solutions for b.
LEAKY CANALS AND UPSTREAM FIELDS
The same principles apply for interceptor drains catching leakage from irrigation canals
of losses from upstream fields.
For leaky irrigation canals, the best way is to reduce the water losses by lining. Where
that is impossible, and damage is occurring by nearby waterlogging or salinization,
interceptor drains are a second option. Then, the incoming flow per metre, q1, is half of
the losses from the canal. These losses can be estimated by measuring the fall in water
level in an isolated section.
However, these losses are
proportional to the difference in
FIGURE A21.3
head between the canal water and
Drain trench or ditch in low-permeable topsoil
the nearby groundwater. Therefore,
drainage will increase both head and
inflow (Figure A21.4). Lowering the
groundwater increases the flow with
a factor h2/h1.
h
The incoming inflow can be
b
calculated if the original loss and the
factor h2/h1 and q0 are determined:
(8)
where:
q0 = original outflow from
canal (m2/d);
q1 = outflow from canal after
interceptor drainage (m2/
d);
D = thickness of aquifer (m);
h1 = hydraulic head in the
canal (m) above original
groundwater level;
h2 = hydraulic head in the canal
(m) above drain level.
On the other hand, losses from
upstream irrigated or rainfed lands
will not be influenced by interceptor

Slightly permeable layer

K1

Highly permeable layer

K2

FIGURE A21.4

Leakage from a canal

Canal

h2
h1

Groundwater before
and after drainage

Drain

Groundwater before
D

Impermeable base

Drain

and after drainage

206

Guidelines and computer programs for the planning and design of land drainage systems

FIGURE A21.5

Cross-section of flow from higher lands with interceptor

drainage
Irrigated lands

Recharge and flux constant

Original groundwater
Interceptor

Groundwater after drainage

(same flux)

Impermeable base

drainage. This is because these losses

are a component of the upstream
water balance, as can be observed
from the cross-section shown in
Figure A21.5.
These types of losses can be
estimated from water balances or
by applying Darcys Law to the
resulting groundwater current.
Where the canal or field losses
are known, the programs INCEP
and INCEP2 can be used to find
the necessary trench width for the
interceptor drain.

RESULTS
In many cases, the width is such that a regular drainage is to be preferred, for which the
program ARTES gives some guidelines. Alternatively, a wide ditch can be considered,
especially at intermediate values for the required width. However, as side slopes tend to
become unstable under such circumstances, it is often necessary to stabilize them. This
can be achieved by covering the side slopes with a gravel cover or by making a wide,
gravel-filled trench provided with an outlet pipe.
REFERENCE
Donnan, W.W. 1959. Drainage of agricultural lands using interceptor lines. J. Irri. Drain. Div.
Proc. ASAE, 85, IR 1:1323.

207

Annex 22

Drainage by vertical wells

INTRODUCTION
Vertical drainage is possible under
favourable geological circumstances:
a good aquifer underneath;
an aquifer containing water
with a low salt content, so that
the water can be used;
not too large resistance between
soil and aquifer.
Figure A22.1 gives a sketch of the
method.
Two types of wells are considered:
those fully penetrating the aquifer;
and non-penetrating cavity wells.
They are supposed to form a large
array of squares (Figure A22.2)
or triangles (Figure A22.3). In
Figures A22.2 and A22.3, for one
well, the flow region and the sphere
of influence are indicated.
This method is mainly used in
arid regions where use of the water
for irrigation has often led to serious
overpumping. In some areas, the
lowering of the water levels in
the aquifer has led to attraction of
salty water from elsewhere, often
from deeper layers, sometimes
from the sea. In the long run, in
an arid climate, salt will inevitably
accumulate. However, this process
is usually very slow, owing to the
large amount of water stored in an
aquifer. Thus, vertical drainage may
be a temporary solution to a high
water table situation.
Nevertheless, the method can be
used to control groundwater levels.
This is illustrated by the following
(steady-state) theory.

FIGURE A22.1

Flow to a well in a semi-confined aquifer

5
Well

Surface

Groundwater
Head aquifer

-5

Resistance

-10

Aquifer with filter

Cavity well

Permeability K

Fully
penetrating
well

Impermeable base

Thickness D

-15

-20
0

10

15

20

FIGURE A22.2

Quadratic well net

FULLY PENETRATING WELLS

An area is drained by an array of evenly spaced deep wells tapping an aquifer
(Figure A22.1). This array may be quadratic or triangular and contains a large
number of wells that penetrate the entire aquifer. Each of them drains an equivalent

25

Guidelines and computer programs for the planning and design of land drainage systems

208

square (Figure A22.2) or a hexagon

(Figure A22.3), depending on the
Triangular well net
array pattern (quadratic or triangular, respectively). This outer
limit is approached by a circle of
equal area, with radius R, and the
flow is cylindrical towards the
well. The entire well-field is very
large and exchange of water with
the surroundings may be ignored.
Recharge is from the surface.
The aquifer is overlain by
a relatively thin layer of low
permeability, which separates it
from the shallow phreatic water. It
offers a certain resistance to flow
between groundwater and aquifer,
but does not prevent it entirely.
Thus, pumping lowers not only the
hydraulic head in the aquifer, but
also the shallow groundwater level.
The aquifer has a permeability K (metres per day) and a thickness D (metres), and,
thus, a transmissivity T = KD.
Between the aquifer and the groundwater is a semi-permeable layer of low vertical
permeability K and thickness d. This leads to a certain resistance c = d/K, which is
considered independent of the water levels. If K is in metres per day and d in metres,
c is in days.
Through this layer, the aquifer is recharged by rainfall or irrigation, with an intensity
q (metres per day).
A first estimate about the square spacing of wells is that it should be of the order of
a characteristic length of the aquifer system:
FIGURE A22.3

(1)
where:
resistance of semi-confining layer (d);
D = thickness of aquifer (m);
d = thickness of semi-confining layer (m);
K = permeability of aquifer (m/d);
K = permeability of semi-confining layer (m/d);
= characteristic length (m).
Greater insight is obtained from formulae describing the lowering of the
groundwater when an aquifer is pumped by a network of wells under the following
conditions (Figure A22.1):
the wells are fully penetrating and tap the aquifer over its entire depth;
between groundwater and aquifer, there is a layer of low permeability that gives a
certain resistance to vertical flow, but still allows its passage;
there is equilibrium between the amounts pumped and the recharge (steady
state);
no water is entering the well-field laterally from outside.

Annex 22 Drainage by vertical wells

209

The yield of each well Qw is taken to be positive, as is the flow Q towards the
well. According to Darcys Law and taking absolute values for Q, for the flow in the
aquifer:
Q = 2 rKD

dH

(2)

dr

On the other hand, the rainfall or irrigation excess should create the same flow:

Q = R2 r 2 q

(3)

so that both expressions for Q are equal, provided that there is no lateral inflow
from around the well-field.
Finally, the vertical resistance c of the layer between groundwater and aquifer leads
to a recharge:
q=

h(r ) H(r )
c

(4)

where, in these equations:

h = groundwater level (m);
H = head in aquifer (m);
q = recharge (m/d);
= flow towards well, absolute value (m3/d);
Qw = discharge of well, absolute value (m3/d);
r = distance from well centre (m);
rw = radius of well (m);
R = radius sphere of influence of well (m);
in which Q, h and H are functions of r.
At the watershed boundary with other wells, r = R and Q = 0. At this critical point,
h should have a prescribed maximum level. If h and H are expressed with respect to
soil surface, the groundwater should be at a certain depth (e.g. 2.0 m), so that h(R)
= -2.0. Then, with a given recharge q and resistance c, H(R) can be calculated from
Equation 4.
Then, it follows from the basic equations that:
or

(5)

Integration gives for the head H in the aquifer:

(6)
R is taken as the radius of a circle with the same area as the quadrangular or
triangular region served by one well.
Under these conditions, the following equation is valid for the groundwater
height h:
for

rw r R

(7)

Midway between the surrounding wells, the groundwater table should be lowered to
the required depth, but it will be deeper near the well. The head in the aquifer is lower
than the groundwater level because of the resistance between the two. If more water
is being pumped than the recharge, there will be overpumping, leading to a gradual
depletion of the aquifer. Although this is usually not sustainable, overpumping can be

