49th INTERNATIONAL COURSE ON LAND DRAINAGE

Wageningen - The Netherlands 2009

DRAIN SPACING

WORKBOOK

Authors
H.P Ritzema
K.J. Lenselink
ILRI
(revised May 2009)

Lecture notes for the International Course on Land Drainage are not official publications. They
may be altered from year to year.
 2

Contents

- EXECUTIVE SUMMARY .......................................................................................... 3

1 INTRODUCTION ........................................................................................................ 4
2 TEACHING OBJECTIVES ........................................................................................... 4
3 RELATED COURSE SUBJECTS................................................................................... 5
4 TEACHING METHODS .............................................................................................. 6
5 BASICS OF GROUNDWATER FLOW ........................................................................... 8
6 SPACING OF HORIZONTAL SUB-SURFACE DRAINS ................................................. 18
7 USING EXCEL FOR DRAIN-SPACING CALCULATIONS .............................................. 23
8 EXERCISE SUB-SURFACE DRAINAGE ..................................................................... 28
9 SHORT TEST .......................................................................................................... 35

(Separately:)
10 SOLUTIONS TO THE PROBLEMS ............................................................................. 37
11 SOLUTION TO THE EXERCISE ................................................................................. 48
12 ANSWERS TO THE SHORT TEST.............................................................................. 60
 3

EXECUTIVE SUMMARY

Course: International course on land drainage/ICLD

Subject: Drain spacing

Lecturer: P.W. Vehmeyer, ILRI

Time allocation: Lectures : 10 hours

Exercises : 16 hours
Test : 2 hours
Study time : 7 hours

Lecture notes: Portions of ILRI Publication 16 (second completely revised edition

1994) and this workbook.

Teaching method: Lectures, exercises, a video presentation, and computer use and
demonstration.

Objective: To introduce the basics of groundwater flow, and the principles and
the application of the most common equations for calculating the
drain spacing under various soil and hydrological conditions.

Related lectures: Drainage for Agriculture

Soils and Drainage
Water in the Unsaturated Zone
Auger-Hole Measurements
Sub-Surface Drainage Design
Environment
Design Exercise Sub-Surface Drainage
 4

1 INTRODUCTION

Sub-surface Sub-surface drainage systems are used in agricultural lands (i) to

drainage control the water table or (ii) to control the amount of salts in the root
zone. A sub-surface drainage system induces the excess of water and
salts to flow through the soil towards wells, mole drains, pipe drains,
and/or open drains, from where it can be evacuated.

Theory In this series of lectures we shall discuss the basics of sub-surface flow
to drains under saturated soil conditions. The lectures follow the
theory presented in the completely revised 2nd edition of ILRI
Reference Publication 16 "Drainage Principles and Applications" 1994. For a
definition of the technical terms used in this workbook, please see the
Glossary of Publication 16, on pages 1095-1106.

This work book The purpose of this workbook is to guide you, the course participant,
through the relevant theory of Publication 16. Chapter 2 of this
workbook states the objectives of this series of lectures, Chapter 3
indicates the links with other lectures in the course, and Chapter 4
discusses the teaching methods. Fundamental physical concepts
governing the flow of groundwater to drains and wells are treated in
Chapter 5. These physical principles underlie the equations that are
used to describe (i) the flow of groundwater to parallel drains or wells
and (ii) a number of seepage problems. Such drainage equations relate
drain properties, like depth and spacing, to hydrological soil
parameters, like the depth of the water table, the soil’s hydraulic
conductivity, and the corresponding drain discharge. The most
common steady- and unsteady-state equations are discussed in
Chapters 6 and 7. Their application under different soil and
hydrological conditions will be shown in a special exercise
(Chapter 8). Some test questions are given in Chapter 9.
 5

2 TEACHING OBJECTIVES

Goals After this series of lectures and exercises you, as a participant, can:
 explain the flow of water through saturated soil, using common
physical properties and basic laws;

 solve simple groundwater flow problems based on Darcy -,

Laplace -, and Dupuit-Forchheimer theory;
 quote and apply steady-state drainage equations of Hooghoudt
and Ernst;.
 make and use a Hooghoudt spreadsheet;
 reproduce and apply non-steady drainage equations of Glover-
Dumm and De Zeeuw-Hellinga;
 explain the advantages, limitations, and interconnections of the
above drainage equations.

Approach In studying this subject, it is important to realise that your main aim
should be to understand the analytic approach of a drainage problem,
and not to learn facts and figures by heart.
 6

3 RELATED COURSE SUBJECTS

During the design stage of a sub-surface drainage system, you must

Computations interact with many related subjects, and not only apply equations. Of
course, drainage equations are needed when you want to design a sub-
Horizontal systems surface drainage system (Lecture Sub-surface Drainage Design and
Publication 16, Chapter 21). Their application is based on the
agricultural needs (Lecture Drainage for Agriculture and Publication
Agriculture 16, Chapter 17) and the prevailing soil and hydrological conditions, of
which the hydraulic conductivity is the most important one (Lecture
Hydr. conductivity Drainage Research and Publication 16, Chapter 12). These soil
hydrological constants are difficult to measure (Workshop Auger-Hole
Method). The elevation of the water table is also influenced by the
Unsaturated flow flow of water in the unsaturated zone (Lecture Water in the
Unsaturated Zone and Publication 16, Chapter 11).

With the introduction of a sub-surface drainage system you are

manipulating the water table and the water and salt balance in your
project area in order to increase or to sustain agricultural production.
However, these manipulations trigger side-effects in the surrounding
Environment environment (Lecture Environment and Publication 16, Chapter 25).
You should also consider these effects when you design a sub-surface
drainage system.
 7

4 TEACHING METHODS

Blocks The course material of our subject is divided into five blocks. It will
be presented as a combination of the following elements: lectures,
short problems to solve, self-study, a video presentation, computer use
and demonstration, a test, and a larger exercise.

After the first introductory classroom lecture, the second block on

groundwater flow is a combination of lectures, short problems and
Video self-study, while also the video "A viscous fluid model for
demonstration of groundwater flow to parallel drains" will be shown.
Block 3 has an identical structure (except for the video). The
Lectures classroom sessions will be used to introduce the subject, to explain the
theory, and to highlight the most important aspects.

Problems Short problems, contained in this workbook, will also be introduced in

Blocks 2 and 3 during classroom sessions, after which you, as a
Working sessions participant, can start working on them (working session). These
sessions will include the use of spreadsheets on the computer.
Questions during these working sessions will give rise to comments
and explanations from the lecturer. You are supposed to finalise the
exercise as a home assignment, so as to be ready by the next classroom
Individual study session (individual study). In this classroom session, the completed
exercise will be discussed, and a short summary of what was learnt
and how it relates to the next topic will be given.

Final exercise Block 3 is followed by block 4, which is a special Exercise Sub-

Surface Drainage, in which the practical use and limitations of the
drainage equations are illustrated. The subject is concluded by a
Test written test (block 5), so that you yourself and the course management
can assess your level of understanding (marking is done for this
purpose only!). Table 1 presents an overview of the approximate time
per block and per teaching element.
Materials During lecturing, overhead sheets will be used, as well as the
whiteboard. For the short problems and the Exercise you need a
pocket calculator. For part of the exercises you will need a personal
computer. For the explanation of the Exercise (Block 4), the lecturer
will use a spreadsheet program on the personal computer (and a
beamer).
 8

Table 1 Approximate study load in hours

Block Topic Lect- Exer- Indiv.
ures cises study
1 Introduction 1 - -
2 Basics of groundwater flow 2 2 1
3 Drain spacing 7 8 2
4 Exercise sub-surface drainage - 6 3
5 Test - 2 1
Total 10 18 7
 9

5 BASICS OF GROUNDWATER FLOW

 Study the theory presented in Chapter 7 of Publication 16

Water properties First, a number of physical properties of water are defined, and some
fundamental laws are illustrated, with emphasis on the conservation
Basic laws of mass, the continuity equation, and the types of energy (Publication
16, Sections 7.1 - 7.3).

