Groundwater

An Introduction to Hydrology and Water Resources Engineering in Uganda
14. GROUNDWATER
14.1 Introduction
Groundwater can be defined as: subsurface water that occurs beneath the water table in soils and
geological formations that are fully saturated or subsurface water at or greater than atmospheric pressure.
But the term groundwater is sometimes used to imply the subsurface water regime in a broader sense to
include:
i) The near-surface, unsaturated soil moisture zone
ii) The deeper saturated zones that influence geological processes (e.g. earthquakes) and whose
water may no longer be part of the hydrological cycle.
I. Characteristics
i) Groundwater is an areally distributed resource in contrast to surface water the wide
distribution of groundwater often makes it more readily available for development.
ii) Time scale: Groundwater is a slow response, low velocity medium. Even small aquifers have
time constants in the order of weeks or months. Very large aquifers may have time constants
of many years.
iii) Variability of flow and supply is therefore small: A groundwater aquifer behaves as a large
natural reservoir having a heavily damped response. In this sense, Groundwater supplies are
often independent of droughts. In contrast, surface water has much higher variability and
requires Hydraulic Structures to create artificial storage.
iv) Energy is usually required to lift groundwater. Pumping costs may form a substantial
proportion of the cost of delivering the water. In general, ‘Operating costs’ tend to be higher
for groundwater that for surface water from which gravity is often the natural delivering
force.
v) Economics: Groundwater, however, usually requires relatively low capital investment in
contrast; capital costs for surface water are often very high when dams and other hydraulic
structures are needed. For this reason, groundwater sources are often suitable for staged
development. The marginal costs for extra capacity are relatively low.
vi) Quality
Groundwater Surface Water
a) Virtually free of pathogenic organisms; Often easily contaminated

little or no treatment needed
b) Effectively self purifying; soil or Treatment usually necessary

ground matrix is a good
biological/chemical and physical filter
c) Dissolved salts content can be high; Dissolved salts content usually much
water is mineralized by rocks through lower
which it moves and may not be easy to
treat.
d) If groundwater does get polluted, it can Pollution effects more readily treated
be difficult to clean up
e) Relatively constant temperature Variable temperature; can vary

diurnally
398
f) Colour, turbidity good Colour, turbidity may change readily,

e.g. after a storm
vii) Data: Field data for modelling or analysis in groundwater are usually scarce, often at a few points
only. Extensive interpolation and inference may be necessary, based on experience and judgment.
viii) Legal: An example of the water rights is available in the Uganda Water Act 1997, Section 7 (1)
where it states : A person while temporarily at a place or being the occupier of or resident on any
land, where there is a natural source of water, may use that water for domestic use fire fighting or
irrigating a subsistence garden (2) In addition to the above, the occupier of land or resident on land
may, with approval of the authority responsible for the area, use any water occupied under the land
occupied by him or her on which he or she is resident on or any land adjacent to that land. The
above rights do not permit/authorize a person to construct any works (Johnston, 1983).
II. Aquifers
Refer to Figs 14.1a and 14.1b
i) Aquifer: A saturated permeable geological unit which can transmit significant
quantities of water under ordinary hydraulic gradients.
ii) Aquiclude: A saturated geological unit that is incapable of transmitting
(aquifuge) significant quantities of water under ordinary hydraulic gradients e.g. granite clay.
iii) Aquitard: A ‘less permeable’ bed in stratigraphic sequence (may be significant
in a regional groundwater balance but insignificant for a local production well).
N.B.: These terms are relative
Aquifers take two basic forms
iv) Confined Aquifer: An aquifer confined between two aquitards
Fig 14.1a A perched water table.
399
v) Artesian Aquifer: or aquiclude or vice versa such that pressures in an aquifer are everywhere greater
than atmospheric.
The recharge is from rainfall in the area, where the aquifer forms outcrops or intersects the land surface.
From slow flow or ‘leakage’ through a confining aquitard, the aquifer is then termed a leaky confined
aquifer and the aquitard, a leaky confining layer.
vii) Unconfined Aquifer: an aquifer with a free water surface as the upper boundary
viii) Water table aquifer or phreatic surface, at which pressure is atmospheric. The recharge percolation is
from rainfall in the immediate area.
Note: The terms confined and unconfined relate to the hydraulic characteristics of the Aquifer, not to the
geological formation itself. Particular water bearing geological formation may be confined in one zone
and unconfined in another as shown in Fig 14.1b.
ix) Perched Aquifer: A saturated lens of relatively low permeability bounded by a perched water table as
shown in Fig 14.1a. Perched Aquifers can provide minor sources of supply but can suffer rapid changes in
water level since the storage involved is usually relatively small. However, unconsolidated sediments
(e.g. sands, silts, clays) are commonly intergraded, making the determination of the true water table
sometimes very difficult in practice.
CONFINED UNCONFINED
PIEZOMETRIC
LEVEL
SPRING
CLAY
AQUIFER
CLAY
Fig. 14.1b An confined aquifer adjacent to an unconfined aquifer and the water table.
x) Piezometric surface or potentiometric surface:

The piezometric surface or potentiometric surface of a confined aquifer is an imaginary surface
coinciding with the hydrostatic pressure level of the water in the aquifer. The water level in a well
penetrating a Confined aquifer defines the elevation of the piezometric surface at that point. If the
piezometric surface is above ground a flowing well results. A confined aquifer becomes an unconfined
aquifer when the piezometric surface falls below the upper confining bed.
400
14.1.1 The Groundwater Column

The groundwater column consists of three main zones, namely; i) the zone of aeration, ii) the capillary
water zone and iii) the groundwater zone. On most of the land masses of the earth, a single zone of
aeration overlies a groundwater zone as shown in Fig 14.2.
Fig 14.2 The Groundwater column.
i) Zone of Aeration
In the zone of aeration vadose (shallow) water occurs. The zone of aeration consists of interstices
occupied partially by water and partially by air. The general zone may be further subdivided in
descending order into; soil water pendular zone, intermediate vadose (funicular) zone and the capillary
zone or ‘fringe’.
In the Pendular Zone the liquid is held in isolated rings around the grains, hygroscopic water. This water
is held by surface tension forces and is referred to as capillary water. This zone extends from the ground
surface down through the major root zone and is particularly important to soil scientists and
agriculturalists, because it is the source of crop moisture. Water in the soil water zone exists at less than
saturation except when the water reached the surface as from rainfall or irrigation. The thickness of the
zone varies with soil type and vegetation. The amount of water is dependent upon the recent exposure of
the soil to moisture (antecedent conditions).Under hot arid conditions; water vapour equilibrium tends to
become established between the ambient air and surfaces of the grained soil particles. In the intermediate
vadose (funicular) zone the liquid is continuous through the pore space and entirely encloses the vapour
phase, pellicular water. This zone extends from the lower edge of the soil water to the upper limit of the
capillary zone. The thickness may vary from zero where the bounding zones emerge with high water table
approaching ground surface to more than 100 m under deep water table conditions. Non moving vadose
water is held in place by hygroscopic and capillary forces. Temporary excesses of water migrate
downward as gravitational water.
ii) Capillary Zone

In the capillary zone or fringe: there is full saturation with pore water pressure less than atmospheric
pressure. The capillary zone (or capillary fringe) extends from the water table to the limit of capillary rise
of water.
401
Water table: That surface in a groundwater body at which the water pressure is atmospheric. This is
defined as the surface of atmospheric pressure and appears as the level at, which water stands in a well
penetrating the aquifer.
iii) Groundwater (Saturated) Zone