210

Guidelines and computer programs for the planning and design of land drainage systems

a temporary solution for water scarcity (groundwater mining), high groundwater

tables, and soil salinization.
For a quadratic pattern (Figure A22.2) with well spacing distances L, the area A
served per well is:
or

(8)

For a triangular array (Figure A22.3), the region drained by a well is hexagonal,
where:
or

(9)

CAVITY WELLS
In some areas, wells are made by removing sand from the aquifer by heavy pumping.
A washed-out cavity is formed at the top of the aquifer, which remains intact during
the following period of less heavy abstraction (Figure A22.1, in blue). Compared with
fully penetrating wells, they encounter an extra resistance, but their diameter is larger,
although the actual size is rarely known.
The cavity is supposed to be a half-sphere with radius rw. In its vicinity, the flow is
spherical and an extra resistance occurs. This effect is estimated by assuming that the
flow to such non-penetrating wells breaks down as follows:
cylindrical flow from the outer limit R to a distance rd from the well, so that
Equation 6 can be used for r > rd; arbitrarily, rd can be taken as the lowest value of
D or R;
spherical flow from distance rd to the spherical cavity with radius rw.
For rd, arbitrarily:
(10)
where:
D = thickness of aquifer (m);
and D < R.
For very thick aquifers or a very dense network, D can become larger than R. Then,
for D > R:
(11)
The cylindrical part of the flow is described by Equation 6 for rd < r R.
The head in the aquifer is calculated (or approximated) by:
rw r rd

(12)

There are several assumptions involved, but the greatest uncertainty lies in the
unknown diameter (thus, radius rw) of the cavity. Although this is an approximation,
the errors are small enough for practical purposes.
APPLICABILITY OF THE METHOD
If more water is being pumped than the recharge, there will be overpumping, leading
to a depletion of the aquifer. Moreover, an equilibrium abstraction will also not be
sustainable in an arid region. This is because its use for irrigation will lead ultimately
to a harmful accumulation of salt in the aquifer. However, both overpumping and
equilibrium abstraction may be used as temporary solutions for water scarcity, high

Annex 22 Drainage by vertical wells

groundwater, and soil salinity. The time horizon depends on the local circumstances
and requires further study.
The program WELLS is based on these considerations. The differences between
fully penetrating and cavity wells relate to an extra radial resistance in the vicinity of
the latter (red and blue lines in Figure A22.1). This extra resistance is caused by flow to
a sphere instead of a long cylinder.

211

213

Annex 23

Computer programs for drainage

calculations
GENERAL CONSIDERATIONS
The programs first mention their name and purpose. Then, the following three
questions appear:
Notation of decimals
The use of the decimal separator in your country, point or comma, is requested.
Answer 1, 2 or 9. If a comma, a warning is given to ENTER all decimal data with a
point as separator. Using a comma would lead to serious errors. Answer the question
with 9 if you like to quit.
Project name
A project, or a section of it, must be indicated by a name of at most four characters,
which will form part of the output filename. The limited length allowed is because of
the limited size of filenames under DOS.
Certain rules must be followed:
The program asks for a project name, put between single quotation marks.
A maximum of four characters are allowed between those quotes, so that
abbreviations are often needed (e.g. proj for project). It is advisable to divide
large projects into sections and use section names (usually one or two characters)
as the project name. The single quotes indicate that the name is entered as a
character string, even if it is a pure number (23).
Project names with less than four characters are padded with minus signs in order
to obtain filenames of equal length. Thus, A2 automatically becomes A2--.
When the session is finished and the program closed, the data are saved in a file.
The filename has two characters indicating the kind of program, followed by this
project name and the extension TXT, for example, file SPA2--.TXT for program
drain spacings (SP) with project name entered as A2.
However, as new data become available, this existing file cannot be used again,
because this project name is already occupied. If tried, a warning is given that the
name is already in use and that a new name must be given. Thus, it is advisable
to end with a number, so that (for example) project A2 can be followed later by
A3, where both cover the same area A.
Location
After this short indication for the project (or part of it), the program asks for the
location within. Each project file can store observations from different locations, which
are indicated by a name of at most ten characters (letters and numbers).
Again, the name must be between single quotation marks. The location can be
a plot number (123, C14), a name (Johnson, Bahawalpur), or a combination
(7aq2n4).
If processed in the same session, the data for several locations within the same project
are combined into one file, which contains the name of the project (A2--- .TXT in the
earlier example). This project file contains all locations treated and is closed automatically
at the end of the session. As mentioned above, the name cannot be used again.

214

Guidelines and computer programs for the planning and design of land drainage systems

All project files obtained are listed in a file LIST**, beginning with LIST, followed
by two characters for its kind (LISTSP.TXT contains all drain spacing [SP] calculations
made).
Output files
For each project, the results are written to a file, the name of which is mentioned by
the program.
If reading in DOS, take care to copy this indication literally, including the signs -, --,
and --- used if some of the four positions are blanks (project A leads to file A---.TXT,
and project AB to file AB--.TXT).
Under Microsoft Windows, this difficulty is avoided. Just double-click the icon.
GUMBELS METHOD
GUMBEL, for estimating extreme values
After the three general questions (notation of decimal, project name, and location), the
program moves on to specifics:
Input of the extreme values (e.g. the highest three-day precipitation in a given
month, in millimetres) from keyboard or from data file. They are processed using
Kendalls method.
The return period (T) related to hydrological data (usually in years). The program
gives the expected values.
End the series of T with 999. A graph appears on screen with the data on the
vertical axis, and the Gumbel distribution on the horizontal, with the data plotted
according to Kendall. The Kendall line is shown in red. The graph is useful to
visually detect upward or downward trends, which make the prediction less valid
and indicate that the method may not be applicable in this case: too low if upward,
too high if downward.
Leave the graph with ENTER.
Continuation, output and example
The process can be repeated in a new case belonging to the same project. With another
project or END, the files are closed and the results written to file GU****.txt, where
GU stands for Gumbel and **** is the abbreviated project name. These filenames
are mentioned in LISTGU.TXT.
Figure A23.1 gives the output for extremes of total precipitation occurring during 1
to 7 successive days (1d to 7d) in an area in eastern Spain. The climate is Mediterranean,
with heavy rainfall in autumn.
PERMEABILITY MEASUREMENTS
AUGHOLE, for permeability from auger-hole measurements
After the three general questions (notation of decimal, project name, and location), the
program moves on to specifics:
Which unit is chosen? Answer 1, 2 or 3. Recommended is 2, the use of centimetres,
in contrast to most other programs.
Diameter and depth of the auger hole in the chosen units?
Location of the impermeable base?
Groundwater present of no? This determines the method: normal or inverse (less
reliable).
Normal method
For the normal method, the initial depth of the water in the hole is measured after
equilibrium. Then, some water is pumped out and the position of the water table is
given at different times:

Annex 23 Computer programs for drainage calculations

215

Equilibrium groundwater depth?

Water depth at time t1?
Water depth at time t2? (should be less).
Time interval t2 - t1 in seconds?

Inverse method
In dry soils, the groundwater may be
too deep to measure the permeability
of the upper layers. In this case, the
inverse method can be used. Water
is poured in, and its lowering is
measured over time. The method
is less reliable and should be used
only if there is no other possibility.
Moreover, some soils swell slowly
and have a lower permeability in the
wet season.
Option no groundwater is
followed, and the fall of the water
level and the time interval are
entered.
Continuation
The resulting permeability appears
on screen.
Next items:
Same or new auger hole or
END? The first option allows
another measurement in the
same auger hole, e.g. in the
subsequent interval. The other
two finish the calculation and
show the mean value and its
standard deviation on screen.
The next item can be in the
same project or not. In the
first case, the existing project
file is continued. Otherwise,
it is closed and the filename
mentioned on screen as
AU****.txt where AU denotes
auger hole and **** is the
abbreviated project name.
This name is also added to
the listing LISTAU.TXT,
mentioning all existing augerhole files.
If Other project or END
is selected, new names are
required for project and
location; END returns the
user to the initial screen.