The fundamental equation describing one-dimensional flow of

Darcy's Law groundwater is Darcy's Law (Publication 16, Chapter 7.4). The
proportionality factor in this equation is known as the hydraulic
conductivity, which is the reciprocal of the resistance to flow. The
Water conductivity hydraulic conductivity depends on properties like the temperature of
the water, the grain-size distribution and the porosity of the soil, and
the size and shape of its grains. It is difficult to measure accurately
and it can vary at short distances (Lecture Soil Hydrological
Constants and Practical Auger-Hole Method).

Darcy's Law can be used to solve simple one-dimensional

groundwater flow problems, as illustrated in Problems 1 & 2.
 10

Problem 1 Calculation of lateral seepage flow

In a sloping area, a sandy layer lies on top of an impervious clay layer. Two piezometers
have been installed to monitor the piezometric pressures in the sandy layer (Figure 1).
 Indicate the direction of groundwater flow in Figure 1.
 Assuming the groundwater flow takes place only through the sand layer, calculate the
rate of flow through an area 500 m wide perpendicular to the shown section.
Note: The water table runs parallel to the impervious layer.

Figure 1 Groundwater flow in a sloping sandy area

 11

Problem 2 Calculation of upward seepage

In an agricultural area, a thick clay layer with a low permeability (i.e. an aquitard) overlies a
permeable aquifer. The piezometric level in the aquifer exceeds the phreatic level in the clay
layer as shown in Figure 2.
 Calculate the upward seepage rate.
 How high is the vertical resistance of the clay layer?

Figure 2 Upward seepage in a clayey aquitard

 12

Until now we have only treated one-dimensional flow problems,

however most groundwater flow is two or three dimensional. Two-
dimensional flow can be described by a set of partial differential
Streamlines equations: streamlines and equipotential lines (Publication 16,
Equipotential lines Chapter 17.6). Streamlines describe the flow path of water particles,
whereas equipotential lines connect points with equal energy levels.

Streamlines and equipotential lines intersect each other at right

angles, thus forming a flow net which divides the flow region in
Flow nets "approximate squares", so-called flow nets. By definition, each
square represents an equal amount of flow and an equal decrease in
energy level, so that small squares indicate areas of high flow
intensity and large squares of low flow intensity.

It is hard to visualise the flow of groundwater, but with the viscous

Flow model fluid model groundwater flow towards parallel drains can be
simulated. In the soil, the water flowing towards the drains meets
(1) external resistance from friction with the walls of the pores and
(2) internal resistance due to viscosity.
In the model this resistance is simulated by:
 replacing the flow through the pores by flow between two
narrow-placed parallel plates; and
 replacing the water by an oil with a higher viscosity.

The use of the model is demonstrated in a video showing the

Video groundwater flow to parallel drains using an oily fluid. For more
information see: Homma, F. 1968. A viscous fluid model for
Reference demonstration of groundwater flow to parallel drains.
ILRI Bulletin 10, 31 p.
 13

Problem 3 Groundwater flow to parallel drains

The video illustrated how the viscous fluid model was used to simulate the flow of
groundwater to parallel drains. Streamlines were made visible by using a blue tracer.
 Can you draw the streamlines for the two situations presented in Figure 3?
 In which situation will the influence of the impervious layer be more pronounced?

Figure 3 Simulation of groundwater flow using the viscous fluid model

 14

Laplace equation Two-dimensional flow can be calculated using the Laplace equation.
However, a solution to a particular groundwater flow problem can
Boundary only be obtained if the conditions at the boundaries are known
conditions (Publication 16, Section 7.7).

Mathematically exact solutions to two-dimensional flow problems

are rather complex. Furthermore, the boundary conditions are often
not known exactly, the soil is - in general - heterogeneous, or the
recharge to the groundwater is not uniform distributed. Under these
circumstances, results of the same accuracy can be obtained from
simplified methods based on a number of assumptions. Several
Dupuit- examples, based on the Dupuit-Forchheimer theory, are presented in
Forchheimer Section 7.8 of Publication 16. In Problem 4, the Laplace Equation is
used to solve a two-dimensional groundwater flow problem.
 15

Problem 4 Calculation of a curved water table

Figure 4 shows a strip of land drained by two parallel ditches, with constant, but different
water levels. A constant flow, R, recharges the water table.
When R = 0.008 m/d h1 = 8.0 m L = 50 m
K = 0.2 m/d h2 = 5.0 m
and knowing that the equation of the curved water table reads as below, answer:

R 2  h 2 - h1 RL 
2 2
2

h =- x + +  x + h12
K  L K 


 At which distance x from the left-hand ditch reaches the water table its highest point?
 What is the maximum value of h?
 What is the discharge into the right-hand ditch, per 100 metre length along the ditch?
 What happens after the recharge ceases?

Figure 4 A strip of land drained by two parallel drains

 16

Groundwater flow problems can only be solved with the Darcy or

Laplace Equation if the hydraulic conductivity is known. It has
Hydraulic already been mentioned that the hydraulic conductivity is not a
conductivity constant and that it is difficult to measure (see also Lecture 14 Soil
Hydrological Constants or Publication 16, Chapter 12). An example
of measuring the hydraulic conductivity in the laboratory using the
Darcy equation is given in Problem 5.
 17

Problem 5 Calculation of the hydraulic conductivity

Figure 5 (overleaf) shows the vertical section of a container filled with three different soil
layers. A constant flow of water is maintained through the container. The rate of flow is
measured by collecting the discharge at the outflow in a cylinder of known volume over a
3
certain time period: Q = 80 cm in 5 min. The cross-sectional area of the container is 1520
cm2. The hydraulic head is observed at various levels by means of eight piezometers:
h1 = 2.365 m h5 = 1.587 m

h2 = 2.345 m h6 = 1.364 m

h3 = 2.338 m h7 = 1.362 m

h4 = 2.034 m h8 = 1.360 m

 Calculate the hydraulic conductivity of each soil layer.

 18

Figure 5 Laboratory set-up to measure the hydraulic conductivity of a three-layer soil profile
 19

6 SPACING OF HORIZONTAL SUB-SURFACE DRAINS

 Study the theory presented in Chapter 8 of Publication 16

Drainage equations Drainage equations relate technical criteria to soil characteristics and
to agricultural design criteria (Publication 16, Section 8.2.4).
Technical criteria Technical criteria or drain properties are the depth, size and spacing
of drains or wells. Soil characteristics are the hydraulic conductivity,
Soil properties the drainable porosity and the depth of the impervious layer. And
agricultural criteria are the required depth of the water table and the
Agricultural criteria corresponding design discharge. For more information on drainage
design criteria see Chapter 17 of Publication 16.

All equations discussed in the following lectures are based on the

theory and assumptions presented in Chapter 5 of this workbook
(Basics of Groundwater Flow). There are two types of drainage
equations: steady-state equations and unsteady-state equations. Both
types are derived using the Dupuit-Forchheimer theory (Publication
16, Section 8.2). In addition to this, they are based on the following
Assumptions assumptions:
 the pattern of flow to the drains is one- or two-dimensional;
 recharge to the groundwater is uniform over the field;
 there is no spatial variability within a soil layer, and;
 the entrance resistance of the drains can be neglected.

Steady-state equations are based on the assumption that the recharge

to the groundwater, stemming either from precipitation or irrigation,
Steady-state equals the drain discharge. We shall discuss only the three most
equations common steady-state equations, i.e. the Donnan equation, the
Hooghoudt equation and the Ernst equation.
 20

Donnan The Donnan equation is based on a flow pattern of parallel, horizontal

equation streamlines (Publication 16, Section 8.2.1). In addition, a
homogeneous soil profile is assumed and the drains are placed on top
of the impervious subsoil. There is only flow above the drain level.
These assumptions limit the use of the Donnan equation.

Hooghoudt The Hooghoudt equation (Publication 16, Section 8.2.1) is also based
equation on the Dupuit-Forchheimer assumptions, but can be used for drains
installed at any depth in a homogeneous soil or in a two-layered soil
profile, provided that the interface between the layers is at drain level.
This means that flow to the drains takes place above and below drain
level, and that the flow lines will converge towards the drains.
Hooghoudt accounted for this additional radial head loss by replacing
the actual depth of flow by a smaller, equivalent depth.