In the groundwater zone all interstices are saturated and filled with water at hydrostatic pressure. The
groundwater zone extends from the upper surface of saturation down to underlying impermeable rock. In
the absence of overlying impermeable strata the water table or phreatic surface forms the upper surface of
the zone of saturation.
Saturation actually extends slightly above the water table, due to capillary attraction, water is held here at
less than atmospheric pressure. Water occurring in the zone of saturation is termed groundwater or
phreatic water.
In the saturated zone, groundwater fills all the interstices, hence effective porosity provides a direct
measure of water per unit volume. A portion of the water can be removed from subsurface strata by
drainage or pumping. However molecular and tension forces hold the remainder of the water in place.
Specific Retention of rock and soil Sr is the ratio of the volume of water it will retain after saturation
against the force of gravity to its own volume.
Sr = Wr (14.1)
V
where: Wr is the volume occupied by retained water and V is the bulk volume of soil or rock
Specific Yield of soil or rock Sy is the ratio of the volume of water that after saturation can be drained by
gravity to its own volume.
Sy= Wy/V (14.2)
where: Wy is the volume of water drained. Values of Sr and Sy can be expressed as percentages. Because
Wr and Wy constitute the total volume of water saturated material,
α =Wr + Wy where all pores are interconnecting (14.3)
Values of specific yield depend on grain size, shape and distribution of pores compaction of the stratum
and time of drainage.
Fined grained materials yield little water whereas coarse grained materials permit a substantial release of
water hence serve as aquifers. Specific yield can be measured by a variety of techniques including
laboratory – field and estimating techniques. Methods based on well pumping tests give the most reliable
results. The usual range of values is 0.01 – 0.30.
Storage Coefficient or Storativity is defined as the volume of water that an aquifer releases from or takes
into storage per unit surface area of aquifer per unit change in the component of head normal to that
surface. The storage coefficient for an unconfined aquifer corresponds to its specific yield (Todd, 1980).
The of range of values in confined aquifers is from 0.005 to 0.00005
402
14.2 The Occurrence of Groundwater

Rain that falls on to the ground surface and infiltrates eventually percolates through the voids of the rock
and therefore it is inevitable that the mode of occurrence of groundwater depends largely on the type of
formation (geology) of the area.
14.2.1 Sources of Groundwater

a) Types of Groundwater
i. Meteoric-this is water that is in the atmosphere or has recently been in the atmosphere for example
oceanic water. It is recharged into the ground from rain and snow.
ii. Connate-this is water that has been out of contact with the atmosphere for at least an appreciable
part of the geologic period and has been trapped in sediments during deposition.
iii. Juvenile-this is new water that has not been previously part of the hydrosphere and is created by
volcanic activity.
It is necessary to have a clear understanding of how water occurs in a particular rock before any water,
which may be contained in that rock can be abstracted efficiently. Below is a summary of the major rock
types and the way water exists in them.
b) Rock types as aquifers
i) Igneous Rocks
They are generally poor aquifers in solid form because they are very compact, dense and hence relatively
impermeable. Where such rocks occur, near the surface under weathered conditions, they have been
developed by small wells for domestic water supply.
.
Intrusive: Such as granite, gabbro and dolerite.
Fresh rock – almost no water exists. The little water there exists in fissures.
Weathered rock – water in weathered zones and in fissures opened up by weathering. Usually only
moderate supplies are obtained. The fissures are mainly joints produced by cooling stressed or tectonic
forces.
Extrusive: All lavas

Fresh rock – little water except in porous zones which may exist near the surface of each lava flow and
caused by the escape of gas. At times the pores do not interconnect and so contain no water. Basalt can
form highly permeable aquifers. These are a result of:
i) interstitial spaces in clinkery lava at the tops of flows,
ii) cavities between adjacent lava beds. iii) shrinkage cracks, lava tubes, gas vesicles, fissures, resulting
from faulting and cracking and holes left by burning trees overwhelmed by lava. The values of hydraulic
conductivity range from 10-13 to 10-10 m/s for unfractured rocks and from 10-6 to 10-2 m/s for fractured
rocks.
ii) Sedimentary rocks

Sand and gravel: Water contained in pore spaces. In gravels the pore spaces may be so large that no
filtering of the water can take place so the water may contain bacteria. Very large yields are possible. The
values of hydraulic conductivity range from 10-4 to 102 m/s.
Sandstone and conglomerate are cemented forms of sand and gravel. As such their porosity and yield
have been reduced by cement. The actual open volume of the pore space depends upon the grain size and
distribution, grain shape and percentage of cementing material. The best sandstone aquifers yield waters
through their pores and fissures. Frequently large to medium volumes of water are obtained.
403
Conglomerates have limited distribution and are unimportant as aquifers. The values of hydraulic
conductivity range from 10-7 to 10-3 m/s.
Limestone: Limestone rock is impermeable so all water is contained in fissures, which because of the
solubility of the rock in slightly acid water are frequently very large. Caves containing large streams of
rivers are not uncommon. Because of the very large size of the fissures there is no filtering of bacteria and
any pollution spreads very rapidly throughout the extent of aquifer. Limestone rock varies widely in
density, porosity and permeability, depending on the degree of consolidation and development of
permeable zones after deposition. Those most important as aquifers contain sizeable proportions of the
original rock that have been dissolved and removed.
Openings in limestone range from microscopic original pores to large solution caverns (karstic systems)
forming subterranean channels to carry entire flow of stream. Streams disappear underground in
limestone terrain and large springs are found in limestone areas. By dissolving rock, water tends to
increase the pore space and permeability with time. Values of hydraulic conductivity range from 10-7 to 1
m/s.
Clays: In clays the porosity is high but the effective porosity is almost zero. Clay and coarse materials
with clay are generally porous but the pores are so small, that they may be regarded as relatively
impermeable. Clayey soils can provide small domestic water supplies from shallow, large diameter wells.
Values of hydraulic conductivity range from 10-10 to 10-6 m/s.
Shales: Water may sometimes exist within the laminations and flow parallel to them. The permeability
parallel to the laminations may thus be moderate whilst the permeability at right angles may be zero.
iii) Metamorphic rocks

There are various types but typically found as the following:
Gneiss: This has very similar hydrological properties to intrusive igneous rocks.
Schist : These have similar water bearing properties to gneiss except that permeability parallel to the
planes of schistocity is sometime moderate.
Quartzite: These only contain water in fissures and since the rock is very resistant to weathering fissures
are frequently tight.
The values of hydraulic conductivity are similar to those of igneous rocks and typically range from 10-13
to 10-10 m/s for unfractured rocks and from 10-6 to 10-2 m/s for fractured rocks.
iv) Superficial deposits

Laterite: this is a weathering product of various geologic materials. These can be very porous and so
produce large volumes of water like sand, gravels and clays.
v) Geological structures
Faults: These, if they are filled with clay material can act as subsurface dams producing quite marked
changes of water level on either side. In other cases when they are open they act subsurface flow paths for
water and can produce mixing of water from different aquifers which are normally separated by
imperious layers.
Folds: The act of folding often produces associated jointing in rocks so increasing their water containing
capacity (Blyth and de Freitas, 1980).
404
14.2.2 Groundwater Occurrence in Uganda

Groundwater occurrence in Uganda (Tindimugaya, 2005) is greatly influenced by the varied geological
conditions, which consist of very old precambrian rocks that underlie over 90% of the country and
cenozoic rift valley sediments and tertiary and pleistocene volcanics that occur in a few areas and cover
less than 10% of the country. Hydrogeological conditions are typical of Precambrian basement terrain and
aquifers occur in the weathered overburden (regolith) and in the fractured bedrock, forming one
discontinuous aquifer system. In most of Uganda, boreholes are typically drilled into fractured bedrock, while
shallow wells are drilled in the regolith. In-situ weathering is the primary process in overburden
development, and weathering is usually most intense in fractured rocks. A conceptual model of the
weathered –fractured basement aquifer system is shown in Fig.14.3.
Fig 14.3 Conceptual model of the weathered – fractured aquifer system in Uganda.
Source (Tindimugaya, 2005)
Related studies (Taylor and Howard, 1999, 2000) show that the hydrogeological characteristics of deeply
weathered crystalline rock derive from and are hence related to long term tectonically controlled
geomorphic processes. Permeability in the bedrock arises from fractures, which are attributed to the
decompression resulting from the removal of overlying rock in solution (deep weathering) and by
colluvial and fluvial erosion (stripping). The overlying unconsolidated mantle is a product of deep, in-situ
weathering. Sand sized clasts predominate at the base of the mantle and form an aquifer that is an order of
magnitude (T = 5-20 m2/d) higher than underlying bedrock fractures (T~ 1m2/d).
405
It has also been suggested (Bradley et al., 2008) that the corridors of preferential groundwater flow occur
in relict river channels like the Katonga river valley, reversed by tectonic activity associated with the
development of the East African Rift System (EARS). In the mid- Pleistocene, uplift parallel to the rift
escarpment exceeded the rate of river basin incision and led to the reversal of the surface flow. On
surfaces of low relief east of the upwarp, coarse grained sediments in these channels were overlain by low
energy fine grained deposits that currently support extensive wetlands. The research is ongoing and
intends to examine the relict river channels in the Katonga drainage basin, with the aim of understanding
the groundwater resource potential and facilitate the management of wetlands.
14.3 Water Movement in Soils (Saturated Flow)
14.3.1 Steady Flow of a Perfect Fluid