FIGURE A23.1

Printout of program GUMBEL

****** Gumbel Distribution ******
=======================================================
project: Pego; location: P-1d; case: Pego01.txt
return period value
2.0 111.3565
5.0 188.6375
10.0 239.8043
20.0 288.8846
****** Gumbel Distribution ******
=======================================================
project: Pego; location: P-2d; case: Pego02.txt
return period value
2.0 136.6437
5.0 232.8728
10.0 296.5849
20.0 357.6990
****** Gumbel Distribution ******
=======================================================
project: Pego; location: P-3d; case: Pego03.txt
return period value
****** Gumbel Distribution ******
=======================================================
project: Pego; location: P-4d; case: Pego04.txt
return period value
2.0 162.4361
5.0 273.7000
10.0 347.3664
20.0 418.0289
****** Gumbel Distribution ******
=======================================================
project: Pego; location: P-5d; case: Pego05.txt
return period value
2.0 171.9708
5.0 283.2239
10.0 356.8832
20.0 427.5389
****** Gumbel Distribution ******
=======================================================
project: Pego; location: P-6d; case: Pego06.txt
return period value
2.0 177.9110
5.0 284.9797
10.0 355.8684
20.0 423.8667
****** Gumbel Distribution ******
=======================================================
project: Pego; location: P-7d; case: Pego07.txt
return period value
2.0 186.2486
5.0 291.2184
10.0 360.7175
20.0 427.3827

216

Guidelines and computer programs for the planning and design of land drainage systems

FIGURE A23.2

Printout of program AUGHOLE

****** Calculation of K from auger hole data ******
=====================================================
project: OFL1; location: Swifterb; case: OFL101.txt
diameter depth groundwater depth of position of
cm
cm depth cm base cm hole bottom
----------------------------------------------------8.0 150.0 50.0 200.0 above base
----------------------------------------------------number water level cm time
K stand.err.
meas.
1
2
s
m/d of mean
--- direct method ---------------------------------1 85.0 83.0 20.0
.63
2 80.0 78.0 24.0
.60
3 70.0 68.0 31.0
.67
-------------------mean .63 .02

Example
In the project OFL1, at location
Swifterb, an auger hole of 8 cm in
diameter and 150 cm deep is made.
The impermeable base is at a depth
of 200 cm. Groundwater establishes
a water level in the hole at a depth
of 50 cm. Several measurements are
taken after lowering to 90 cm below
the surface. This gives K = 0.63 m/d,
as shown by Figure A23.2.

PIEZOM, for permeability from

piezometer measurements
After the three general questions
(notation of decimal, project name,
and location), the program moves on
to specifics:
Which unit is chosen? Answer 1, 2, or 3. Recommended is 2, the use of centimetres,
in contrast to most other programs.
Diameters of protection pipe and cavity in the chosen units?
Length of protection pipe and cavity in the chosen units?
Location of the impermeable base?
Equilibrium groundwater depth below top of pipe?
Then, some water is pumped out and the position of the water table is given at
different times:
Water depth at time t1?
Water depth at time t2? (should be less).
Time interval t2 - t1 in seconds?
The inverse method is not included.

Continuation
The resulting permeability appears on screen.
Next items:
Same or new piezometer hole or END? The first option allows another measurement
in the same piezometer, e.g. in the subsequent interval. The other two finish the
calculation and show the mean value and its standard deviation on screen.
The next item can be in the same project or not. In the first case, the existing
project file is continued. Otherwise, it is closed, and the filename mentioned
on screen as PZ****.TXT where PZ indicates piezometer and **** is the
abbreviated project name.
This name is also added to the listing LISTPZ.TXT, mentioning all existing
piezometer files.
If Other project is selected, new names are required for project and location.
END returns the user to the initial screen.
Output
The output is similar to that of AUGHOLE. Figure A23.3 gives an example.
CALCULATION OF DRAIN SPACINGS
SPACING, for drainage under normal (non-artesian) conditions
After the three general questions (notation of decimal, project name, and location), the
program moves on to specifics:

Annex 23 Computer programs for drainage calculations

217

How is the size of drains

expressed, (as diameter, as

FIGURE A23.3
radius, as width of open
Printout of program PIEZOM
ditches)? ENTER 1, 2 or 3.
****** Calculation of K from piezometer data ******
The size itself, in metres?
=======================================================
project:d; location: da nang; case: d---01.txt
Divide centimetres by 100
========================================================
and always use a point for the
Piezometer
decimal.
diameter
length
groundw. position
pipe cavity pipe cavity depth
bottom
The design discharge, in metres
cm
cm
cm
cm
cm
cavity
per day. Divide millimetres per
----------------------------------------------------------day by 1 000.
8.0 5.0 200.0 25.0 40.0
above base
---------------------------------------------------------- The required groundwater
number water depth cm time
K stand.err.
depth at this recharge, in metres
meas.
1
2
s
m/d of mean
below surface.
----------------------------------------------------1
120.0
115.0
12.0
3.29
The depth of drains (pipes
2
115.0
110.0
13.0
3.25
or ditch bottoms), in metres
-------------------below surface.
mean 3.27 .02
These general data appear on
screen. If correct, ENTER 1; else 9
to restart the questions. Then:
The number of layers distinguished: the first above drain level, the remaining
strata below.
Their thickness. That of the first is known, being the drain depth; for the others,
it must be given.
Their anisotropy. As this will seldom be available, it is advisable to use 1 above
drain level, and below 4 if not clearly layered and 16 if so. This is a better guess
than neglecting anisotropy.
Their permeability, as measured by auger hole or piezometer or estimated from
profile characteristics.
The soil data are shown and, if
correct, the necessary calculations
FIGURE A23.4
are made.
Printout of program SPACING

Continuation
The project can be continued and
then the data for the new location
are added to the same file. If a new
project is taken or the existing one
is ended, the files are closed and the
filename is mentioned on screen and
added to LISTSP.TXT. Any new
project needs another name.
Output and example
The results are visible on screen
and put on file SP****.TXT, where
SP denotes spacing and ****
the abbreviated project name.
Figure A23.4 gives an example of
the output for project aa, location
amandabad. The radial resistance Wr
can be used as input in the programs
NSDEPTH and NSHEAD.

******Drain spacings, steady state******

Artesian influences not significant
=======================================================
project: aa; location: amandabad; case: aa--01.txt
************ GENERAL INPUT DATA for SPACING ************
effective diameter of drain
.08 m
design discharge of drain
.015 m/d
design groundwater depth midway
.30 m
design head above drain level
1.20 m
design drain depth
1.50 m
********************************************************
****************** Soil data ***********************
thickness layer 1, above drains
1.50 m
thickness layer 2, below drains
2.00 m
anisotropy factor layer 1
1.00 -anisotropy factor layer 2
4.00 -horiz. permeability layer 1
1.00 m/d
horiz. permeability layer 2
2.00 m/d
-------------------- Results ----------------------available head
1.20 m
radial resistance Wr
.97 d/m
flow above drains/total flow
.20 -drain spacing L-Hooghoudt
43.
m
********************************************************

218

Guidelines and computer programs for the planning and design of land drainage systems

NSABOVE, for drain spacing at non-steady flow above drain level only
After the three general questions (notation of decimal, project name, and location), the
program moves on to specifics:
Thickness of permeable layer (equal to drain depth or ditch bottom).
Pipe drains or ditches. For pipes and dry and almost dry ditches, the Boussinesq
approach is followed; for water-holding ditches, the Schilfgaarde method is used.
For pipe drains and nearly dry ditches, there is choice between an elliptic
initial situation, where the shallowest depth is midway between drains, or a total
ponding of the entire area.
In the elliptic case, the initial groundwater depth midway is asked (in ponding it
is zero everywhere). In the Schilfgaarde method, the shape is initially elliptic.
The required groundwater depth at time t and the value of t.
For water-holding ditches, the (constant) water depth must be specified.
If these data are correct, the soil characteristics are required: the permeability
and the available storage (moisture volume fraction between saturation and field
capacity).
Calculations are made and the resulting drain spacing appears on screen.
If initially ponded, a lag time is mentioned, an estimation of the time span
between total saturation and the first lowering midway between drains.