Ernst The Ernst equation uses a two-dimensional flow pattern (Publication

equation 16, Section 8.2.2). The flow path and the corresponding head losses
are divided into a vertical, a horizontal and a radial component.

Application of The question which equation to use in a particular situation depends

steady-state mainly on the soil profile and the relative position of the drains in this
equations profile (Publication 16, Section 8.2.3). The Ernst equation is
generally applicable for two-layered soil profiles, although the
Hooghoudt Equation gives better results when the calculated drain
spacing exceeds the depth of the permeable layer four times or more.
 21

Problem 6 Drain spacing calculation using Hooghoudt

A farmer in the north-eastern part of The Netherlands wants to drain his pasture fields. The
following data are available:
Soil profile:
0 - 0.20 m below soil surface K = 10.0 m/d
0.20 - 1.20 m below soil surface K = 0.35 m/d
1.20 - 6.20 m below soil surface K = 0.70 m/d
Below 6.20 m, the soil is considered impervious.

The rainfall surplus, being the rate of recharge to the groundwater, is 8 mm/d, and the
required depth of the water table is 0.40 m below soil surface. Drainage will be by means of
pipe drains with a radius of 0.10 m and a depth of 1.20 m below soil surface.
 Make a drawing of the situation, and;
 Calculate the drain spacing using the Hooghoudt equation.

Problem 7 Drain spacing calculation using Ernst

As an alternative to the situation in Problem 6, we lower the drain depth to 1.50 m below soil
surface. The values of q, r0 , and the required water-table depth remain the same.
 Make again a drawing of the situation and calculate the drain spacing using the
appropriate form of the Ernst equation.
 What is the contribution of the vertical, horizontal and radial flow to the total head loss
(in %)?
 22

Unsteady-state The assumption that the recharge to the groundwater equals the drain
equations discharge is not always justified. Common examples are irrigated
fields, or humid areas with high-intensity rainfall. Under these
conditions, the unsteady-state equations are sometimes more
appropriate (Publication 16, Section 8.3).

Glover-Dumm The Glover-Dumm equation describes the fall of the water table after
equation an instantaneous rise above drain level.

De Zeeuw-Hellinga The De Zeeuw-Hellinga equation describes a fluctuating water table,

equation i.e. when the recharge to the groundwater varies with time.

The use of the unsteady state equations is restricted because it is even

Drainable porosity more difficult to measure the drainable porosity in the field than to
measure the hydraulic conductivity. For more information see Section
11.3.5 of Publication 16.

Problem 8 Drain spacing calculation using Glover-Dumm

In an irrigated area the hydraulic conductivity of the soil is 0.25 m/d. The depth of the
impervious layer is assumed to be at 11.80 m below soil surface. The drainable porosity of
the soil is 8% (µ = 0.08). The fields are irrigated every 10 days. The field application depth is
100 mm of which 25% is assumed to become percolation loss. The water table is not
permitted to rise within 1.0 m depth from the soil surface. Drains with a radius of 0.1 m will
be installed at a depth of 1.80 m below the soil surface.
 Calculate the drain spacing with the Glover-Dumm equation.
Between two irrigations, the water table drops from 1.0 m to 1.3 m.
 What is the total discharge during this period in mm and what is the average discharge
per day if one does not take the shape of the water table into account?
 What is the actual discharge rate q (mm/d) as a function of time?
 23

Comparison between Steady State and Unsteady State Equations

Steady state equations are derived under the assumption that the
recharge to the groundwater equals the drain discharge and
Publication 16,
consequently that the watertable remains in the same position. In
Chapter 17.3.4:
reality, the recharge to the watertable varies both in space and time.
Dynamic Consequently the flow of groundwater towards the drains is not
Equilibrium steady: thus it does not seem logical to use steady state equations. For
several reasons, however, steady-state equation can be used:
1. In areas, where the drainage problem is caused by excess rainfall,
we can assume that the rainfall is evenly distributed of the drainage
unit. Furthermore, in humid and temperate climate zones, there are
Publication 17.5:
periods of rather uniform, medium intensity rainfall and during
Example 17.1 The these periods the watertable will be more or less constant and
Netherlands consequently the discharge will be equal to the recharge. Under
these “average state” or “dynamic equilibrium” conditions steady
state equations can be used.

In arid, irrigated areas, the watertable often rises sharply during the
application of irrigation and than recedes gradually until the next
irrigation. Under these conditions it seems more logical to use
unsteady state equations. However, there are again various reasons
why under these conditions also the steady state approach can be used,
i.e.:
2. In irrigated areas, especially in small-scale agriculture, a series of
parallel field drains often serves more than one fields. In general,
these fields are not irrigated at the same time or even not on the
same day. Thus the drawdown of the watertable after irrigation in a
particular field is mainly caused by the flow of groundwater
towards the neighbouring fields (not irrigated at the same time/day)
and not so much by flow towards the drains. The flow towards
these drains depends more on the average depth in the drainage (or
irrigation) block. Again it is not the daily drop of the watertable in
a particular fields that is important, but much more the average
watertable in all fields covered by a particular drain over a period
of e.g. an irrigation interval.
 24

3. In irrigated areas it is much more logical to control the watertable

Publication 16, by managing the irrigation supply than by creating extra drainage

Chapter 15: Salinity capacity (compare: when your bathroom is flooded because your
bathtub is overflowing, you first close the tap instead of making
Control
another waste-pipe). By increasing irrigation efficiency and/or
lowering irrigation gifts waterlogging problems can often be
solved. Under these conditions, subsurface drainage is mainly
required to control the salts imported by the irrigation water, i.e.
the drainage design discharges is based on the leaching
Publication 17.5: requirement. Changes in soil salinity, however, are not so much
influenced by the daily fluctuation of the watertable, but more by
Example 17.3
the total inflow of salt by irrigation over the season. The subsurface
Egypt
drainage requirements should be calculated over this longer period,
over which the change in storage (= change in water table) is small
compare to the total in- and outflow. Again the steady state
approach can be used to describe this “quasi” steady condition.

The question of whether to use the steady-state or the unsteady-state

Publication 16,
approach further depends on the availability of data. The steady-state
Chapter 8.4:
approach required less data and is thus preferred for large scale
Availability of input
implementation, because less data requirements results in costs for
data investigations.

So maybe you wonder what the use is of unsteady state equations?

Unsteady-state equations are not used for the design of subsurface
drainage systems, but mainly to analyse the functioning of such
systems, as they better describe the variation in time of the watertable
fluctuation. Thus to derive the drainage design criteria the unsteady-
state approach has to be used. Fortunately, it is possible to combine
Equation 8.40 both approaches so that the corresponding criteria can be converted
into one other. In this way results from field experiments, conducted
under unsteady-state conditions, can be converted into steady-state
criteria that can be applied on project scale.
 25

7 USING EXCEL FOR DRAIN-SPACING CALCULATIONS

You have learnt that we compute the drain spacing with steady-state equations (Donnan,
Hooghoudt, Ernst) or with unsteady-state equations (Glover-Dumm, De Zeeuw-Hellinga). In
problems 6 (Hooghoudt), 7 (Ernst), and 8 (Glover-Dumm) you have exercised with the most
important ones. There are quite some computations involved, which become tedious
especially if the spacing needs to be calculated frequently. Therefore, it is useful to automate
these computations, which can conveniently be done with a spreadsheet program like Excel.
Below, we start with making a calculation template for Hooghoudt’s equation, as an example
of what is possible. It can be considered as a simple alternative for Van der Molen and
Wesseling’s Fortran program (1991), mentioned in Publ. 16, page 268.

The exercise: Use Excel for making a simple Hooghoudt calculation program
See also: Publication 16, Chapter 8, Section 8.2.1: The Hooghoudt Equation

Your task is to complete a worksheet for a simple Excel program, which allows you:
(i) to input data for the different variables; and then
(ii) to perform an iteration to find the drain spacing according to Hooghoudt's equation:
q*L2 = 8*Kb*d*h + 4*Kt*h2.