Consider a control volume (incompressible, no energy losses, velocity and pressure functions of position
only) as shown in Fig 14.4.
V2t P2
V2
V1t
V1
P
Z2
Z1
Datum
Fig.14.4 A control volume with steady flow.
From the continuity equation, q = av = constant, where a is the cross sectional area and v is the velocity
at the point of interest
a1v1=a2v2=q (14.4)
The work (Force x Distance) input in time t by p1
p1a1v1δt = p1q1δt (14.5)
The work expended in time t against p2
p2a2v2δt = p2q2δt
( pa1v1t ).g.( z2  z1 )   gqt ( z2  z1 ) increase in potential energy (mgΔh) (14.6)
1
2
 a1v1t v22  v12   1 2  q t v22  v12  increase in kinetic energy (14.7)
Equating work input in δt by P1 to work expended in δt against P2
406
p1  p2   g ( z2  z1 )  1  (v22  v12 )
2
p1 v12 p2 v22
or z   z  = h a constant (14.8)
 g 1 2g  g 2 2g
for a viscous fluid

A term must be added to account for the work done against viscous forces, that is,
p1 v2 p v2
 z1  1 = 2 +z2 + 2 +viscousheadloss (14.9)
g 2g ρg 2g
For seepage flows, water velocities of about 1cm/sec are large hence the term v2/g is negligible
v2 0.012
i.e. = = 5.1x10-6  negligible with respect to z and p/g
2g 2x9.81
Then it follows that
p1 p
+ z1 = 2 + z 2 + Viscous head loss (14.10)
ρg ρg
Viscosity
The fluid shear strain rate is given by;
(dx) (dx) 1 dv
 
(dy)/dt dt dy dy
This shear strain rate is caused by a viscous shear stress 
The coefficient of viscosity  defined as
viscous shear stess 
μ= 
fluid shear strain rate dv
dy
dv
Hence  =  (14.11)
dy
14.3.2 Viscous Flow between Parallel Plates
The average velocity of flow between parallel plates separated by a distance 2 is given as:
_
δ 2 ρg dh
V =
3 μ ds
407
dh
and is a measure of the viscous head loss
ds
Note vmax occurs at y = 0 at the central position:
 2  g dh
v max =
2  ds
_ (14.12)
V 2
=
vmax 3
and discharge per unit width:
_
2 ρg dh
Q = V. 2 = . 3 . . (14.13)
3 μ ds
14.3.3 Viscous Flow through a Tube

The average velocity of flow through a tube or radius R is given by:
_
R 2 ρg dh
V =
8 μ ds
(14.14)
_
 R 4 ρg dh
 Q = VA = V . R = 2
4 μ ds
Normalise by introducing mean hydraulic radius RH

Flow Area
RH =
Wetted Perimeter
for parallel plates for a tube

2 1 πR2 R
RH = = RH = =
2 2πR 2
1 ρg dh 1 ρg dh
v= RH2 v= RH2
3 μ ds 2 μ ds
i.e. v = shear factor x size factor x fluid factor x driving force (Wheater, 1984)
So for any shape or cross section, we can write:

2 ρg dh
v = CsR H (14.15)
μ ds
408
pore area
For soil pores, R H =
pore perimeter
If the porosity is n, for a given cross sectional area A of material.

Pore area = n. A. If the pore perimeter is P
nA
Then, R H =
P
Consider two sections, ds apart.
Pore Volume nAds
Surface area of grains  Pds
nAds Pore volume e. Vs
Then = = = R H , where e is the void ratio
Pds Surface area of grains As
Substitute for RH in equation 14.12,
 e Vs  ρg dh
2
_
So V = Cs  
 A s  μ ds
But we cannot measure the pore velocity. Only the total discharge flow rate Q and the total sample area A
and its porosity can be measured.
So define Discharge Velocity = Q = V

A
_
n
and V = n V , void ratio e =
1-n
 n Vs  ρg dh
2
So V = nCs  ( )  (14.16)
 1-n A s  μ ds
Now for spherical soil grains of diameter Ds:
Vs 1/6 D3s Ds
= =
As D 2
s
6
Substitute for Vs/As in equation (14.12):
 n Ds  ρg dh
2
 V = nCs  ( ) 
 1-n 6  μ ds
(14.17)
 C D2 ρg n3  dh
=  s s 
 36 μ (1-n)2  ds
409
dh
or V = - K
ds (14.18)
This is Darcy’s Law and K is the hydraulic conductivity. The negative sign indicates that flow is in the
direction of increasing head.
ρg
If K = k
μ
n3
k = CD2s Kozeny – Carman Formula (14.18a)
(1-n)2
k is the coefficient of permeability
In practice the structure of porous media is more complex than assumed in the Kozeny Carman
formulation and Darcy’s law is normally considered empirically as in the original experimental
derivation. Table 14.1 shows the typical range of values of K.
Table 14.1 Typical range of values of K
Soil Type K (m/sec)

Clean gravel  10-2
Sand 10-6 – 10-2
Silt 10-9 – 10-4
Clay  10-7
Source: (Freeze and Cherry 1979)
14.4 Permeability
14.4.1 Intrinsic Permeability

The permeability of a rock or soil defines its ability to transmit a fluid. This is a property of the medium
only and is independent of fluid properties. To avoid confusion with hydraulic conductivity which
included the properties of groundwater, and intrinsic permeability k may be expressed as
Kμ
k= (14.19)
g
Where K is hydraulic conductivity,  is dynamic viscosity,  is fluid density, g is acceleration of gravity.
Inserting Darcy’s equation into the above.
dh
V = -K  K= - V/(dh/ds)
ds
μ V
 k=- which has units of
ρg (dh/ds)
410
- (kg/ms)(m/s)
k= 3 2
= m2
(kg/m )(m/s )(m/m)
The negative sign indicates that the flow is in the direction of decreasing head.
The intrinsic permeability possesses units of area. Because values of k are so small it is often expressed
as square micrometers (m)2 = 10-12 m2.. In the petroleum industry, the value of k is measured by a unit
termed Darcy.
1 Darcy= 0.987 (m)2
which closely corresponds to the square micrometer unit.
The intrinsic permeability can be estimated from gain size distribution data and from the Fair Hatch
(1933) formula as reported by Todd (1980).
1
1  (1-n)2  B 
2
F  V
k=  3  100  d  (dh/ds)  (14.20)
A   m 
 n 
Where n is the porosity, A is a dimensionless packing factor found to be approximately 5 and B is a
particle shape factor equal to 6 for spherical particles and 7.7 for highly angular ones. The factor F is the
percent by weight of the sample between two arbitrary particle sizes and dm, is the geometric mean of the
particle sizes corresponding to F. the Fair Hatch formula takes account of the distribution in grain sizes.
Other equations require only a characteristic grain size d.
For example the formula k = (6.54 x 10-4 )d2 (14.21)
yields the intrinsic permeability in cm2 as a function of a characteristics particle diameter in (Harleman et.
al 1963). The above equation is nearly valid for materials of very uniform particle size and shape.
14.4.2 `Hydraulic Conductivity

For practical work in groundwater hydrology, where water is the prevailing fluid, hydraulic conductivity
is employed. A medium has a unit hydraulic conductivity if it will transmit in a unit time a unit volume of
groundwater at the prevailing kinematic viscosity through a cross section of unit area at right angles to
the direct of flow, under a unit hydraulic gradient. From Darcy’s law, the units are:
V m/day
K=- = = m/day
dh/ds m/m
14.4.3 Transmissivity
This term is widely employed in groundwater hydraulics. It is the rate at which water of prevailing
kinematic viscosity is transmitted through a unit width of aquifer under a unit hydraulic head. (It is equal
to the discharge rate at which water is transmitted through a unit width of an aquifer under a unit
hydraulic gradient)
2
T = Kb (m/day)(m) = m /day (14.22)
Where b is equal to the saturated thickness of an aquifer it is equal to the depth of a confined aquifer and
equal to the average thickness of the saturated zone of an unconfined aquifer.
411
14.4.4 Methods of Measuring Permeability

i) Constant Head Permeameter (coarse-grained soils, sands and silty sands)
Water under a constant head is allowed to percolate through a sample contained in a cylinder of cross
sectional area A. The quantity of water passing through the sample in time t is collected in a measuring
cylinder to give Q the flow rate. Manometers tapped into the side of the sample cylinder give the loss of
head H over a sample length l and hence the hydraulic gradient i.
Inlet
Waste
h1
Constant
head H
h2
Filter
Overall length L
Sample
Q c.s.a:A x Length
of
sample L
perforated disc
Measuring and wire gauze
cylinder
Filter
Fig 14.5 Constant Head Permeameter.