FIGURE A23.5

Printout of program NSABOVE

****** Non-steady flow above drain or ditch bottom ******
=======================================================
project: a; location: a1; case: a---01.txt
******************************* Drains **********
drain depth
1.40 m
depth impermeable base
1.40 m
Properties of permeable layer
permeability (horiz.=vert.) 2.00 m/d
storage coefficient
.12 ------------------------ Results -----------------groundw.depth
at t= .00 d
.00 m [everywhere]
groundw.depth midway at t= 1.00 d
.20 m
drain spacing . . . . . . . . .
L
19. m
estimated lag time
.41 d
********************************************************
****** Non-steady flow above drain or ditch bottom ******
=======================================================
project: a; location: a2; case: a---02.txt
**************** Ditches Schilfgaarde **********
ditch water depth below surface
.80 m
ditch bottom depth below surface 1.40 m
depth impermeable base
1.40 m
Properties of permeable layer
permeability (horiz.=vert.)
storage coefficient

2.00 m/d
.12 --

----------------------- Results -----------------groundw. depth midway at t= .00 d

.00 m [elliptic]
groundw. depth midway at t= 1.00 d
.20 m
ditch spacing . .
L-Schilfgaarde 22. m
estimated lag time
.00 d
********************************************************

Continuation
The process can be repeated in a new
case belonging to the same project.
With another project or END, the
files are closed and the results written
to file NA****.txt, where NA stands
for Nonsteady Above and ****
is the abbreviated project name.
These filenames are mentioned in
LISTNA.TXT.
Output and examples
Figure A23.5 gives results at two
locations in project a, of which
location a1 has pipe drains, location
a2 water-holding ditches. In the first
case, the surface is considered ponded
at the beginning; in the second case,
the water table is initially elliptic. The
difference in lag time to reach a
nearly elliptic shape explains most of
the difference in drain spacing.
NSDEPTH and NSHEAD, for
drains above impermeable base
NSDEPTH gives the depth of
the groundwater below surface,
NSHEAD gives the head above
drain level.
After the three general questions
(notation of decimal, project name,
and location), the programs move on
to specifics:

Annex 23 Computer programs for drainage calculations

The
The
The
The
The

permeability (equal above and below drain level), in metres per day.
storage coefficient, as volume fraction.
drain depth, in metres below surface.
thickness of the layer below the drains, in metres.
initial groundwater depth, the same everywhere: ponded or specified. If
ponded, it is automatically zero; if specified, the initial depth is required.
The radial resistance Wr near the drain (d/m). An estimate can be obtained from
the program SPACING. The entrance resistance, met by flow into the drain, is
ignored. For ditches, it is near zero; for good working drains, it is negligible, of
the order of 0.1 d/m.
For abnormally high discharges, the outflow system can be handled by the pipes and
ditches, but at higher heads and water levels. The following data allow an estimate:
The design discharge of the outflow system, in metres per day. Divide millimetres
per day by 1 000.
The design head loss in this system, in metres. At high discharges, higher head
losses are to be expected, leading to higher levels in this system.
After a heavy rain (or snowmelt), evaporation may help to lower the groundwater
tables, but the influence diminishes the deeper they are. The following items allow an
estimate:
The potential evaporation, in metres per day. Divide millimetres per day by
1 000.
The relationship of potential evaporation with groundwater depth, linear or
exponential.
The depth where evaporation becomes zero (linear) or the characteristic depth
where it is reduced to 1/e times the value at the surface (exponential).
Check the input. If correct, continue with:
Proposed drain spacing, in metres.
Number of days to be calculated.
Time-step for the calculation (lower than a given maximum), in days.
NSDEPTH shows the resulting groundwater depths on screen, with t is the time,
dp the groundwater level in the drainpipe, d0 the groundwater level near the drain and
d1d10 the depths between the drain and midway, where d0 is drain and d10 is midway.
Finally, d11 is equal to d9 (symmetry).
If unsatisfactory, other drain spacing can be taken. A slow retreat in dp values
suggests an insufficient main system or unsatisfactory performance of the drainpipe.
Large differences between dp and d0 indicate a considerable influence of the radial
resistance Wr.
NSHEAD is similar, but it gives the heads above drain level instead of the
depths.
Continuation
After ending with 999, the process can be repeated for a new case belonging to the
same project. With another project or END, the files are closed and the results written
to file ND****.txt or NH****.txt, where ND stands for Nonsteady Depth, NH
for Nonsteady Head and **** is the abbreviated project name. These filenames are
mentioned in LISTND.TXT and LISTNH.txt.
Output and examples
Figure A23.6 and A23.7 show examples from NSDEPTH and NSHEAD for project
aa, location aa1. The first shows the groundwater depths as function of time, the
second the heads above drain level. Together they form the drain depth of 1.50 m.
The initial depth of the water table was 0.2 m below surface, giving the initial head
as 1.30 m.

219

220

Guidelines and computer programs for the planning and design of land drainage systems

ARTES, for drainage under

artesian conditions
After the three general questions
FIGURE A23.6
(notation of decimal, project name,
Printout of program NSDEPTH
and location), the conditions are
****** Non-steady flow, groundwater depths ******
mentioned and the program moves
=======================================================
on to specifics:
project: aa; location: aa1; case: aa--01.txt
How is the size of drains expressed
****** GENERAL INPUT DATA for NSDEPTH ****
(as diameter, as radius, as width of
soil permeability
2.000 m/d
open ditches)?
storage coefficient
.150 -- The size itself, in metres? Divide
drain depth below surface
1.500 m
2.000 m
thickness soil below drain level
centimetres by 100 and always use
initial groundw. depth below surface
.200 m
a point for the decimal.
radial resistance Wr
.500 d/m
outflow system, design capacity
.0100 m/d
The design discharge, in metres
outflow system, design head
.500 m
per day. Divide millimetres per
max. evaporation
.0050 m/d
day by 1 000.
groundwater depth where E=.43E0
.500 m
****************************************************
The required groundwater depth
at this recharge, in metres.
**** Results of NSDEPTH, non-steady depth ****
The required groundwater depth
******* Depths below soil surface *******
if there is no recharge (important
t=time, dp=depth in drain, d0=outside drain
for salinization in times that there
d10=midway, d0-d11=proportional distances from drain
is no irrigation and no rainfall).
20.00 m
Drain spacing
L
This depth must be greater than
Radial resistance Wr
.50 d/m
the former.
t dp d0 d1 d2 d3 d4 d5 d6 d7 d8 d9 d10 d11
The depth of drains (pipes or
ditch bottoms), in metres below
.00 .39 .21 .20 .20 .20 .20 .20 .20 .20 .20 .20 .20 .20
.15 .47 .29 .26 .25 .23 .22 .22 .22 .21 .21 .22 .22 .22
surface.
.30 .50 .33 .30 .28 .26 .25 .24 .23 .23 .23 .22 .22 .22
These general data appear on
.45 .53 .36 .33 .31 .29 .28 .26 .25 .25 .24 .24 .24 .24
screen. If correct, ENTER 1, else 9
.60 .55 .38 .36 .33 .31 .30 .28 .27 .27 .26 .26 .26 .26
.75 .58 .41 .38 .36 .34 .32 .31 .29 .29 .28 .28 .28 .28
to restart the questions.
.90 .60 .43 .40 .38 .36 .34 .33 .32 .31 .30 .30 .30 .30
Then, data are required about
1.05 .62 .45 .42 .40 .38 .36 .35 .34 .33 .32 .32 .32 .32
soils and hydrology:
1.20 .64 .47 .44 .42 .40 .38 .37 .36 .35 .34 .34 .34 .34
1.35 .66 .49 .47 .44 .42 .40 .39 .38 .37 .36 .36 .36 .36
The thickness of the top layer
1.50 .67 .51 .49 .46 .44 .42 .41 .40 .39 .38 .38 .38 .38
of low permeability, above and
1.65 .69 .53 .51 .48 .46 .44 .43 .42 .41 .40 .40 .40 .40
1.80 .71 .55 .52 .50 .48 .46 .45 .44 .43 .42 .42 .42 .42
below drain level. Above, it is
1.95 .73 .57 .54 .52 .50 .48 .47 .46 .45 .44 .44 .43 .44
already given by the drain depth
2.10 .75 .59 .56 .54 .52 .50 .49 .48 .47 .46 .46 .45 .46
and mentioned as such. Below,
2.25 .77 .61 .58 .56 .54 .52 .51 .49 .48 .48 .47 .47 .47
2.40 .78 .63 .60 .58 .56 .54 .52 .51 .50 .50 .49 .49 .49
it must be entered or estimated.
2.55 .80 .65 .62 .60 .58 .56 .54 .53 .52 .52 .51 .51 .51
However, where the thickness
2.70 .82 .66 .64 .61 .59 .58 .56 .55 .54 .53 .53 .53 .53
below is only a few decimetres,
2.85 .83 .68 .66 .63 .61 .59 .58 .57 .56 .55 .55 .55 .55
3.00 .85 .70 .67 .65 .63 .61 .60 .59 .58 .57 .57 .56 .57
it is better to put the drains
somewhat deeper, so that they tap
the underlying aquifer. This avoids
many problems with seepage.
The anisotropy above and below drain level. Often this is unknown. If not
visually layered, put 1 above and 4 below, else 16 below.
The horizontal permeability above and below, in metres per day, as can be
measured by auger hole or piezometer.
The resistance between top layer and aquifer, in days. This is thickness divided
by permeability of the layer between top layer and aquifer. A minimum is 25
50 days, a thin layer of tight clay has already 1 0005 000 days. If unknown and
no clay or compressed peat interferes, input 200 or try several values to see the
effect.