Procedure:
1. Making a specific calculation program can be done in one of the programming
languages. An example of a Fortran program can be found on your floppy disk as
A:\HOOG\HOOG.EXE. Run this program and enter the following data:
Maximum hydraulic head (h) 1.50 m
Drain radius (ro) 0.05 m
Hydraulic cond. above drain (Kt) 0.10 m/d
Hydraulic cond. below drain (Kb) 1.0 m/d
Depth of impervious layer (D) 3.0 m
Drain discharge (q) 0.005 m/d
Initial value of drain spacing (L) 20 m
and find the final calculated drain spacing, L, as 76 m.
 26

2. Such a small calculation program can also be made in Excel, and we are going to make
it in this small exercise. Start Excel 97 and Open file A:\HOOG\HOOG_1.XLS.

3 Note that we already prepared a template with seven variables, a column where you
must enter INPUT DATA, and a section for the iteration to save time. If necessary,
make a situation drawing on paper and note that:
h = drain depth - allowed water table;
D = aquiclude depth - drain depth;
Kt = upper conductivity (t for top)
Kb = lower conductivity (b for bottom);
q = steady recharge = drain discharge;
ro = drain radius (calculate equivalent for open drains from ro = u/);
L = drain spacing.
The entered values for the variables relate to the same situation as under 1) above.

4. Now press the [] key a number of times to show the screen to the right (addresses K1,
etcetera). This screen holds the Calculation procedure, which is the basis for the small
program. Remember that the calculation of L is not straightforward, since the
equivalent depth d must be calculated first and d is a function of the wanted L. Chapter
8 mentions a direct solution procedure for d, based on Van der Molen & Wesseling
(1991). Let us start entering the missing equations (in cells N11-N15, N17, N18, N20,
and E20).

5. For the calculation procedure we need values for D (aquiclude depth minus drain
depth), for the drain radius, ro, and an initial value for L. Both D and ro are found from
the input data, while the initial value for L must be specified under Assume in the first
template. This assumed value is then copied to the Calculation procedure screen. So:
Make the contents of cell D20 appear in cell N11. Then calculate depth D as =E11-E9
in cell N12. Then calculate x in cell N13 by entering the appropriate formula.
 27

6. Calculating F(x) requires a bit of work. Press [] again a number of times and see
columns V and W, which can assist us in calculating F(x) as a series. In column V we
have entered a number of n-values (numbers 1,2,3,..., 39), in column W we have
entered =ln(1/tanh(n*x)), because the hyperbolic cotangent is not present in Excel’s
functions, but the hyperbolic tangent is (and cotanh = 1/tanh). See Eq. 8.11 in Publ. 16
(page 270). Note that the higher terms are very small indeed, and certainly may be
neglected (n = 39 as a maximum was chosen rather arbitrarily). We find the F(x) by
taking double the sum of all Series terms: Put this equation in cell N14 of the
Calculation procedure screen.

7. Calculate the value of the equivalent depth, d, in cell N15 according to the given
equation (which is Eq. 8.9 of Publ. 16).

8. Now calculate the value of h in cell N17, and the value of 8*Kb*d*h+4*Kt*h2 (=q*L2)
in cell N18. Divide it by q (watch the units !) and take the square root in cell N20. Make
the contents of N20 appear in cell E20 of the original template.

9. The computation model is now ready. To use it, change the initial value of 20 in the
first template to something closer to 60 and keep on changing until the two are equal
(this is demonstrating the iteration process). The final value should be 76 m, the same
as we found with the Fortran program.

10. Your simple Hooghoudt calculation program is now ready, apart from a possible
protection against accidental overwriting. You may protect the worksheet, but you have
to unprotect ranges E9..E15 and D20, otherwise you cannot enter input data. Note that,
upon opening the file you see nothing of the Calculation procedure screen, nor of the
Series calculation columns. You do not have to bother about them from now on. The
only thing you do is enter input data and then change assumed L-values until you get
identical values: that is your drain spacing.
 28

11. You may find that you have invested quite some time in making a simple calculation
program: that is correct, but your benefit lies in the time-saving you will realise if you
have to calculate drain spacings for many different sets of variables. If you only have to
calculate an occasional value, writing a computer program does not make sense !

12. Save your Hooghoudt program as A:\HOOG\HOOG_2 on your floppy. You now have
HOOG_1.XLS, HOOG_2.XLS, and HOOG.EXE on your floppy under subdirectory
A:\HOOG\.

13. There are ways to avoid the manual iteration, as we may demonstrate later.

At the end of this exercise, you may be aware that embellishments (adding colour, adding
graphs, avoiding impossible input data, etc.) are possible. You can do that on your own. An
example is given on your floppy as A:\HOOG\HOOG_4.XLS. A print of the template page is
shown on page 27 (Figure 6).

You may also realise that similar templates can be made for other drain-spacing equations. A
draft of such other templates may be given to you later (called SPACING.XLS), which
contains templates for different drain-spacing equations in different sheets of the same file
(Hooghoudt, Ernst, Glover-Dumm, and De Zeeuw-Hellinga). NOTE that this is a draft
version, which may not work error-free !!!
 29

Figure 6 A possible Excel calculation template for Hooghoudt’s drain-spacing eqaution

 30

8 EXERCISE SUB-SURFACE DRAINAGE

Input data To apply drainage equations we need three sets of input data:
 the agricultural criteria: these are the criteria which are used to
translate the agricultural objectives in parameters that can be
used in the design;
 the technical criteria: these are the criteria which specify the type
of drainage system that we want to install, and;
 the soil characteristics, which specify the soil and soil’s
hydrological conditions.

Fixed inputs For a design engineer, the agricultural criteria and the soil
characteristics are data, which can not be changed: they are fixed input
Design variables data. On the other hand, the technical criteria are the design tools for
the engineer: for an area with fixed soil characteristics, the design
engineer can design alternative systems by selecting different types of
drainage systems, different construction methods or materials, each
satisfying the same agricultural criteria. For more information on
drainage criteria, see Publication 16, Chapter 17.

Simplification For this design, the engineer needs the drainage equations. These
drainage equations do not describe the complex reality of a drainage
problem, as we find it in the field, but they only describe a simplified
situation. In other words, we have to simplify the complex reality
before we can apply a drainage equation (Figure 7).

Judgement To select the appropriate equation and to make the correct

simplifications requires a thorough knowledge of the basics of
groundwater flow and the assumptions and boundary conditions on
which the equations are based. This is the most difficult part of a
design.
 31

Exercise In this chapter of the workbook a drainage exercise is given, in which

you can see that the final result of your calculations depends
very much on the assumptions and simplifications you have to make
in order to apply the appropriate drainage equation.

Figure 7 Application of the drainage equations in a complex

situation
 32

Exercise sub-surface drainage: Application of the drainage equations

In an irrigated area, high water tables occur due to the recharge by irrigation water losses.
Design a pipe drainage system to control the water table under the following
conditions:
Agricultural drainage criteria:
 a steady design discharge of 1 mm/d;
 the water table should be controlled at 1.0 m below soil surface.
Technical criteria:
 drains will be installed at 2.0 m depth;
 pvc pipe drains with a radius (r0) of 0.10 m will be used.

In the area the following soil investigations have been conducted:

 the texture of the soil has been established from samples of an augering up to the
impermeable layer (Table 2);.
 thirty auger-hole measurements at a depth ranging from 2.0 to 3.8 m below soil surface
(Table 3).