Source: (Todd, 1980)
From Darcy’s Law, V = Ki But Q = VA

Q
 = Ki
A
H
-i = where H  h1  h2
L
Q L
K=
A H
The above apparatus in Fig.14.5 is known as a constant head permeameter. The water may be arranged to
flow up the sample as shown but some permeameter permit downward flow with the same principles
applying. A sand filter is incorporated above and below the sample to help prevent it from washing away.
Note:
‘Quick’ Condition (Piping)
This condition occurs when the pore pressure becomes equal to the total stress (vertical). There is then no
effective contact between soil particles.
412
Consider soil at depth x

Total Stress = P2 + ρT gx T = B (bulk density) (14.23)
Pore Pressure  P2 + P1 - P2 .x (14.24)

L
When Equation. 14.23 equals Equation 14.24 P1 = P2 = Tg (14.25)

This occurs at icrit
h2 - h1  P2 P1  P1 - P2
icrit = =  +L-  /L = 1- (14.26)
L  ρwg ρwg  ρ w gL
i.e. from Equation.14.25 and Equation.14.26

T g
icrit =1 -
w g
T
-icrit = -1
w
Most sands have T close to 2w -icrit  1

ii)Falling head permeameter (fine-grained soils silts and silty clays)
Water flows through fine-grained soils at a much slower rate than through coarse material, consequently,
it is not possible to obtain a measurable amount of water within a reasonable time. In this case a falling
head or variable head permeameter is used as shown in the Fig 14.6.
When the stopcock is opened, water will pass through the sample and the level in the standpipe will fall.
Once steady conditions have been obtained two readings of H are taken H1 and H2 at a time interval t.
Now during any increment of dt the variation in head is -dH and hence the quantity of water flowing
through the sample in dt, Q = -adH.
H1
t Standpipe
H2
H3
t c.s.a. s
Variable
head
H
Length L
Sample
c.s.a. A 100mm dia.
cutting ring
perforated disc
and wire gauze
Fig 14.6 Falling Head Permeameter.
413
dH
Flow = a. in the pipe
dt
A-H Darcy’s Law; v = -ki
= k through the soil
L
 q = Av = -Aki
Continuity
adH AHK
=-
dt L
Separating variables and integrating between limits
H2 t2
dH AK
H1
 H
=-
aL  dt
t1
H1 AK
ln =- (t 2 - t 1 )
H2 aL
So aL H  (14.27)
K= ln  1 
A(t 2 - t 1 )  H2 
Precautions
Minimize disturbance of soil sample by taking the sample in field in a coring device which can also serve
as a permeameter cell. Beware of gaps between sample and cell wall; a rubber sleeve can be used.
Used boiled water (de-gassed) air immigrates to larger pores and blocks flow. Match laboratory water to
field water quantity to avoid chemical effects particularly for clays.
Ensure sample is saturated, but note K can decrease for a soil subjected to prolonged flooding due to
growth of slimes.
iii) Auger Hole Method
The auger hole method involves the measurement of the change in water level after rapid removal of a
volume of water from an unlined cylindrical hole as Fig 14.7. If the soil is loose, a screen may be
necessary to maintain the hole. The method is relatively simple and is most adaptable to shallow water
table conditions. The value of K obtained is essentially that for a horizontal direction in the immediate
vicinity of the hole
dv
K = C. 14.28)
dt
dv
where is the rate of rise of water table at depth and C is a function of r, s, H, y, where r is the radius
dt
of the hole, s is the depth from the bottom of the hole to the impermeable layer, H is the depth from the
water table to the bottom of the hole and y is the height from the water table to the level of water in the
hole.
Various solutions have been presented by several researchers including Boast and Kirkham in 1971.
414
Ground surface Water table
2rw
impermeable or highly impermeable layer
Fig 14.7 Auger hole cross section.

iv) Pumping Tests of Wells

The most reliable method for estimating aquifer hydraulic conductivity is by pumping tests of wells.
Based on observations of water levels near pumping wells an integrated K value over a sizeable aquifer
section can be obtained. Furthermore, because the aquifer is not disturbed, the reliability of such
determinations is superior to laboratory methods. These are discussed in Sections 14.7 and 14.9.
Generally, if the stratum being tested is above the water table, water is pumped into the borehole and the
rate of flow to maintain a constant head is measured. If the stratum is below the water table, either
pumping in or pumping out tests may be used, in conjunction with a casing to the borehole extending into
the permeable stratum.
The U.S Bureau of Reclamation (1974) uses the following expression for cased boreholes.
q
K= (14.29)
5.5rh
where: r = radius of casing
h =the head of water maintained above water table
q = rate of flow of water into the borehole to maintain constant head above the water table
14.5 Heterogeneity and Anisotropy of Hydraulic Conductivity

An important extension of Darcy’s law to groundwater flow is its application in three dimensions (Freeze
and Cherry, 1979). Darcy’s law gives the specific discharge in the single x direction, and it is assumed
that the medium does not change in character so that K remains constant in that direction. However, the
ground structure may differ radically in other directions and therefore at a point in the ground three
specific discharges may be defined;
415
Vx = -K x dh/dx
Vy = -K y dh/dy
Vz = -K z dh/dz
(14.30)
with K having different values in the x, y and z directions.
Hydraulic conductivity values usually show variations;
 through space within a geologic formation. This property is termed homogeneity.
 with the direction of measurement at a given point in a geologic formation. This is termed
anisotropy.
Homogeneity and Heterogeneity

Most geological formations are not uniform in composition and the properties of that porous medium will
vary both in space and direction. Consider hydraulic conductivity, K;
Homogeneous: K is Independent of position in a geological formation
K (x, y, z)= C (a constant)
Heterogeneous: K is dependent on position within a geological formation
K (x, y, z)≠ C
14.5.1 Types of Heterogeneity Formations
a) Layered geological strata
A
K=103
K=102
A’
K=10
”
K=1
0
` x
103
102 x
10 x
A A’
Fig. 14.8 Layered geological strata

Source: (Cherry and Freeze,1979)
416
Layered heterogeneity is common in sedimentary rocks and unconsolidated lacustrine and marine
deposits. In this case, individual layers making up the formation each have a homogeneous conductivity
value K1,K2,…, but the entire system can be thought as heterogeneous.
b) Discontinuous heterogeneity: Faults are an example.
c) Trending heterogeneity: Trends are possible in any type of geologic formation, but they
are common in response to the sedimentation processes that create deltas, alluvial fans and glacial
outwash plains. Fig. 14.8 is a sketch map that shows a case of trending heterogeneity.
K may change 2-3 orders of magnitude in a few Km.
Isotropy and Anisotropy
Isotropic: K is Independent of direction of measurement in a geological formation.