Annex 23 Computer programs for drainage calculations

221

The hydraulic head in the

aquifer in metres, above drain

depth in cases that upward
seepage occurs. For negative
seepage (natural drainage),
input negative values.
These data appear on screen.
ENTER 1 if correct, 9 otherwise.
If correct, the necessary calculations
are made.
Continuation
The project can be continued and
then the data for the new location
are added to the same file. If a new
project is taken or the existing one
is ended, the files are closed and the
filename is mentioned on screen and
added to LISTAR.TXT. Any new
project needs another name.
Output and example
The results are visible on screen and
put on file AR****.TXT, where
AR denotes artesian and ****
the abbreviated project name. The
smallest drain spacing is critical and
should be taken. The filename is
mentioned on screen and added to
LISTAR.TXT.
As an example, Figure A23.8
describes a seepage area under
irrigation in project a, location
adana. If irrigated, downward
water movement causes removal of
salts, but if no irrigation is given
the situation is critical, because of
upward movements. Therefore, the
drain spacing should not exceed
17 m, the smallest spacing given.

FIGURE A23.7

Printout of program NSHEAD

****** Non-steady flow above drain or ditch bottom ******
=======================================================
project: aa; location: aa1; case: aa--01.txt
****** GENERAL INPUT DATA for NSHEAD ****
******* all heads above drain level *******
soil permeability
2.000 m/d
storage coefficient
.150 --thickness of soil below drain level
2.000 m
initial groundwater head
1.300 m
radial resistance Wr
.500 d/m
outflow system, design capacity
.0100 m/d
outflow system, design head
.500 m
max. evaporation
.0050 m/d
groundwater depth where E=.43E0
.500 m
****************************************************
**** Results of NSHEAD, non-steady flow ****
******* Heads above drain level *******
t=time, hp=head in drain, h0=outside drain
h10=midway, h0-h11=proportional distances from drain
Drain spacing L
20.00 m
Rad. resistance Wr
.50 d/m
t

hp

h0

h1

h2

h3

h4

h5

h6

h7

h8

h9

.00 1.11 1.29 1.30 1.30 1.30 1.30 1.30 1.30 1.30 1.30 1.30
.15 1.03 1.21 1.24 1.25 1.27 1.28 1.28 1.28 1.29 1.29 1.29
.30 1.00 1.17 1.20 1.22 1.24 1.25 1.26 1.27 1.27 1.27 1.28
.45 .97 1.14 1.17 1.19 1.21 1.22 1.24 1.25 1.25 1.26 1.26
.60 .95 1.12 1.14 1.17 1.19 1.20 1.22 1.23 1.23 1.24 1.24
.75 .92 1.09 1.12 1.14 1.16 1.18 1.19 1.21 1.21 1.22 1.22
.90 .90 1.07 1.10 1.12 1.14 1.16 1.17 1.18 1.19 1.20 1.20
1.05 .88 1.05 1.08 1.10 1.12 1.14 1.15 1.16 1.17 1.18 1.18
1.20 .86 1.03 1.06 1.08 1.10 1.12 1.13 1.14 1.15 1.16 1.16
1.35 .84 1.01 1.03 1.06 1.08 1.10 1.11 1.12 1.13 1.14 1.14
1.50 .83 .99 1.01 1.04 1.06 1.08 1.09 1.10 1.11 1.12 1.12
1.65 .81 .97 .99 1.02 1.04 1.06 1.07 1.08 1.09 1.10 1.10
1.80 .79 .95 .98 1.00 1.02 1.04 1.05 1.06 1.07 1.08 1.08
1.95 .77 .93 .96 .98 1.00 1.02 1.03 1.04 1.05 1.06 1.06
2.10 .75 .91 .94 .96 .98 1.00 1.01 1.02 1.03 1.04 1.04
2.25 .73 .89 .92 .94 .96 .98 .99 1.01 1.02 1.02 1.03
2.40 .72 .87 .90 .92 .94 .96 .98 .99 1.00 1.00 1.01
2.55 .70 .85 .88 .90 .92 .94 .96 .97 .98 .98 .99
2.70 .68 .84 .86 .89 .91 .92 .94 .95 .96 .97 .97
2.85 .67 .82 .84 .87 .89 .91 .92 .93 .94 .95 .95
3.00 .65 .80 .83 .85 .87 .89 .90 .91 .92 .93 .93

h10 h11
1.30
1.29
1.28
1.26
1.24
1.22
1.20
1.18
1.16
1.14
1.12
1.10
1.08
1.07
1.05
1.03
1.01
.99
.97
.95
.94

WELLS, for vertical drainage

Vertical drainage requires special conditions and is seldom a durable solution as it usually
leads to overpumping and mobilization of salts from elsewhere. However, if required, a
first estimate for well spacings can be obtained, based on steady-state equilibrium.
The program starts with the three general questions (notation of decimal, project
name, and location) and then moves on to specifics:
The minimum groundwater depth at the points furthest from the wells.
The type of well, fully penetrating the aquifer or cavity well.
The spacing of wells, in metres.
Their diameter, in metres.
The permeability of the aquifer and its thickness, in metres per day and in metres,
respectively.

1.30
1.29
1.28
1.26
1.24
1.22
1.20
1.18
1.16
1.14
1.12
1.10
1.08
1.06
1.04
1.03
1.01
.99
.97
.95
.93

222

Guidelines and computer programs for the planning and design of land drainage systems

The recharge (by rain or irrigation

FIGURE A23.8

Printout of program ARTES

****** Drainage under artesian conditions ******
=======================================================
project: a; location: adana; case: a---01.txt
************* GENERAL INPUT DATA for ARTES ***********
effective diameter of drain
.10 m
design recharge R (by rain or irrig.)
.005 m/d
design grw. depth midway at R
1.40 m
design grw. depth midway at R=0
1.80 m
design drain depth
2.40 m
design entrance resist. into drain
.00 d
********************************************************
************* Data for case a---01.txt *******************
Properties of top layer
thickness above drain level
2.40 m
thickness below drain level
5.00 m
anisotropy above drain level
1.00 -anisotropy below drain level
4.00 -hor.perm. above drain level
.20 m/d
hor.perm. below drain level
.40 m/d
Hydrology
resistance of aquitard
200.00 d
hydraulic head in aquifer
2.00 m
recharge (by rain or irrig.) R=
.005 m/d
--------------- Results of case a---01.txt -----------------recharge (by rain or irrig.) R =
.0050 m/d
seepage (neg. if downward)
.0048 m/d
spec. discharge above drain level .0023 m/d
spec. discharge below drain level .0075 m/d
head midway, at drain level
.98 m
groundwater depth midway
1.39 m
drain spacing . . . .L-Brug. =
19.
m
---------------------------Values for recharge R=0
recharge (by rain or irrig.)
.0000 m/d
seepage (neg. if downward)
.0061 m/d
spec. discharge above drain level .0010 m/d
spec. discharge below drain level .0051 m/d
head midway, at drain level
.60 m
groundwater depth midway
1.80 m
drain spacing . . .
L-Brug. =
17.
m
* * * Take SMALLEST value for spacing L * * *
********************************************************