Table 2 Soil texture analysis

Depth (m -ss) Sand (%) Silt (%) Clay (%) Texture

0.0 - 0.5 6 34 60 clay

0.5 - 1.5 8 20 72 clay

1.5 - 2.0 11 20 69 clay

2.0 - 2.5 34 29 37 clay loam

2.5 - 3.0 26 41 33 clay loam

3.0 - 4.0 6 61 33 silty clay loam

4.0 - 5.0 15 65 20 silt loam

5.0 - 6.0 8 44 48 silty clay

6.0 - 6.8 23 8 69 clay

> 6.8 3 12 85 clay

 33

Table 3 Hydraulic conductivity measurements

N Auger-hole bottom Water depth in Water table elevation Hydraulic conductivity

(m -ss) auger hole (m) (m -ss) (m/d)
1 2.00 0.86 1.14 0.07
2 2.00 0.87 1.13 0.02
3 2.00 0.49 1.51 0.09
4 2.00 0.90 1.10 0.02
5 2.00 0.79 1.21 0.11
6 2.00 0.90 1.10 0.08
7 2.00 0.85 1.15 0.03
8 2.00 0.90 1.10 0.09
9 2.00 0.47 1.53 0.04
10 2.00 1.17 0.83 0.10
11 2.00 0.82 1.18 0.08
12 2.00 0.50 1.50 0.05
13 2.00 0.85 1.15 0.05
14 2.00 1.35 0.65 0.05
15 2.80 0.98 1.82 2.00
16 3.00 1.38 1.62 0.03
17 3.00 1.42 1.58 2.27
18 3.00 1.32 1.68 2.64
19 3.03 1.00 2.03 0.01
20 3.03 1.00 2.03 0.02
21 3.03 1.20 1.83 0.02
22 3.10 1.10 2.00 0.92
23 3.10 0.96 2.14 2.10
24 3.10 1.13 1.97 1.60
25 3.20 1.16 2.04 0.01
26 3.45 1.05 2.40 0.12
27 3.50 0.95 255 1.10
28 3.50 1.10 2.40 1.04
29 3.50 1.12 2.38 0.88
30 3.80 1.00 2.80 0.73
 34

To complete the exercise, apply the following procedure:

1. Make a drawing of the simplified soil profile:

a) How many different soil layers can you distinguish?
b) What is the thickness of the different soil layers?
c) Calculate the “average” hydraulic conductivity of each soil layer in two ways, i.e.
by using the geometric and the arithmetic means.
d) Calculate the ratio of the hydraulic conductivity above/below drain level.
e) What will be the geometric and arithmetic means if we assume that the soil is
homogeneous?
f) Which value(s) of the hydraulic conductivity is (are) representative?

If you want, you may use Excel to do your computations. The data of Table 2 and Table
3 are in a file named A:\COND\COND_1.XLS on your Excel floppy.

2. Proceed to calculate the drain spacing:

a) Select the appropriate drainage equation.
b) Select representative values of the hydraulic conductivity.
c) Calculate the drain spacing.

3. Can you also solve the problem by using the simplified Hooghoudt equation?
a) Select a representative value of the hydraulic conductivity.
b) Calculate the drain spacing.

4. Which spacing do you obtain if you assume that the soil profile is homogeneous?
a) Select a representative value of K for the homogeneous soil profile.
b) Calculate the drain spacing.
c) What will be the drain spacing if you use the arithmetic mean of the hydraulic
conductivity?
 35

5. Now compare the drain spacings calculated in questions 2c, 3b, 4b, and 4c and draw
your conclusions. You may also want to know that, if we would (erroneously) use the
Ernst equation for this situation, we would find L = 77 m. What conclusion would that
lead to? Which equation is the most appropriate for this case and why?

6. Let us now, for this question only, assume that the available drainage machines can
only install drain pipes up to a maximum depth of 1.50 m below the soil surface. Then
the technical criteria change: you have to reduce the drain depth from 2.0 m to 1.5 m.
The width of the trench made by this machine is 0.25 m.
a) Which drainage equation should now be used?
b) Calculate the drain spacing.
c) Calculate the drain spacing with the same equation, assuming a homogeneous soil
profile (although it is an unrealistic exercise).
d) Calculate the drain spacing assuming a homogeneous soil profile and using
Hooghoudt.
e) Compare the results and comment.

7. In questions 2, 3, 4, 5, and 6 you have calculated the drain spacings assuming steady-
state conditions. Now check if the calculated drain spacing can also cope with
unsteady-state conditions. For applying the unsteady-state approach an additional
agricultural criterion is required:
 the maximum permissible height of the water table is 0.75 m below the soil surface
(whereas in the steady-state equations we used an average depth of 1.00 m).

Furthermore, the following information is available:

 during the rainy season, an observation well was installed in the area. It was
observed that in a rainy period a single rainstorm of 15 mm caused the water table
to rise 0.30 m;
 during the irrigation season water is applied every 20 days.
 36

Calculate:
a) the drainable pore space µ.
b) the instantaneous recharge assuming that the steady-state drainage criterion
(equalling 1 mm/d) over a irrigation cycle can be transformed into an instantaneous
recharge.

8. Select an equation to calculate the fluctuation of the water table and enter the known
parameters, so that the water table depth only depends on day t.

9. Calculate the fluctuation of the water table for days 0, 5, 10, 15, 20, assuming that the
water table is at 0.75 m below soil surface after the first irrigation (which gives you
h0). Note the height of the water table at day 20, just before the next irrigation.

10. Check if there is a relation between the steady-state conditions (considering the
simplified Hooghoudt equation) and the unsteady-state conditions (considering the
reaction factor).
 37

9 SHORT TEST

To check your knowledge on the subject a test is included in this series of lectures. To assess
your current level of understanding yourself, try to answer the following questions.

1. What are the three components of the total energy head? Which one of the three can be
neglected in groundwater flow problems?

2. Why do we need the Dupuit-Forchheimer assumptions to solve most groundwater flow

problems?

3. The viscous fluid model can be used to simulate the flow of groundwater to parallel
drains. How can we simulate a flow problem in a soil with a lower hydraulic
conductivity than in the video?

4. How did Hooghoudt account for the extra head loss caused by the converging flow near
the drainpipe?

5. In the Hooghoudt equation the square of the spacing, L2, is proportional to 1/q. What is
the reason that, when q is taken 4 times larger, L does not become exactly half?

6. Why have we used the Hooghoudt equation in Problem 6 and the Ernst equation in
Problem 7?

7. What is the reason that steady-state equations can also be used for drainage problems in
irrigated areas, where the flow is highly unsteady?

8. Which two equations can be used to relate a steady-state criterion with a unsteady-state
criterion? Derive that relationship.

9. A farmer has determined that his crop will give the highest yield if the water table is
 38
maintained at 0.50 m below the soil surface. He has engaged a contractor to install
pipe drains using a trencher, which makes trenches of 0.20 m wide. Corrugated plastic
pipes, r0 = 0.06 m, wrapped with a 7-mm thick envelope will be used. The drains will
be installed at a depth of 1.0 m. The design discharge is 7 mm/d and the soil has a
uniform hydraulic conductivity of 1.5 m/d up to a depth of 7.0 m. Below this depth, the
soil is considered to be impervious.

Determine the drain spacing with and without considering the drain trench.
 39

10 SOLUTIONS TO THE PROBLEMS

Problem 1 Calculation of lateral seepage flow

The flow rate can be calculated with Darcy's Equation, being

h
Theory in Q=K A (7.19)
L
Publication
where
16, Chapter
Q = flow rate in m3/d
7.4
K = hydraulic conductivity = 1.5 m/d
h = head loss = 2.0 m
L = flow-wise length of section = 300 m
A = area of flow = 15 * 500 = 7500 m2

Substitution of the above data gives

2.0
Q = 1.5 7500 = 75 m 3 /d
300
 40

Problem 2 Calculation of upward seepage

Again the Darcy Equation is used, now written as

Theory in h
v=Ks=K
Publication 16, D

Section 7.4 where

v = discharge per unit area (= apparent velocity) in m/d
K = hydraulic conductivity = 0.05 m/d
s = hydraulic gradient
h = head loss = 0.84 m
D = length of flow = thickness of clay layer = 12.18 m

Substitution of the above data yields for the seepage rate

0.84
v = 0.05 x = 0.0034 m/d = 3.4 mm/d
12.18
The vertical resistance of the clay layer is
D 12.18
C= = = 244 d
K 0.05
 41

Problem 3 Groundwater flow to parallel drains

Theory in First determine the boundary conditions (Figure 8):

Publication 16,  the impervious layer, i.e. the bottom of the model;
Section 7.7  the planes of symmetry, i.e. the vertical planes through the
drain axis and midway between the drains (dashed lines).

The streamlines do not cross these boundary lines. The water table is
also a boundary line, but because there is a constant recharge q, the
water table is neither an equipotential line nor a streamline, and the
starting points of the streamlines are at regular horizontal distances.
Situation A: DL, streamlines not influenced by the impervious layer.
Situation B: D < ¼ L, streamlines deformed by the impervious layer.