Anisotropic: K is dependent on direction of measurement in a geological formation.
Anisotropy implies if K = K (θ), in a two-dimensional vertical section, θ is angle between the horizontal
and direction of movement.
If coordinate axes are set up in the principal directions Kx, Ky, Kz are the conductivities
Isotropic: Kz = K x = Ky Anisotropic: Kz ≠ K x ≠ Ky
Transversely Isotropic: A layered formation where hydraulic conductivity can be considered as

having only horizontal and vertical components
Kx= Ky # Kz (14.31)
14.5.2 Layered Heterogeneity and Anisotropy

If a layered formation can be considered as a sequence of discrete layers, as shown in Fig 14.9 each
homogeneous and isotropic with hydraulic conductivity Ki , the whole can be considered as a single
homogeneous anisotropic layer.
d1 K1 K1
z
d d2 K2
 Kx
x
dn
Fig. 14.9 Layered formation
- From continuity Vin = Vout (specific discharge through any layer)

- If head loss across each layer = ∆h1, ∆h2, …, ∆hn
Total head Loss= ∆h = ∆h1 + ∆h2 +…+ ∆hn
To establish the desired equivalence, from Darcy’s Law
417
K1Δh1 K 2 Δh2 Kn Δhn K z Δh

V= = = .... = =
d1 d2 dn d
Vd Vd
 Kz = =
Δh Δh1 + Δh2 + ... +Δhn
Vd (14.32)
=
Vd1 Vd2 Vd
+ + ... + n
K1 K2 Kn
d
Kz = n
di
K
i=1 i
Horizontal KX : Ifh = head loss over horizontal distance

Q = discharge/unit thickness of the system = Sum of discharge through the
Layers
Q n
kidi Δh Δh
V=
d
= 
i=1 d l
= Kx
Δl
(14.32a)
n
kidi
Kx =  i=1 d
Darcy’s Law in three dimensions

Vx = -K x h/x Kx, Ky, Kz: Hydraulic conductivity in x, y, z directions, respectively
Vy = -K y h/y Vx, Vy, Vz: Velocities / Specific discharges
Vz = -K z h/z
14.6 Factors affecting the Hydraulic Conductivity
Hydraulic conductivity K, is a measure of the ease with which water can flow through soil. K is
dependent on the pore fluid and soil type. For laminar flow, in groundwater flow in a given direction, the
flow follows Darcy’s Law as given in Equation 14.18.
When water is the pore fluid and its temperature range is not large, the hydraulic conductivity, K will
describe fully the permeability of the soil in a given flow direction.
K is influenced by other factors (Jackson and Dhir, 1988) that include;
i) Particle size, shape and grading;

ii) Arrangement of particles and stratification and fissures;
iii) Density of packing;
iv) Composition;
v) Consolidation pressure;
vi) Migration of particles; and
vii) Presence of air or gas.
418
i) Particle size, shape and grading

These factors determine the size, shape and continuity of the pore spaces in the soil. Therefore, they
determine the hydraulic conductivity K, since they affect the ease with which water can flow through soil.
In graded soils, K is mainly influenced by the finer sizes of soil present. From experimental work on filter
sands of medium permeability, Hazen showed that K could be estimated by:
K = D102 x 104 µms-1 (14.33a)

where: D10 is the effective size of the sand in millimetres. The expression above does not contain the
parameters describing either the density of packing or particle size distribution particles but it gives good
estimations.
ii) Arrangement of soil particles

The process of soil formation results in different variations of the soil structures that affect the
permeability. For example, alluvial soils tend to have a less resistance to horizontal flow of water along
the strata than to vertical flow. In uniform gravels, the presence of seams, even of small thickness, greatly
increases seepage flows and if undetected may lead to serious and unexpected water losses in hydraulic
construction works.
iii) Density of packing of soil

When the pore volume in a given soil decreases as a result of densification or compaction, the intrinsic
permeability also decreases. This is shown in the relationship below by Casagrande and quoted by Peck
and Terzaghi (1948) of permeability and void ratio;
K =1.4K0.85e2 (14.33b)
where: K is the hydraulic conductivity at void ratio e in a clean sand whose hydraulic conductivity is K0.85
at a void ratio of 0.85. It might be deduced that K  e2. A plot of e against log K for soil is a straight line
for nearly all soils.
iv) Composition of soil

In clays, mineral composition affects a range of properties that include; swelling, consolidation, strength
and permeability. The montmorillonite clays exist at high void ratios yet their permeability is very low.
When clayey soils are used as fill materials the water content at the time of mechanical compaction
appears to influence the microstructure formed and hence the permeability.
v)Consolidation pressure
An increase in effective stress leads to volume decrease in soil and hence a reduction in void ratio and the
value of K. The influence of consolidation on the coefficient of permeability is significant owing to a
reduction in void ratio and a closing or tightening of fissures. When testing soils samples, it is very
important that the same stress that exists at the original place be applied during tests. This aims at
preventing disturbance of the soil. Stress changes may also occur during or as a consequence of
construction works.
vi)Migration of fine particles
When water flows though soil particles, it exerts a drag force that forces finer particles to move through
the void spaces in the coarser fraction. This process is called internal erosion. This mainly occurs in
coarse uniform sands containing fines. The movement of the fines leaves more spaces making the
remaining material more permeable hence increasing the hydraulic conductivity. In case the flow is
beneath a hydraulic structure, a failure may occur. This can be controlled by use of filters, presence of
foreign matter that clogs the pore spaces and hence reduces the hydraulic conductivity.
419
vii) Air or gas in the soil

If the sand is not fully saturated, reductions occur in the coefficient of permeability. A drop from 100
percent to 85 per cent in the degree of saturation being accompanied by a drop of around 50 per cent in
the value of K. The permeability will also change with changes in the pore pressure.
14.7 Steady Groundwater Hydraulics
14.7.1 Hydraulic Conductivity Ellipsoid
Vz
Vs

x
Vx
Fig.14.10 Hydraulic Conductivity Ellipsoid.
Specific Discharge, Vs in an arbitrary direction of flow Hydraulic conductivity ellipse

Assume: A homogeneous, anisotropic medium in the x-z plane
To determine: Hydraulic Conductivity Ks in any arbitrary direction
Given: Hydraulic Conductivity in the x and z directions (Kx, Kz) as shown in Fig.14.10
h
Vs = -K s Darcy’s Law
s
h
Components of Vs are: Vx = -K x = Vs Cos (14.34)
x
h
Vz = -K z = Vs Sin (14.35)
z
h h x h z
Since h = h (x, z) = . + . (14.36)
s x s z s
x z
By geometry: = Cos = Sin (14.37)
s s
x z h h h
Substituting for , , , and into Equation (14.36) together with (14.34), (14.35) and
s s x z s
(14.37).
In rectangular coordinates, x  r cos , z  r sin .
420
Equation of an ellipse (14.38)
Axes = , if r =
,
Similarly in 3 dimensions, hydraulic conductivity ellipsoid has axes
, , (Freeze and Cherry,
1979).
i) Flownets
Flow nets are graphical representations of families of streamlines and equipotential lines are widely used
in groundwater studies to determine quantities, rates and directions of flows. The use of flow nets is
limited to steady incomprehensible flow at constant viscosity and density for homogenous segments.
Darcy’s Law must be applicable to the flow conditions. A flow net can be determined exactly if the
velocity potential Φ (x,y) and the stream function Ψ(x,y) are known before hand. This is often not the
case and as a result, graphically constructed flow nets are widely used. More about their construction can
be found in the literature like Watson and Burnett (1993).
14.7.2 The Dupuit-Thiem Theory

The groundwater is generally extracted from the aquifers by drilling the wells and using pumps for lifting.
The analysis of flow of groundwater towards a well was first analysed by Dupuit in 1863 in the study of
steady flow to wells and ditches, and developed further by Forchheimer in 1886 and later modified by
Thiem in 1906. The following assumptions are made in the Dupuit-Thiem Theory:
1. The aquifer is homogeneous, isotropic and of uniform thickness
2. The aquifer has a uniform lateral extent
3. The well fully penetrates the aquifer and receives water from the entire thickness
4. The flow lines are radial. The flow is horizontal and uniformly distributed in the vertical direction
along the thickness of the aquifer
5. The pumping has been continued for sufficiently long period at a uniform rate to attain an
equilibrium stage (or steady state)
6. Flow is steady and Darcy’s law is applicable
7. The slope of the hydraulic gradient line is small, and it can be taken equal to tan instead of sin,
where is the angle between the hydraulic gradient line and the horizontal
8. The well is of small size and has negligible storage, and, therefore, all the pumped water comes
out of the aquifer.
9. The water is released from the storage in aquifers as soon as the water table falls.
The above assumptions are not strictly valid under actual field conditions. They are however required in
order to simplify the analysis, and to obtain an approximate solution of the complex problem of the well
hydraulics. On the basis of these assumptions, the equations are derived for the discharge (Q) from the
wells. The expressions for the discharge wells are derived separately for unconfined aquifer and the
confined aquifer in the following sections.
Before pumping is started, the water level in the wells is stationary. In an unconfined aquifer, the water
level is at the level of the water table; whereas in a confined aquifer, it is at the level of the piezometric
421
surface. As pumping is started, the water level in the wells drops. The radial flow occurs into the well,
and the water table (or piezometric surface) assumes the shape of an inverted cone around the well as the
pumping is continued. This cone is known as the cone of depression. The area influenced by the cone of
depression is known as the area of influence, and its radius is known as the radius of influence.
The drop in the water table level (or the piezometric surface) at any point from the original static level is
called the drawdown (s). The maximum drawdown (sw) occurs at the well and as the radial distance from
the well increases, the drawdown decreases. When the rate of pumping remains constant, initially the
drawdown curve changes gradually with time. This phase is called the unsteady flow state because the
water table level (or piezometric surface level) at a given point is a function of time. However if the
pumping is continued for a long time, an equilibrium state (i.e. steady state) is reached when the cone of
depression becomes stationary. At that stage, the rate of inflow into the well is exactly equal to the rate of
outflow from it due to pumping.
Examples of the use of the Dupuit-Thiem Theorem to simulate common groundwater problems are
provided in the following sections a) to f) as well as in Section 14.8 on the superposition of elementary
groundwater solutions.( DeWeist, 1965; McWhorter and Sunada,1977; Bouwer, 1978; Wilson,1996;
Arora 2007,)
a) Steady, One Dimensional Unconfined flow between two parallel ditches, unequal water levels
in the ditches and no recharge (equivalent to flow through a dam/embankment with vertical faces)
observed
Seepage face
Calculated Q h0
(Dupuit) h
hL
Fig.14.11 Steady, one dimensional, unconfined flow with unequal water levels.
Refer to Fig 14.11. From Darcy’s Law and Dupuit assumptions:
dh By Dupuit assumptions, Q through any vertical plane is