losses), in metres per day. Divide

millimetres per day by 1 000.
The resistance of the overlying
layer, either directly (in days)
or from its permeability and
thickness.
The shape of the network (quadratic
or triangular arrangement of
wells).
The input is shown. If correct,
the heads far from and near the well
are given on screen. These heads are
expressed with respect to the soil
surface, because there is no drain
level in this case.
Continuing gives a table with
expected aquifer heads at various
distances, again with respect to the
soil surface.
Continuation
The project can be continued and
then the data for the new location
are added to the same file. If a new
project is taken or the existing one
is ended, the files are closed and the
filename is mentioned on screen.
Output and example
The results are visible on screen and
put on file WN****.TXT, where
WN denotes well network and ****
the abbreviated project name. The
filename is mentioned on screen and
added to LISTWN.TXT. Any new
project needs a different name.
An example is given in
Figure A23.9.

DRAIN DIAMETERS
DRSINGLE, for single drain
After the three general questions (notation of decimal, project name, and location), the
program moves on to specifics:
Type of drains: options are available for laterals and collectors. The latter are
characterized by greater spacing, and often also greater length.
Type of pipe: smooth (theoretical) (1); technically smooth (in practice) (2); or
corrugated (two options, general (3) or according to Zuidema for small pipes,
[maximum diameter 0.12 m]). Option general (3) will ask for the spacing of
corrugations.
Maintenance status, that is the amount of sediment to be expected in this soil
under usual maintenance. In some soils, drains will keep clean, even without or
with infrequent maintenance; in others, the pipes will clog with iron hydroxides,

Annex 23 Computer programs for drainage calculations

sediments, or roots, even with

regular (e.g. annual) cleaning.
The first will have a good
status, the second a poor one.
The quantity must be estimated
from earlier experience. Where
unknown, try 3.
Required
items:
length,
diameter, maximum spacing
allowed, head loss in drain, all
in metres and maximum specific
discharge (discharge divided by
area served) in metres per day.
According to this choice, all
other quantities except the
unknown will be required. The
result is shown on screen and
all data are written to file.
ENTER to continue. The
program calculates the results
and asks for a new item or to
end.
Same project, other one, or
end? The first option allows
another measurement in the
same project. The others finish
the calculation.
Continuation
In the same project case, the
existing project file is continued.
Otherwise, it is closed, and the
filename is mentioned on screen as
DS****.TXT, where DS denotes
Drain, Single and **** is the given
project name. All these names are
collected in the file LISTDS.TXT.
If Other project is selected,
new names are required for project
and location. With END, the user
returns to the initial screen.
Output and example
Figure A23.10 is an example for a
collector of 1 000 m in length in an
arid area.
DRMULTI, for multiple drain
The different materials of a multiple
drain, consisting of sections with
different diameters or materials
(cement, smooth or corrugated
plastic) must be specified, together

223

FIGURE A23.9

Printout of program WELLS

=======================================================
Drainage by array of wells, steady state
project: b; location: babel; case: b---01.txt
Fully penetrating well
Requirement on groundwater depth
min. depth
2.00 m
Well
diameter
.20
m
Aquifer
permeability
10.00 m/d
40.
m
thickness
recharge
.0030 m/d [3.0 mm/d]
System
aquifer transmissivity 400. m2/d
overlying resistance 200. d
characteristic length 283. m
Network
quadratic, spacing
200.
m
influence radius
113.
m
120.
m3/d (equilibrium)
discharge per well
head aquifer, limit
-2.60 m
head aquifer, well
-2.91 m
radius m head m [ surface=0. ]
groundwater aquifer
.10
-2.31
-2.91
.11
-2.31
-2.91
.28
-2.26
-2.86
1.01
-2.20
-2.80
2.99
-2.15
-2.75
7.15
-2.11
-2.71
14.71
-2.07
-2.67
27.17
-2.05
-2.65
46.28
-2.02
-2.62
74.07
-2.01
-2.61
112.84
-2.00
-2.60
---------------------------------------------------

FIGURE A23.10

Printout of program DRSINGLE

****** Dimensions of single drain ******

======================================================
project: abba; location: Saltabad; case: abba01.txt
----------------------------------------------------Drain pipe design: Single diameter
----------------------------------------------------Collectors
Technically smooth pipe, a-Blasius=0.40
Maintenance status: good
Input data
Drain length
1000.00 m
Collector spacing
300.00 m
Design head loss
.30 m
Design spec. disch.
.0030 m/d
Results
Min. inner diameter
.200 m
-----------------------------------------------------

224

Guidelines and computer programs for the planning and design of land drainage systems

with the available diameters, total length and spacing. The program then calculates the
length of the different sections.
After the three general questions (notation of decimal, project name, and location),
the program moves on to specifics:
Type of drains (laterals, collectors, or interceptor drains).
For laterals and collectors, data are asked for allowed head loss in drain and
specific discharge; for interceptors allowed head loss and inflow per m length
(obtained from INCEP or INCEP2).
The number of different sections is required.
Type of pipe used in each section: smooth (theoretical) (1), technically smooth (in
practice) (2), or corrugated (two options, general (3) or according to Zuidema for
small pipes [maximum diameter 0.12 m]). Option general (3) will ask for the
spacing of corrugations.
Maintenance status for the entire drain. This is the amount of sediment to be
expected in this soil under usual maintenance. In some soils, drains will keep
clean, even without or with infrequent maintenance; in others, the pipes will clog
with iron hydroxides, sediments, or roots, even with regular (e.g. annual) cleaning.
The former will have a good status, the latter a poor one. The quantity must be
estimated from earlier experience. Where unknown, try 3.
Diameter of each section.
For laterals and collectors: spacing and length; for interceptors: their length only.
Results
The necessary calculations are made and the result appears on screen, first for two
sections only. Then:
ENTER to see a graph showing the head at design discharge and the slope of the
drain.
ENTER again to leave the graph.
If more than two sections are being considered, this procedure is repeated for all
sections involved: lengths of all sections on screen, followed by a graph. Then:
ENTER to continue.
Same project, other one, or end? The first option allows another measurement in
the same project. The others finish the calculation.
Continuation
In the same project case the existing project file is continued. Otherwise, it is closed,
and the filename is mentioned on screen as DM****.TXT, where DM denotes Drain,
Multiple and **** is the given project name. All these names are collected in the file
LISTDM.TXT.
If Other project is selected, new names are required for project and location. With
END, the user returns to the initial screen.
Output and example
Figure A23.11 gives an example for laterals of 350 m in length in a humid climate.
MAIN DRAINAGE SYSTEM
BACKWAT, for backwater effects in the outlet channel of the main system
If an open channel of the main drainage system discharges via an open connection
or sluice into a river, lake or sea, fluctuations in outside water level will influence
the level in that channel. Especially high outside levels have an unfavourable and
sometimes disastrous effect. Apart from a steady-state influence, also non-steady
effects can be important in such cases. However, to form an idea of such effects,

Annex 23 Computer programs for drainage calculations

a steady-state approach is useful

in cases where storage of water
inland is not too important and the
fluctuations are relatively slow.
For such situations, the program
BACKWAT gives a solution.
Thus, travelling waves cannot be
calculated. Therefore, application is
limited to downstream sections and
sections above weirs that are of not
too great length and that receive a
constant flow from upstream.
Both high and low outside levels
are covered, and data about positive
of negative backwater curves are
given.