Figure 8 Streamlines of the groundwater flow to parallel drains

 42

Problem 4 Calculation of a curved water table

Theory in The general equation of the curved water table is given as

Publication 16,
Section 7.8 R 2  h 2 - h1 RL 
2 2
h2 = - x +  +  x + h1 2
K  L K 

or, with substitution of the known data

h2 = - 0.04 x2 + 1.22 x + 64

At the highest point of the watertable d/dx of h2 = 0.

Differentiation of the above equation gives
d(h 2 )
= - 0.08 x + 1.22
dx
Hence the highest point of the watertable occurs when

- 0.08 x + 1.22 = 0
or
1.22
x= = 15.25 m
0.08
The (maximum) value for h is at that point is found from

h2 = - 0.04  15.252 + 1.22  15.25 + 64 = 73.30

thus: h = 8.56 m

The recharge to the right-hand ditch is found by considering that the

recharge over a width of L - x = 50 - 15.25 = 34.75 m is drained into
the right-hand ditch. The discharge is therefore

Q(r) = R(L - x) 100 = 0.008  34.75  100 = 27.8

 43
3
m /d

If the recharge ceases suddenly, we will first enter into an unsteady

situation, because supposedly the discharge continues, and the
watertable will drop. If also the discharge stops, we get a situation
comparable with Figure 7.27 in Publication 16, to which Dupuit’s
formula applies (equation 7.73) and we get a Q of:

Q=
K
2L
2 2

h1 - h 2  100 =

0.2
= (64 - 25)  100 = 7.8 m 3 /d
100

This gives the same result as substituting R = 0 in the earlier-used

equation, leading to

h2 = - 0.78 x + 64

and the discharge into the right-hand ditch is found by applying

Darcy's law for x = L = 50 m
dh
Q=-K h  100 =
dx
1 d( h 2 )
= - 0.2 x  100 =
2 dx

= - 10 (-0.78) = 7.8 m 3 /d
 44

Problem 5 Calculation of the hydraulic conductivity

In an experiment the following observations were made

Theory in
Q = 80 cm3 in 5 min
Publication 16,
h1 = 2.365 m h5 = 1.587 m
Section 7.4
h2 = 2.345 m h6 = 1.364 m
h3 = 2.338 m h7 = 1.362 m
h4 = 2.034 m h8 = 1.360 m

The hydraulic conductivity of each layer is calculated with the Darcy's

equation
h
Q=K A (7.19)
L
Substitution the appropriate data of each layer, converted to units of m
for length and day for time

80 cm 3 80  106
Q= =  60 x 24 m3 /d = 23.04  103 m3 /d
5 min 5

Q L 23.04  10 3 0.47
K1 =  =  = 3.6 m/d
A  h(1-2) 0.152 0.02

Q L 23.04  10 3 0.2
K2 =  =  = 0.04 m/d
A  h(3-5) 0.152 0.751

Q L 23.04  10 3 0.45
K3 =  =  = 17.1 m/d
A  h(6-8) 0.152 0.004

It is clear that a small error in the head reading for the more permeable
layers will result in a large difference in the K-value that is calculated.
 45

Problem 6 Drain-spacing calculation using Hooghoudt

Theory in A schematic presentation of the problem is given in Figure 9.

Publication 16,
Section 8.2.1

Figure 9 Drain-spacing calculation using Hooghoudt

We can use equation 8.7

2 8K b dh+ 4K t h 2 8  0.70  d 0.8 + 4 x 0.35 x 0.82

L = = = 560 d + 112
q 0.008

1st estimate L = 50 m 
D = 6.20 - 1.20 = 5.0 m  d=
3.02 m

L2 = 502 = 2500 << 560 x 3.02 + 112 = 1803 m too wide

2nd estimate L = 40 m 
D = 5.0 m  d = 2.75 m

L2 = 402 = 1600  560  2.75 + 112 = 1652 O.K.

thus L = 40 m
 46

Problem 7 Drain-spacing calculation using Ernst

Theory in q = 8 mm/d = 0.008 m/d Dv = 0.80 m

Publication 16, ro = 0.10 m Dr = 4.70
Section 8.2.2 m
h = 1.50 - 0.40 = 1.10 m u =  ro = 0.314 m
Db
= 5.0 m
Because the drains are situated in the bottom layer (Figure 10), we
have to use equation 8.22 to calculate the drain spacing
D 2
L 
h = q  v + L + ln Dr 
 K t 8K b D b  K b u 

(8.22)

 0.80 L
2
L 4.70 
1.10 = 0.008  + + ln 
 0.35 8  0.70  5.0   0.70 0.314 

137.5 = 2.29 + 0.0357 L2 + 1.23 L

0.0357 L2 + 1.23 L - 135.2 = 0
thus

- 1.23  1.232 + 4  0.0357  135.2

L= = 46.7 m
2  0.0357

Check if Db and Dr < ¼ L: Db = 5.70 m 

Dr = 4.7 m  O.K

¼ L = 12 m 

The head loss caused by the vertical, horizontal and radial flow is
respectively
Dv = 0.008  0.80 = 0.02 m
hv = q
Kv 0.35
 47
2 2
L 46.7
hh = q = 0.008  = 0.62 m
8 K b Db 8  0.70  5.0

L 46.7 4.70
hr = q ln Dr = 0.008  ln = 0.46 m
 Kr u   0.7 0.314

and the total head loss

ht = hv + hh + hr = 0.02 + 0.62 + 0.46 = 1.10 m

This is OK. So, hv contributes 2%, hh 56% and hr 42% to the total
head.

Figure 10 Drain-spacing calculations using Ernst

 48
Problem 8 Drain-spacing calculation using Glover-Dumm

Theory in During irrigation, the water table rises 0.3 m (= 25%  100 mm  1/
Publication 16, = 0.25  100 1/0.08 = 312.5 mm). As the water table is not permitted
Section 8.3.1 to rise to within 1.0 m of the soil surface, the depth of the water table
just before irrigation must be minimally at 1.3 m (Figure 11).

Figure 11 Drain-spacing calculation using Glover-Dumm

Calculation of drain spacing using the Glover-Dumm equation

1 1
-
 K d t 2  h 2
L =     ln 1.16 o  =
    ht 

1 1
 0.25  d  10  2 
-
0.8  2
=    ln 1.16   = 22.3 d
 0.08   0.5 

We know D and ro, thus we can obtain the equivalent depth d from
 49
Table 8.1

1st estimate: L = 50 m 
D = 10 m d=
3.74 m

L = 22.3 3.74 = 43 m  50 m is too wide.

2nd estimate: L = 40 m 
D = 10 m d = 3.23 m

L = 22.3 3.23 = 40 m  O.K.

The total discharge between two successive irrigations is:

(1.3 - 1.0)  0.08 = 0.024 m = 24 mm water depth (compare with the
deep percolation loss) or an average of 2.4 mm/d.

The actual discharge as a function of time is (equation 8.36, applying a

fourth degree parabola, as done by Dumm, 1960)

with  = 
2
6.89 Kd - t Kd
qt = ho e
L
2
 L2
6.89  0.25  3.23
thus q t = 2
 0.8 e t  0.0028 e αt
40
and
 2 Kd  2  0.25  3.23
= = = 0.062 d 1
 L2 0.08  402
Thus
qt = 0.0028 e-0.062 t

At day 1: q1 = 0.0026 m/d = 2.6 mm/d is calculated, but

the condition that t > 0.2 is not fulfilled. This does
not happen until day 3 (approximately).
At day 10: q10 = 0.0015 m/d = 1.5 mm/d
 50

11 SOLUTION TO THE EXERCISE

1a&b Based on the textural classification (Table 2) we distinguish

Publication 16,
five different soil layers. We can use an Excel spreadsheet bar
Section 3.4.1
chart to further simply the soil profile by only looking to the
clay content (Figure 12):

Depth (m) Texture

Average clay content (%)
0.0 - 2.0 clay
67
2.0 - 5.0 clay loam
Excel97 exercise 31
> 5.0 clay
67

Figure 12 Bar chart of clay contents made in Excel

 51
Note the very large variation in measured K-values within one layer.
Despite this wide variation there is a clear difference between the
hydraulic conductivity values of the two layers. Again you can use a
Excel spreadsheet to visualize this variation, by plotting the
cumulative frequency of the K-values of each layer in a graph (Figure
13).