Q  Kh (14.39) the same. i.e. Q is independent of xQL = Q0 = Q
dx
Integrating between the limits x =0, h = h0 and x = L, h = hL

K 2 2
Q (h 0  h L ) Dupuit Formula (14.40)
2L
which gives the equation for the flow rate (per unit length of the ditch)
422
To find an equation for the water table elevation, h at distance x, re-arrange and use x instead of L,
substituting the above equation for Q;
2Q (h 2  h L2 )
h 2  h 02  x  h 02  0 x (14.41)
K L
(h 02  h L2 )
The equation h 2  h 02  x has the form of a parabola (y2 = cx + b)
L
and expresses the shape of the water table under the Dupuit assumptions i.e. with no vertical flow
component.
The deficiencies in the Dupuit assumptions are apparent at the seepage face, x = L. If the downstream
water level were drawdown to hL =0, the area of flow on the vertical face would be zero and dh/dx , a
situation for which the Dupuit assumption does not hold.
The water table, with no vertical re-charge and under steady state conditions, is a streamline. The
vertical face of the aquifer below water at x = L is an equipotential line (pressure head + elevation head =
constant). Therefore at channel water level at x =L, the stream line and equipotential line should intersect
at right angles, but as in the diagram, this is not supported by the water table described by the Dupuit
assumption. However, the errors in using the Dupuit assumption in practice, are small especially when the
horizontal distances are large compared to the vertical distances and (dh/dx)2<<1.
b) Steady, One Dimensional Unconfined Flow with Recharge, Equal water levels
W
Two parallel ditches with

vertical recharge (W),
Hydraulic Conductivity (K)
h0
h0
x
L/2 L/2
Fig.14.12 Steady, one dimensional unconfined flow with recharge, equal water levels.
Refer to Fig 14.12. Under the Dupuit assumptions, the flow through any vertical plane is
dh
Q  Wx  Kh
dx (14.42)
Separating variables:  Wxdx    Khdh

x2 h2
W  K  C
2 2 (14.43)
423
With boundary conditions x = L/2, h = h0

WL2 Kh 02
C  (14.44)
8 2
Substituting back for C in Equation 14.43
W 2
h2  h02  (x  (L / 2)2 )
K (14.45)
This an ellipse equation describing the form of the water table
At x = 0, water table is at maximum height and the vertical at the point, halfway between the ditches is a
groundwater divide
WL2
 h max   h 02
2
(14.46)
4K
If the recharge rate can be considered a constant, W (e.g. continuous irrigation rates or rainfall over long
time periods), this equation can be used to get an idea of the spacing of agricultural drains – an estimate
of L is to be found so that hmax is less than some given value for a particular crop.
This parallel ditch – constant recharge configuration can also be used to represent a variety of
unconfined flow problems e.g. the ditch may be taken as a stream and the groundwater divide can be
taken as the relatively impermeable side of an alluvium filled valley (since there is no flow across a
groundwater divide) as shown in Fig.14.13.
h0
L/2 L/2
O
Fig.14.13 A river valley with constant recharge.
Example 14.1
A horizontal impermeable clay layer exists 5.1m below the surface of the soil in a humid region where the
annual precipitation exceeds the annual evaporation by 30cm. A subsurface drainage system composed of
equally spaced parallel drains is required to maintain the maximum water table elevation 1.1m below the
surface. Stating any simplifying assumptions derive the equation describing the water table between
adjacent drains. If the drains are placed 2.1m below the surface and K=2.0 x 10-4 cms-1. What is the
spacing between the drains?
424
Solution
w
GL
2.1m 1.1m
5.1m
hmax
ho= = 4.0m
3.0m h
x L/2 Impermeable clay

dh
i
dx
Dupuit’s assumptions Darcy’s law
dh
Q  wx   Kh V  Ki
dx
Q  vh  hKi
rearranging, integrating and applying boundary conditions
The expression for the water table is as in Equation 14.45
w  L2 
h 2  ho2    x 2 
K 4 
at x=0 h  hmax and this is a groundwater divide as in Equation 14.46
wL2
 h0 
2 2
h max
4K
From the above figure
hmax = 4.0m
ho = 3.0m
w =30cm/year
30x10 2
 = 9.513x10-9m/s
365x 24x3600
K= 2.0x10-4cm/s = 2.0x10-6m/s
Substituting known values to obtain L2
L2 
4K 2
w

hmax  ho2 
4 x2 x106

9.513x10 9
4 2  32 
= 5886.68
L = 76.72m = 77m which is the spacing of the drains.
425
Example 14.2
A flat alluvial valley bounded by impermeable shale averages 2150 m in width with a stream flowing
through it in the approximate centre. The valley contains an aquifer in hydraulic connection with the
stream. During the irrigation season recharge can be steady and uniform over the valley. Observation
wells indicate that during irrigation, the water table at a distance 110m from the stream is 3.3mabove the
impermeable shale and 1.2m above the stream level. If the discharge of the aquifer to the stream is 3.325
x 10-2 m3/day per unit width, by making any simplifying assumptions and deriving any necessary
equations determine the hydraulic conductivity of the aquifer.
Solution
w
h ho
3.3m 2.1m
965m 110m 1075m

0
Length=2150/2=1075m
The discharge of aquifer to the stream 0.03325m3/day per unit width
The height of water table at any distance x is given by:
w 
2
L
h  h   x2   
2 2 
K  
o
2 
As in Equation 14.46
h=3.3m ho=2.1m x=965m L/2=1075m
According to Dupuit all w goes to stream
Q  wL  wx
.03325m 3 / day .03325m
w 
2150mx1m 2150x 24x3600s
10
w  1.79x10 m / s
Substituting
1.79x1010
3.32  2.12 
K
9652  10752 
1.79x10 10
k
 
9652  10752  2.12  3.32
1.79x1010
2.1  3.3
2 2

9652  10752  K 
K  6.2x10-6m/s or = 0.536 m/day
426
c) Unconfined flow between parallel ditches with vertical recharge and unequal water levels in
the ditches
W
H1
h
Q H2
x
L
Fig.14.14 Unconfined flow across parallel ditches, unequal water levels and vertical recharge
Q is made up from flow through the aquifer under a differential Head (H1 – H2) as well as from recharge
W as shown in Fig 14.14.
Darcy’s Law with Dupuit assumptions:
dh
Q  Kh (14.47)
dx
Continuity dQ  Wdx (Contribution to Q increases with x)
Q  Wx  C1 (Integrated) (14.48)
Equating (14.47) and (14.48) and separating variables
(Wx  C1 )
hdh  dx
K
W 2 2C1
Integrating h 2  x  x  2C2 (14.49)
K K
Boundary conditions h = H1 at x=0
C2  H12 / 2
and h = H2 at x = L
Substituting back into Equation 14.49 for C2, h and x
WL2 2C1L
H2 
2   H2
K K 2
K 2 WL
Solving for C1 C1  (H2  H12 ) 
2L 2
Substitute back in Equation (14.49)
427
W 2 2x K WL
h2   x  ( (H 22  H12 )  )  H12
K K 2L 2
W x WLx
  x 2  ( H12  H 22 )  )  H12
K L K
W x
h2   (Lx  x2 )  ( H12  H 22 )  H12 Equation of the water table (14.50)
K L
Since Q = Wx + C1
WL K
Q  Wx   (H 22  H12 ) Equation for flow rate (per length of ditch) (14.51)
2 2L
d) Steady Radial flow to a well in an unconfined aquifer