225

FIGURE A23.11

Printout of program DRMULTI

****** Dimensions of multiple drain ******
=======================================================
project: ba4; location: Balsa34; case: ba4-01.txt
================================================================
Drain pipe design
--------------------------------------------------------------------Number of sections: 2
Pipe type for lateral
section 1: corrugated, Zuidema (a-Blasius= .77), diameter .05 m
section 2: corrugated, Zuidema (a-Blasius= .77), diameter .08 m
section 3: corrugated, Zuidema (a-Blasius= .77), diameter .12 m
maintenance status: good
Input data
design head loss
.20 m
discharge intensity
.010 m/d
spacing of laterals
50.0 m
length of laterals
350.0 m
Output data
length of section 1:
.00 head loss .0000
length of section 2: 163.64 head loss .0935
length of section 3: 186.36 head loss .0966
length of drain : 350.00 real loss .1901 allowed .2000
--------------------------------------------------------------------Number of sections: 3
Pipe type for lateral
section 1: corrugated, Zuidema (a-Blasius= .77), diameter .05 m
section 2: corrugated, Zuidema (a-Blasius= .77), diameter .08 m
section 3: corrugated, Zuidema (a-Blasius= .77), diameter .12 m
maintenance status: good
Input data
design head loss
.20 m
discharge intensity
.010 m/d
spacing of laterals
50.0 m
length of laterals
350.0 m
Output data
length of section 1: 45.69 head loss .0261
length of section 2: 86.95 head loss .0497
length of section 3: 217.36 head loss .1026
length of drain : 350.00 real loss .1784 allowed .2000
================================================================

Program
After the three general questions
(notation of decimal, project name,
and location), the program moves on
to specifics:
Dimensions of watercourse:
bottom width in metres, side
slopes. The results are shown
on screen and can be corrected
if necessary.
Longitudinal profile: length
of section, land and bottom
elevation, first upstream and
then downstream, in metres.
Water elevation downstream, in
metres. The results are shown
on screen and can be corrected
if necessary.
Discharge from upstream, in cubic metres per second. Correction is possible. The
program gives the equilibrium depth far upstream. As a check, the discharge is
recalculated.
The step size in water depth, in metres, to be used in the numerical calculations.
The program shows the results. ENTER returns to step size so that another value
may be tried. Indicating END at this stage (type 9) leads to a question about the next
item.
Next item and example
Same project, other one, or end? The first option allows another measurement in the
same project. The others finish the calculation and ask for a new project filename for
another abbreviated filename.
In the same project case, the existing project file is continued. Otherwise, it
is closed, the filename mentioned on screen and added to LISTBW.TXT. If Other
project is selected, new names are required for project and location. With END, the
user returns to the initial screen.

Guidelines and computer programs for the planning and design of land drainage systems

226

FIGURE A23.12

Printout of program BACKWAT

****** Backwater curves ******
=======================================================
project: aa ;location: adana; case: aa--01.txt
Backwater curves
Watercourse
bottom width
5.00 m
side slopes 1: 2.00
(1 vertical: 2.00 horizontal)
Elevations
length of section

2000. m
6.00 m
3.00 m

land upstream
land downstream

bottom upstream
4.00 m
bottom downstream .00 m
water downstream

2.00 m

land slope
bottom slope

1.500 o/oo
2.000 o/oo

Discharge from upstream = 10.000 m3/s

Equilibrium depth upstream
Calc. discharge Q
distance
0.
28.
56.
85.
115.
145.
176.
208.
241.
275.
312.
351.
394.
442.
499.
571.
679.
1073.

depth
2.000
1.950
1.900
1.850
1.800
1.750
1.700
1.650
1.600
1.550
1.500
1.450
1.400
1.350
1.300
1.250
1.200
1.150

1.144 m
9.998 m3/s

water & land level

2.000
3.000
2.006
3.042
2.013
3.085
2.021
3.128
2.030
3.172
2.040
3.217
2.051
3.264
2.065
3.311
2.081
3.361
3.413
2.100
2.124
3.468
2.152
3.526
2.188
3.591
2.234
3.663
2.297
3.748
2.391
3.856
2.558
4.019
3.296
4.610

Q-calc
10.000
10.000
10.000
10.000
10.000
10.000
10.000
10.000
10.000
10.000
10.000
10.000
10.000
10.000
10.000
10.000
10.000
10.000

An example
Figure A23.12.

is

given

by

INTERCEPTOR DRAINS
INCEP and INCEP2, for
homogeneous profiles and for a
less permeable top layer
Interception drains are needed in
places where waterlogging occurs
in undulating terrain, especially to
protect the downstream fields. This
waterlogging is usually caused by a
decrease in slope, a change in the soil
profile or an abrupt lowering of the
surface. In other cases, it is caused
by leakage from irrigation canals and
watercourses, or from higher lands.
The program allows changes of this
kind for a profile of permeable soil
on an impermeable base. It calculates
the width of a drain trench or ditch
bottom that is sufficient to catch
the intercepted flow. A separate
calculation is needed for the size of
the drain needed, this can be found
by the program DRMULTI.

INCEP, homogeneous profile

After the three general questions
(notation of decimal, project
name, and location), the programs
moves on to specifics regarding the
upstream conditions:
The source: hillslope, canal or
higher fields.
In the case of hillslopes:
The upstream slope, as the ratio 1:
n (vertical: horizontal) of which n
is required.
The upstream permeability, in
metres per day.
The upstream depth of the impermeable base, in metres below surface.
The upstream depth of the groundwater, in metres below surface.
Depth of drain, below the upstream soil surface, in metres.
In the case of a leaky canal at higher level:
The water losses from the canal, flowing to both sides in the present situation in
square metres per day.
The water level in the canal above the nearby soil surface.
The original groundwater level below surface.
The required future groundwater level below surface.
In the case of flow from higher ground:
The flow from higher lands.
The required future groundwater level below surface.

Annex 23 Computer programs for drainage calculations

These data appear on screen.

If correct, ENTER 1, else 2 to
restart the questions. If correct, the
downstream conditions must be
specified:
Flat or sloping surface?
If there is a further downward
slope 1:n, the downstream n is
required, which must be more
than upstream.
The downstream permeability,
in metres per day.
The downstream depth of the
impermeable base, in metres
below surface.
Depth of drain, below the
downstream soil surface, in
metres. For hill slopes, the
difference with the upstream
value determines the difference
in surface elevation near the
drain.
The required downstream
depth of the groundwater, in
metres below surface.
These data appear on screen.
If correct, ENTER 1, else 2 to
restart the questions. If correct, the
necessary calculations are performed
and the results shown on screen,
the main one being the width of the
drain trench or ditch bottom needed
to catch the intercepted flow. In
most cases, a normal trench width is
sufficient, the main exception being
permeable soils of considerable
depth.
Calculating the lowering of the
groundwater upstream of the drain
is an option for hill slopes.
INCEP2, less permeable topsoil
The program treats a two-layered
soil with an upper layer at least
ten times less permeable that the
second one. Only a change in slope
is considered.
After the three general questions
(notation of decimal, project name,
and location), the program moves
on to specifics. These are similar to
those for INCEP, plus:

227

FIGURE A23.13

Printout of program INCEP

***** interceptor drain, homogeneous soil *****

=================================================
project: a; location: a1; case: a---01.txt
Upstream values
tangent of slope
.05 m/m 1:20.0
diff. surface level at x=0
.00 m
permeability
3.00 m/d
depth to impermeable layer
8.00 m
depth of drain, upstream end 2.00 m
drain above impermeable base 6.00 m
radial resistance near drain
.48 d/m
incoming flow
1.05 m2/d
thickness of incoming flow
7.00 m
depth groundwater upstream
1.00 m
Downstream values
zero slope, flat terrain
diff. surface level at x=0
permeability
depth to impermeable layer
depth of drain, downstream
drain above impermeable base
radial resistance near drain
head from radial resistance
incoming flow
intercepted flow
downstream flow
thickness of outgoing flow
depth groundwater downstream

.00 m
3.00 m/d
8.00 m
2.00 m
6.00 m
.48 d/m
.50 m
1.05 m2/d
1.05 m2/d
.00 m2/d
6.50 m
1.50 m