2.50

2.00
K (m/d)

1.50

1.00

0.50

0.00
0 10 20 30 40 50 60 70 80 90 100
Cumulative Frequency (%)

0.0 - 2.0m 2.0 - 4.0m

Figure 13 Cumulative frequency of the hydraulic conductivity in the top clay layer (0 – 2.0m)
and the underlying clay loam layer (2.0 – 4.0 m).

Combing the soil texture data (Figure 12) and the hydraulic conductivity data (Figure 13)
results the following simplified profile (Figure 14):
 52

Figure 14 Actual and simplified profile

Publication 16, 1c Geometric mean

Section 12.3 *
K = n K1 x K2 x ...... Kn (12.5)
where
n = number of observations.

Arithmetic mean
 Ki
Ka =
n

Excel exercise Layer (m) n Hydraulic

conductivity (m/d)
Geometric
Arithm. St. dev.
00 - 2.0 14 0.06
0.06 0.03
2.0 - 5.0 16 0.30
 53
0.97 0.89
0.0 - 5.0 30 0.14
0.55 0.79

1d Ratio of geometric means

Kt = 0.06 = 1
Kb 0.30 5
Ratio of arithmetic means
Kt = 0.06 = 1
Kb 0.97 16
1e For a homogeneous soil profile
K* = 0.14 m/d
Ka = 0.55 m/d

1f Which value of the hydraulic conductivity is representative

Publication 16, depends on the drainage equation used to calculate the drain
Section 12.3 spacing. Some equations assume a homogeneous soil profile,
i.e. one value of the hydraulic conductivity, others assume a
layered profile, with more than one representative value.

In general, the geometric mean is the most representative

value, although several authors recommend other values.

2) We have assumed a two-layer soil profile (Figure 13) with the

drains at the interface of the two layers, and hence we can use
Publication 16, the Hooghoudt equation to calculate the drain spacing
Section 8.2.1
8 Kb d h + 4 Kt h2
q= 2
(8.7)
L

We shall use the geometric means of the hydraulic

conductivity. Thus we have the following data:
q = 1 mm/d = 0.001 m/d
h = 2.0 - 1.0 = 1.0 m
Kt = 0.06 m/d (soil layer above drain level)
 54
Kb = 0.3 m/d (soil layer below drain level)
D = 3.0 m (Figure 13)

Thus

2 8  0.3  d  1.0 + 4  0.06  1.02

L = = 2400 d + 240
0.001

We can now calculate the drain spacing by trial and error using
Publication 16, Table 8.1.

1st estimate:
L = 100 m 
D = 3.0 m Table 8.1  d = 2.60 m

L2 = 2400  2.60 + 240 = 6.480 m2

6.480 < 10 000 (= 1002) thus assumed spacing is too wide

2nd estimate:
L = 75 m 
D = 3.0 m  d = 2.49 m

L2 = 2400  2.49 + 240 = 6216 m2

6216 > 5625 (= 752) too narrow.

3rd estimate:
L = 80 m 
D = 3.0 m d = 2.52 m

L2 = 2400  2.52 + 240 = 6288 m2

6288  6400 (= 802)
thus the calculated drain spacing is 80 m

Note: if we use the Excel spreadsheet, we find L = 78 m.

 55

3) In the simplified Hooghoudt equation the flow above drain

level is neglected. We can do this because Kt << Kb. So we
select Kb = 0.3 m/d as the representative hydraulic
conductivity:

8 dh
q = Kb 2 (8.6)
L
or

2 8  0.3  d  1.0
L = = 2400 d
0.001

Again we use table 8.1 to calculate the drain spacing:

1st estimate:
L = 90 m 
D = 3.0 m  d = 2.56 m
L2 = 2400  2.56 = 6144 m2 < 8100 m2 too wide.

2nd estimate:
L = 75 m  d = 2.49 m
L2 = 2400  2.49 = 5976 m2 < 5625 m2 too narrow.

3rd estimate:
L = 77.5 m hence d = (2.49 + 2.51)/2 = 2.51 m
L2 = 2400  2.51 = 6012 m2  77.52
thus the drain spacing is 78 m.

N.B. The Excel spreadsheet gives 77 m.

 56

4) If we assume that the soil profile is homogeneous, we use the

geometric mean hydraulic conductivity: K = K* = 0.14 m/d,
thus

2 8  0.14  d  1.0 + 4  0.14  1.02

L = = 1120 d + 560
0.001
1st estimate:
L = 75 m 
D = 3.0 m d = 2.49 m

L2 = 1120  2.49 + 560 = 3348 < 5625 too wide.

2nd estimate:
L = 50 m  d = 2.29 m
L2 = 1120  2.29 + 566 = 3125 > 2500 too narrow.

3rd estimate:
L = 55 m  d = 2.33 m
L2 = 1120  2.33 + 560 = 3170  3025
thus the drain spacing is 55 m (Excel gives 56 m)
Note: neglecting the flow above drain level will introduce a
large error in this case, because Kt = Kb.

If we use the arithmetic mean of the hydraulic conductivity we

get:
K = 0.55 m/d and

2 8  0.55  d x 0.1 + 4  0.55  1.02

L = = 4400 d + 2200
0.001
1st estimate:
L = 150 m  d = 2.72 m
 57
L = 4400  2.72 + 2200 = 14 168 < 22 500
2
too wide.

2nd estimate:
L = 100 m  d = 2.60 m
L2 = 4400  2.60 + 2200 = 13 640 > 10 000 too narrow.

3rd estimate:
L = 120 m  d = 2.60 + 2/5 (2.72 - 2.60) = 2.65 m
L2 = 4400  2.65 + 2200 + 13 851  14 400 (= 1202)
Thus the drain spacing is 120 m (Excel gives 117 m)

5) All assumptions and calculations are summarised below:

Equation Hydraulic conductivity (m/d)

Spacing (m)
Top layer Avge Bottom
Calc. (Excel)
Hooghoudt 0.06 - 0.30
79 (78)
Simplified
Hooghoudt - -
0.30 78 (77)
Hooghoudt - 0.14 -
55 (56)
Hooghoudt - 0.55 -
120 (117)
Publication 16, Ernst 0.06 -
Figure 8.7
0.30 77 (76)
We can conclude that:
 For the given soil profile, with its distinctive different
hydraulic conductivity above and below drain level, we
should use a drain-space Equation that can cope with these
 58
two layers, otherwise the calculated drain spacing is
unrealistic. In this particular case, the interface between the
two layers is at drain level, thus we can use Hooghoudt. We
can even use the simplified Hooghoudt equation, because
the majority of the flow takes place below drain level
because of the low hydraulic conductivity of the top layer
(Kt = 0.06m/d). If the drains are not installed at the
interface of the soil layers, we cannot use Hooghoudt and
we should use Ernst (see next question). Note that in this
case both Hooghoudt and Ernst give the same result. As
Hooghoudt requires less input data than Ernst, we prefer
Hooghoudt.
 The simplification of the soil profile plays a major role: If
we assume a homogeneous soil profile (K = 0.14 m/d), we
underestimate the effect of the bottom layer (Kb = 0.3 m/d),
where the majority of the flow takes place (remember the
viscous fluid model, and Problem 3) which results in an
expensive system (= narrow spacing);
 The selection of a representative value of the hydraulic
conductivity also has a major effect on the final drain
spacing: the arithmetic mean gives a spacing of 120 m and
the geometric mean a spacing of 55 m.
 59
6) In this particular case, the drain level does not coincide
with the interface between the two soil layers, thus we have
to apply the Ernst Equation. The drains are situated in the
top layer, thus we can use Equation 8.23

 Dv L2 L aD 
hq   ln r 
 K t 8(K b D b  K t D t ) πK t u 
where
q = 1 mm/d = 0.001 m/d
Dv = h = 1.5 - 1.0 = 0.5 m
See also Publication Kt = 0.06 m/d
16, Section 12.4.5 Kb = 0.30 m/d
Db = 3.0 m
Dr = 0.5 m
Dt = Dr + h/2 = 0.5 + 0.5/2 = 0.75 m

The geometry factor a can be obtained from Publication 16,

Table 8.2
Kb = 0.30 = 5 and Db = 3.0 = 6
Kt 0.06 Dr 0.50
Publication 16,
Section 3.4.2
So that from Table 8.2  a = 4.6

The wet perimeter is (Equation 8.15)

u = b + 2 ro = 0.25 + 2  0.10 = 0.45 m

Inserting all data in Equation 8.23 gives

 0.5 L
2

0.5 = 0.001  +  +
 0.06 8 (0.30  3.0 + 0.06  0.75) 
 L 4.6  0.5 
+ 0.001  ln 
   0.06 0.45 

L2 + 65.4 L - 3717 = 0
or L = 36 m
 60
Although unrealistic, we can still use Ernst to compute L for a
homogeneous profile, taking Kb = Kt = 0.14 m/d, and use
appropriate values for the various depths. The result will be L =
37 m.