Fig 14.15 shows a well of radius rw, fully penetrating an unconfined aquifer of thickness H below the
water table. The original water table is horizontal. When the water is pumped out from the well at a
constant rate Q for a long time, the water level in the well becomes stable and the equilibrium stage is
reached. At that stage, the contribution of water from the aquifer is just equal to the rate of pumping and
the drawdown curve is established.
Q
r2
Pumped Well
GS r1 Observation Wells
Cone of depression Initial Water - Table

s2
s1
sw r
Drawdown curve P
H Unconfined
h1 h2 Aquifer
hw h
rw
O
ORIGIN
Impermeable Stratum
Fig.14.15 Radial flow in an unconfined aquifer
For the derivation of an expression for the discharge Q, let us consider a cylindrical soil mass around the
well. Let the origin of the cylindrical co-ordinates (r, θ, h) be at the centre of the bottom of the well. Let P
be a point on the drawdown curve at a radial distance of r and at a height of h above the impermeable
stratum. Flow occurs through the vertical sides of the cylindrical surface passing through the point P. The
area of flow A is equal to 2rh.
According to Darcy’s law, the velocity of flow is given by
The discharge Q is equal to the product of the area of flow and the velocity of flow
428
Thus Q  Av
or (14.52)
or (14.53)
Integrating both sides of Equation.14.53 between the limits (r = r1 and h = h1) to (r = r2 and h = h2)
or
(14.54)
or (14.55)
or (14.55a)
Equations 14.55 and 14.55a may be used for the estimation of discharge
e) Steady Flow to a Well in a Confined Aquifer
Fig. 14.16 shows the flow system around a well fully penetrating a confined aquifer thickness b. Let H be
the height of the initial piezometric surface. When the water is pumped out at a constant rate Q for a long
time, the steady flow conditions are established. The water level in the well becomes constant and the
drawdown curve gets stabilised. At that stage, the contribution of water from the aquifer is equal to the
rate of pumping.
Pumped Well
Q
r2
GS r1 Observation Wells
Initial Piezometric surface

s2
s1
sw r
P
Drawdown curve
H
h h1 h2
hw
Confined Aquifer b rw
O
ORIGIN
Fig.14.16 Radial flow in a confined aquifer

429
Let us consider the flow through the vertical sides of a cylindrical surface passing through a point P at a
radial distance r. In this case, because the depth of flow b is constant, the area of flow is equal to 2 rb.
Using Darcy’ law, the discharge is given by
(14.56)
Where: i is the hydraulic gradient,K is the hydraulic conductivity

or
or
(14.57)
Integrating between the limits (r = r1 and h = h1) to (r = r2 and h = h2)
or
(14.58)
or
(14.58a)
Alternative forms
1. Equations 14.58 and 14.58a can also be written in terms of the drawdown s1 and s2. As h2 =H– s2
and h1 = H–s1
(14.59)
2. In terms of the transmissivity T = bk, Equation 14.58a becomes

2T (h 2  h1 )
Q (14.60a)
log e (r2 / r1 )
2T (s1  s 2 )
or Q (14.60b)
log e (r2 / r1 )
Example 14.3
A fully penetrating well of diameter 0.4m is drilled in a confined aquifer 2.5m thick. If the steady state
drawdowns at 10m and 50m are observed to be 2.5m and 0.5m, determine the discharge. Take k = 10 –3
m/s
430
Solution
From Equation 14.59
Example 14.4
A well of diameter 30cm fully penetrates a confined aquifer of thickness 1.5m when pumped at a steady
rate of 30 l/s, the drawdowns observed in wells at radial distances of 10m and 40m, are 1.50 and 1.0m,
respectively. Compute the radius of influence, the permeability, the transmissivity and the drawdown at
the well.
Solution
From Equation 14.59

or
or
Transmissivity, T = bK = 15 x 8.8 x 10 –4 = 1.32 x 10 –2 m2/s
Let sw be the drawdown in the well
From Equation 14.59
2 x15 x 8.8 x10 4 (s w  1.0)

30 x10 
3
5.586
sw  3.02m
Let R be the radius of influence
From Equation 14.59,
2 x15 x 8.8 x10 4 x 3.02

30 x10 3 
log e (R / 0.15)
R  634.0m
431
f) Steady Radial flow in an Unconfined Aquifer with Uniform Recharge

Qw
W
G.L
Water table with

recharge but no Water table before
pumping pumping and no recharge
dh Pseudo –steady state

water table during
pumping and no recharge
he
h
Q+dQ Q
r
dr
re ro
Fig 14.17 Steady radial flow in an unconfined aquifer with uniform recharge
Refer to Fig 14.17

Darcy’s law with Dupuit Approximation:
dh
Q  2 rKh (14.61)
dr
dQ
Continuity  2rW (14.62)
dr
Since recharge increases the flow Q with decreasing radial distance r from the well, continuity gives the
incremental increase in Q for an incremental change in r.
Integrated Q   r 2W  C1 (14.63)
 r 2W  C1 dr
Combined hdh  
2 K r
W 2 C1
Integrated h2   r  ln r  C2 (14.64)
2K K
Use boundary conditions to find C1 and C2
From Equation (14.63), At r = 0, Q=C1=Qw (pumping rate = constant)

At r = re, h = he (At the radius of influence or equilibrium)
432
W 2 Qw
So, from Equation (14.64) C 2  he2  re  ln re
2K K
Substitute back in (14.64) for C2:
he2  h 2  
 

W re2  r 2 Qw  re 
ln  (14.65)
2K K  r 
This gives the equation of the water table for radial flow with constant pumping and constant recharge.
When Qw = 0, ie no pumping from well and the water table is in the form of the top dotted line in Fig
14.17
he2  h 2  

W re2  r 2  (14.65a)
2K
If there is no recharge, i.e. W = 0, then Equation 14.65 reduces to Equation 14.55, which is the equation of
radial flow in an unconfined aquifer and the drawdown is in the form of the lowest curve as in Figs 14.15
and 14.17. The middle dotted line shows the water table before pumping without recharge.
Since Drawdown s = he – h
W Qw (r )
s (re2  r 2 )  ln e (14.66)
2 K (2he  s )  K (2he  s) r
If at steady state conditions, W = Recharge rate = Qw = Pumping rate (which may be true over a long time
period)
Qw   (re2  rw2 )W from radial geometry (14.66a)
Substitute in Equation (14.65), so that at the well (r = rw, h = hw)
Qw
he2  hw2  ln(re / rw )  1 2 (14.67)
K
 K (he2  hw2 )
OR Qw  (14.67a)
ln(re / rw )  1 2
14.7.3 Radius of Influence