Required width of trench needed for groundwater control

width 0.10 m sufficient
WARNING: May not be sufficient for drain discharge!
Use DRMULTI for drain sizes.
Inflow into drain is 1.050 m2/d
==============================================
Upstream lowering by drain
100%= .50 m
lowering % lowering m distance x, m
---------------------------------------------------------100.
.50
.0
90.
.45
13.8
80.
.40
29.2
70.
.35
46.9
60.
.30
67.5
50.
.25
92.0
40.
.20
122.3
30.
.15
161.6
217.3
20.
.10
10.
.05
313.4
==============================================

228

Guidelines and computer programs for the planning and design of land drainage systems

Permeability of top layer, metres

per day.
Permeability of second layer,
Printout of program INCEP2
metres per day.
Thickness of top layer, metres.
***** interceptor drain, two-layered soil *****
================================================= Thickness of second layer, metres.
project: b; location: b1; case: b---01.txt
All entry data appear on screen. If
tangent of slope upstream
.05 m/m 1: 20.0
correct, ENTER 1, else 2 to restart the
downstream slope zero,
flat terrain
questions. If correct, the necessary
calculations are performed and the
no difference in surface level at x=0
permeability top layer
.30 m/d
results shown on screen, the main
permeability second layer
3.00 m/d
one being the width of the trench
thickness top layer
4.00 m
or ditch bottom needed to catch the
thickness second layer
4.00 m
intercepted flow. In contrast to the
depth to impermeable layer
8.00 m
homogeneous case, where a small
width is usually sufficient, a drain
depth of trench or ditch
2.00 m
drain above soil transition
2.00 m
in less permeable topsoil requires
radial resistance near drain
.78 d/m
a much wider trench. As this is
resulting head above drain
.50 m
often not feasible, several drains are
needed. Their mutual distance can be
incoming groundwater flow
.65 m2/d
estimated for the program ARTES
outgoing groundwater flow
.00 m2/d
intercepted by drain
.65 m2/d
for artesian conditions, their number
depth groundwater upstream
1.00 m
from the total flow to be eliminated.
FIGURE A23.14

depth groundwater downstream 1.50 m

7.00 m
thickness of incoming flow
thickness of outgoing flow
6.50 m

Continuation, output and examples

The process can be repeated in a new
case belonging to the same project.
Result: required bottom width
6.83 m
corrected linear approximation 6.84 m
With another project or END,
the files are closed and the results
Use DRMULTI for drain sizes.
written to file ID****.txt, where ID
Inflow into drain is .645 m2/d
stands for Interceptor Drain and
=============================================
**** is the abbreviated project name.
These filenames are mentioned in
LISTID.TXT.
Figure A23.13 gives the output of INCEP for a hillslope in project a, at location
a1. It can be seen that the effect of the radial resistance is negligible in this case, as is
usual for homogeneous permeable soils of rather shallow depth.
Figure A23.14 gives results for a case similar to Figure A23.13, but now with the
upper 4 m of low permeability and for a leaky canal. The increase in necessary bottom
width is dramatic. Although the flow is similar, the required width changes from less
than 0.10 m to more than 6 m. As this is impractical, several drains will be needed.
The hydrological conditions are usually more complicated at such locations and
often poorly known. Therefore, the programs can give rough guidelines only, and
solutions must often be found in the field by trial and error, adding more drains if
needed until the result is satisfactory.
The inflow per m drain can be used as input in the program DRMULTI to find the
necessary dimensions of the drain itself.

FAO TECHNICAL PAPERS

FAO IRRIGATION AND DRAINAGE PAPERS
1

Irrigation practice and water management,

1972 (Ar* E* F* S*)
1 Rev.1 Irrigation practice and water management,
1984 (E)
2
Irrigation canal lining, 1971
(New edition, 1977, available in E, F and S
in the FAO Land and Water Development
Series, No. 1)
3
Design criteria for basin irrigation systems,
1971 (E*)
4
Village irrigation programmes a new
approach in water economy, 1971 (E* F*)
5
Automated irrigation, 1971 (E* F* S*)
6
Drainage of heavy soils, 1971 (E* F* S*)
7
Salinity seminar, Baghdad, 1971 (E* F*)
8
Water and the environment, 1971 (E* F* S*)
9
Drainage materials, 1972 (E* F* S*)
10
Integrated farm water management, 1971
(E* F* S*)
11
Planning methodology seminar, Bucharest,
1972 (E* F*)
12
Farm water management seminar, Manila,
1972 (E*)
13
Water use seminar, Damascus, 1972 (E* F*)
14
Trickle irrigation, 1973 (E* F* S*)
15
Drainage machinery, 1973 (E* F*)
16
Drainage of salty soils, 1973 (C* E* F* S*)
17
Mans inuence on the hydrological cycle,
1973 (E* F* S*)
18
Groundwater seminar, Granada, 1973
(E* F S*)
19
Mathematical models in hydrology, 1973
(E*)
20/1
Water laws in Moslem countries Vol. 1,
1973 (E* F*)
20/2
Water laws in Moslem countries Vol. 2,
1978 (E F)
21
Groundwater models, 1973 (E*)
22
Water for agriculture index, 1973 (E/F/S*)
23
Simulation methods in water development,
1974 (E* F* S*)
24
Crop water requirements, (rev.) 1977
(C* E F S)
25
Effective rainfall, 1974 (C* E* F* S*)
26/1
Small hydraulic structures Vol. 1, 1975
(E* F* S*)
26/2
Small hydraulic structures Vol. 2, 1975
(E* F* S*)
27
Agro-meteorological eld stations, 1976
(E* F* S*)
28
Drainage testing, 1976 (E* F* S*)
29
Water quality for agriculture, 1976
(E* F* S*)
29 Rev.1 Water quality for agriculture, 1985
(C** E* F* S*)
30
Self-help wells, 1977 (E*)
31
Groundwater pollution, 1979 (C* E* S)
32
Deterministic models in hydrology, 1979
(E*)
33
Yield response to water, 1979 (C* E F S)
34
Corrosion and encrustation in water wells,
1980 (E*)

35

Mechanized sprinkler irrigation, 1982

(C E* F S*)
36
Localized irrigation, 1980 (Ar C E* F S*)
37
Arid zone hydrology, 1981 (C E*)
38
Drainage design factors, 1980 (Ar C E F S)
39
Lysimeters, 1982 (C E* F* S*)
40
Organization, operation and maintenance
of irrigation schemes, 1982 (C E* F S*)
41
Environmental management for vector
control in rice elds, 1984 (E* F* S*)
42
Consultation on irrigation in Africa, 1987
(E F)
43
Water lifting devices, 1986 (E F)
44
Design and optimization of irrigation
distribution networks, 1988 (E F)
45
Guidelines for designing and evaluating
surface irrigation systems, 1989 (E*)
46
CROPWAT a computer program for
irrigation planning and management, 1992
(E F* S*)
47
Wastewater treatment and use in
agriculture, 1992 (E*)
48
The use of saline waters for crop
production, 1993 (E*)
49
CLIMWAT for CROPWAT, 1993 (E)
50
Le pompage olien, 1994 (F)
51
Prospects for the drainage of clay soils,
1995 (E)
52
Reforming water resources policy, 1995 (E)
53
Environmental impact assessment of
irrigation and drainage projects, 1995 (E)
54
Crues et apports, 1996 (F)
55
Control of water pollution from agriculture,
1996 (E* S)
56
Crop evapotranspiration, 2006 (E S)
57
Soil salinity assessment, 1999 (E)
58
Transfer of irrigation management services:
guidelines, 1999 (E F R S)
59
Performance analysis of on-demand
pressurized irrigation systems, 2000 (E)
60 Rev.1 Materials for subsurface land drainage
systems, 2005 (E)
61
Agricultural drainage water management
in arid and semi-arid areas, 2002 (E)
62
Guidelines and computer programs for
the planning and design of land drainage
systems, 2007 (E)
Availability: March 2007
Ar
Arabic
Multil Multilingual
C

Chinese
* Out of print
E

English
** In preparation
F

French
P

Portuguese
R

Russian
S

Spanish
The FAO Technical Papers are available through the
authorized FAO Sales Agents or directly from Sales and
Marketing Group, FAO, Viale delle Terme di Caracalla,
00153 Rome, Italy.

ISBN 978-92-5-105670-7

ISSN 0254-5284

789251 056707
TC/M/A0975E/1/02.07/1100