If we have a homogeneous profile, we use Hooghoudt, of course.

Inserting the appropriate values, and using one K-value (= 0.14
m/d), we find approximately 38 m as the drain spacing.

Again the conclusion is that there is hardly any difference

between using the Ernst equation for the two-layered soil and
using a simplified (homogeneous) Hooghoudt equation.
Although this may not be true in all cases, very often there is
little difference between using Hooghoudt and using Ernst, so
that we select the simpler Hooghoudt equation (especially if Kb
and Kt are not too different).

Publication 16, 7) The drainage pore space is per definition

Section 8.3.1
change in amount of soil moisture storage
=
change in the watertable
15 mm
= = 0.05
0.30 m

The instantaneous recharge will be

Ri = 20 days  1 mm/d = 20 mm

8) We shall use the Glover-Dumm equation, because we have an

unsteady-state condition (Equation 8.32)

ht = 1.16 ho e-t

We can only use this equation for a homogeneous soil profile

K = K* = 0.14 m/d
 61
The calculated drain spacing for this situation is found to be (in a
similar way as in Problem 8) to be 48 m and the equivalent depth
is 2.20 m. Thus

 2 Kd  2  0.14  2.20
   0.0267
L2 0.05  482

Each irrigation will result in an instantaneous recharge of

20 mm causing a rise of the watertable
20
 h = Ri = = 400 mm = 0.40 m
 0.05

We assume that after the first irrigation (day 0) the water table
will rise to its maximum level, i.e. 0.75 m below soil surface,
thus

ho = hd - 0.75 = 2.00 - 0.75 = 1.25 m

With equation 8.32 we can calculate the depth of the watertable

after 20 days

h20 = 1.16 ho e-t

= 1.16  1.25 e-0.0267 x 20
= 0.85 m

The drop in the watertable in this 20-days period (1.25 - 0.85 =

0.40 m) is exactly what we want to avoid a gradual rise of the
watertable over the season.

Note that if you take a shortcut, and use the data from question
4b to calculate  (L = 55 m and d = 2.33 m), you get  = 0.021
and h20 becomes 0.95 m. The drop after 20 days is then only 0.3
m and is not sufficient to compensate for the rise caused by the
instantaneous recharge, thus the average watertable will
gradually rise during the irrigation season (Figure 14).
 62

Figure 14 Change of the watertable during the irrigation season

10) The simplified Hooghoudt equation, assuming only flow below

Publication 16, drain level, reads (Equation 8.6)
Section 8.4
8 Kdh 8 Kdh
q= 2
or L2 =
L q

The reaction factor of the Glover-Dumm equation reads

(equation 8.29)

 2 Kd  2 Kd
=  L
2
=
 L2 
 63
Thus
8 Kdh  2 Kd
=
q 
or

= 
2
h
q 8
In the comparable steady-state situation (a homogeneous profile with K
= 0.14 m/d; question 4b), we had h = 1.0 m and q = 0.001 m/d, making
h/q = 1000.

In the above unsteady-state situation, we found that  = 0.0267, so that

the right-hand side of the last equation becomes 924, indicating that the
equation is approximately true.

Résumé
We can conclude that the analysis of the soil and hydrological
conditions needed to translate the complex drainage problem in to a
simplified “model” (= drainage equation) that can be use to calculate a
solution (= drain depth / spacing combination) is the most difficult part:
the simplification of the soil profile in layers of different hydraulic
conductivity and the selection of a representative value of the hydraulic
conductivity for each of these layers (arithmetic or geometric mean) has
a major effect on the calculated drain spacing. Once we have made the
correct analysis, the selection of the appropriate equation is rather
straightforward and the calculation of the required drain spacing
relatively easy, especially when we use spreadsheet.
 64

12 ANSWERS TO THE SHORT TEST

Publication 16 1 The three components of the total energy head are the velocity
Section 7.3.4 head, the pressure head and the elevation head.
Because velocities of groundwater are low the velocity head can
usually be neglected.

2 We need the Dupuit-Forchheimer assumptions because most

Publication 16, groundwater flow problems have the free watertable as a
Section 7.8.1 boundary condition. Thus this free watertable, of which the
elevation is unknown, is both the boundary conditions of, and
the solution to, the drainage problem.

3 In the soil water flowing towards drains meets (1) external

resistance from friction with the walls of the pores and (2)
internal resistance due to viscosity. In the model this resistance
ILRI Bulletin 10 is simulated by (1) replacing the flow through the pores by flow
between two narrowly spaced parallel plates and (2) by replacing
the water with an oil of higher viscosity. A lower hydraulic
conductivity can be simulated by using a fluid with a higher
viscosity or by placing the plates closer together.

4 Hooghoudt accounted for the extra head caused by the radial

flow by the introduction of the equivalent depth theory. He
Publication 16, reduced the two-dimensional flow problem to an one-
Section 8.2.1 dimensional flow problem by the introduction of an imaginary
impervious layer above the real one and by replacing the drains
with imaginary ditches with their bottoms on the imaginary
layer.

5 L2 is not only proportional to 1/q, but also depending on the

 65
equivalent depth d, which in turn is a function of L. This is the
Publication 16, reason that when q is taken 4 times larger, L does not become
Section 8.2.1 half.
For example see Problem 6: when we use the exact solution with
q = 8 mm/d, we find L8 = 38 m. However, for q = 4 x 8 = 32
mm/d, we find L32 = 16 m ¹ 19 m (=½L8).

6 In Problem 6 the depth of the drains coincides with the interface

between the two soil layers, thus we can use Equation 8.7. In
Publication 16, Problem 7 the drains are situated in the lower layer of a two-
Section 8.2.3 layered soil profile, thus we have to use the Ernst Equation.

7 In irrigated areas, steady-state equations can be used because the

relation between the level of the watertable and the agricultural
criterion (e.g. yield) is often considered over a long period (e.g. a
Publication 16, season). In this period the change in water level h is small
Section 17.3.4 compared to the recharge and the discharge, thus the water
balance can be simplified to q = R and the watertable can be
considered at a dynamic equilibrium.

8 The steady-state criterion is the watertable-discharge criterion

q/h and the unsteady-state criterion is based on the rate of
watertable drawdown h0/ht. Combining the Hooghoudt Equation
for flow below drain level (Equation 8.6) and the equation for
Publication 16, the reaction factor  (Equation 8.29) yields a relation between
Section 8.4 these two criteria

= 
2
h
q 8

9 We shall apply the exact solution of the Hooghoudt Equation

using the Excel spreadsheet program. If we do only consider the
radius of the drain (including the envelope) and not the drain
trench, the we have the following data:
 66
· drain depth 1.0 m (-s.s)
Publication 16, · allowed water table 0.5 m (-s.s)
Section 8.2.1 and · aquiclude depth 7.0 m (-s.s)
Excel exercise · upper conductivity 1.5 m/d
· lower conductivity 1.5 m/d
· steady recharge 7 mm/d
· drain radius 0.067 m

and the calculated drain spacing is 53 m.

If we also consider the drain trench, the wet perimeter becomes

(equation 8.16):

u = b + 2(2r0 + m)
= 0.20 + 2(2  0.06 + 0.007) = 0.454 m

and the equivalent radius (Equation 8.14):

u 0.454
ro = = = 0.145 m
 
and the calculated drain spacing is 57 m.

Hence, considering the drain trench increases the drain spacing

(and thus reduces cost) by 7.5%.