The radius at which the cone of depression touches the original water table is called the radius of
influence (R). It is difficult to estimate the radius R accurately because the cone becomes asymptotic to
the water table. R is usually estimated from Sichard’s formula (1927):
R =3000sw√K (14.68)
where: R is the radius of influence (m), sw = drawdown in the well (m), and K is the hydraulic
conductivity (m/s).
433
The radius of influence of most of the wells ranges between 150m and 500m. The expression for the
discharge derived below does not vary much even with a large change in the value of R. Therefore even
an approximate value or R in the above range will not affect the accuracy much. An average value of R
equal to 300 m is generally assumed.
i) Expression for discharge in terms of radius of influence:
For the determination of the discharge by Equation (14.55), it is necessary to determine the height h1 and
h2 of the cone depression at radial distance of r1 and r2, respectively. For this purpose, two observation
wells are required in addition to the main pumping well. To avoid the drilling of observation wells, an
alternative equation is derived by integrating Equation 14.53 between the limits (r = rw and h = hw) to (r =
R and h = H), where rwis the radius of the well, hw is the depth of water in it, R is the radius of influence
and H is the thickness of the aquifer from the impervious stratum to the water table.
or
(14.68a)
or
(14.68b)
or
(14.68c)
Equations 14.68b and 14.68c can also be expressed in alternate forms as follows
Alternative forms
1. Let us express H2  h 2w as (H  h w )(H  h w )
But H2 – hw = sw = drawdown in the well-
and H +hw = (H - hw) + 2hw
Therefore Equations 14.68c can be written as
(14.69)
or
(14.69a)
2. Equations 14.68b and 14.68c can also be written in terms of the drawdowns. Let s1 and s2 be the
drawdowns at radial distances of r1 and r2 respectively. Thus
s1  H - h1 and s2  H - h 2
Therefore h 22  h12  (H  s2 )2  (H  s1 )2
434
From Equation 14.55
(14.70)
Example 14.5
A well penetrates in the centre of an unconfined aquifer bounded externally by a circle of radius 100m
along which the height of the water table is 8m. If at a distance of 10m from the centre of well, the height
of water table is 7.5m when steady conditions are established, determine the discharge of the well. Take K
= 10-4 m/s.
Solution
From Equation (14.55)
Example 14. 6
A 0.4m diameter well fully penetrates an unconfined aquifer whose bottom is 80m below the undisturbed
groundwater table. When pumped at a steady rate of 1.50 m3/min, the drawdown observed in two
observation wells at radial distances of 5m and 15m are; 4m and 2m respectively. Determine the
drawdown in the well
Solution
or
Let hw be the depth of water in the well
or
(782  hw2 )  1209.8
or
hw  69.82
Thus sw  80  69.82 10.18m
435
Example 14.7
The height of the water table above an impermeable base surrounding a circular island of radius 350 m is
10m. If this island receives 1.2 mm/h of rainfall, determine the height of the water table in the borehole of
0.2 m diameter drilled in the centre of the island. Take K = 0.07 m/d and assume there is no pumping.
Qw= 0
W
=fnfn G.L
1///
Water table with 11/1
recharge but no Initial level of Water
pumping Table
dh Pseudo –Steady state

water table during
pumping
he
h
=12 Q+dQ Q
r
dr
re ro
W = 1.2. mm/h = 3.333 x 10-7 m/s

ro = 0.2/2 = 0.1 m
he = 10 m
re = 350 m
K= 7m/day = 8.102 x 10-5 m/s
ho = ?
Solution
From Equation (14.65 a)
he2  h 2  

W re2  r 2  (Assuming no pumping; Qw = 0)
2K
102  h2  

3.333x107 3502  .12 
5
2 x8.102 x10
h2 = 262.14 – 100
h = 12.73m which is the height of the water table in the well
436
14.7.4 Partial Penetration

When the open section of a well casing does not include the full thickness of the aquifer it penetrates,
the well is referred as partially penetrating.
Generally, the flow or discharge as a result of pumping a well is given by the Theim formulas for steady flow
as shown below:
K (d 2  hw2 )
Q (for unconfined aquifers) (14.71)
R
2.3 log10  
 rw 
2KH ( D  hw )
Q (for confined aquifers) (14.72)
R
2.3 log10  
 rw 
where:
rw = radius of the well
K = hydraulic conductivity
R = radius of influence
H = total height of the confined aquifer
D = height of the aquifer below the original piezometric surface for the confined aquifer
d = height of the original water table
hw = artesian pressure in the well
in both cases the flow is radial and horizontal.
However, for a partially penetrating well as shown in Fig 14.18, these formulas are not applicable as the
nature of flow will become three dimensional. It will not only be radial but will also have an upward
component as shown in the figure below. The yield of such a well is found to be more than that of a fully
penetrating well of the same depth.
Well penetrating partially in the
aquifer
Aquifer
Fig 14.18 Flow lines in a well partly penetrating the aquifer.
According to Kozeny correction factor for the discharge of a partially penetrating well in unconfined aquifer
is given as:
437
 
 2 
 .K .( d1  hw  
2
d 
Q
r
1  7. w . cos 1  (14.73)
  R  2 d1 2d 
 2.3 log10    
  rw  
where:
d1= actual penetration depth below the water table
d= Actual depth of the aquifer below the water table
Similarly, the discharge for a partially penetrating well in a confined aquifer:

 
 
 2 .K .H 1 ( D  h)   r H 
Q 1  7. w . cos 1  (14.74)
 R  2 H1 2H 
 2.3 log10    
  rw  
where:
H = Full depth of the confined aquifer
H1 = Depth up to which the well penetrates
14.8 Superposition of Elementary Solutions in Steady Groundwater Flow

Since the equations describing steady groundwater flow are linear, homogeneous partial differential
equations, linear combinations of their solutions can be used to solve more complex problems (Muskat,
1937).
For example consider the equation describing the piezometric surface for radial unconfined flow with
recharge (i.e. an island) as in Equation 14.65a.
where: re = radius of the island at which h = he, the
he2  h2  

W r r e
2 2
 surrounding water level
(14.75)
2K
Consider also the equation for radial unconfined flow to a pumping well as in Equation 14.54.;
Qw  re 
he2  h 2 =
πK  r 
ln (14.76) re = radius of influence
By the principle of super position these two elementary solutions can be combined to give the equation
for the piezometric surface for radial flow to a pumping well with recharge in an unconfined aquifer as in
Equation 14.65;
h h  
2 2


W re2  r 2 Qw  re 
ln 
 (14.77)
K  r 
e
2K
i.e.the sum of the two elementary solutions
The same principle can be used to combine any number of component solutions in order to analyse more
complicated groundwater flow situations. An example is the analysis of flow to a well field i.e. a cluster
of wells all pumping from the same aquifer horizon.
438
An example of a well field in a confined aquifer, all wells fully penetrating

(It is representative of foundation or mine dewatering problems)
The drawdown at any point is found from the sum of the individual contributing drawdowns from each
pumping well in the field.
Qi
si  ln (re / ri ) ri  re where: i refers to the ith well in the field
2T
si is that of the drawdown at the point of interest that can be attributed to the ith well.
ri is the distance from the ith well to the point of interest.
 s = Total Drawdown at the point of interest = sum of the contributing drawdowns from each of the n
wells in the field.
n
Qi
s  ln (re / ri ), ri  re
i 1 2T
To contribute to the drawdown s, a given well in the field must be within range of its radius of influence,
hence ri< re.. If there are a large number of wells in the fields, analysis of the groundwater flow can be
further simplified by representing then n wells or withdrawals points by a single, uniformly distributed
withdrawal rate W= discharge/unit area, applying over the area of the well field.
n
Qi Where A=Area of well field, idealised by a circle of radius R
W  
i 1 A
re
For n well points, total withdrawal = R W = 2
Q, the pumping rate from the well field as a R
whole as shown in Fig 14.19

Well field
The incremental change in drawdown, ds over
an incremental annular strip, width dr in the
circular well field, and area 2rdr
Fig.14.19 A Well Field
dQ  2r drW
(Recall Thiem equation for a single well
2r drW
and  ds  ln(re / r) Q
2T s ln(re / r) ) (14.78)
2T
rdrW
 ds  ln re / r
T
Integrate to find s0, the drawdown at the centre of the well field
WR
T 0
s0  r ln (re / r)dr
(14.79)
WR 2
s0  (ln (re / R)  1 )
2T 2
439
The analysis of well field has been simplified by thinking of it as a single well problem with negative
recharge (an areally uniform withdrawal rate). This solution will clearly not predict the drawdown near
individual wells in the field but it does give an estimate of the overall drawdown caused by the well field
as a whole.
14.8.1 Collector Wells

Collector wells provide another example of a complex flow field, which can be easier solved analytically
by a hypothetical single, ordinary well of large radius as shown in Fig 14.20.
Q
Section Plan re
rw
d Piezometric head
Lateral
SHAFT
Confined Lateral
aquifer
Fig. 14.20 Representation of a Ranney collector well with laterals in a ‘Star’ pattern
Typical dimensions are: 2rw = 1.5 – 4.5m, re= 40 – 50m and d = 20 – 50m
Using the same approach as for a well field, let the total pumping rate or withdrawal from the collector
well be represented by a uniform withdrawal rate W over the whole plan area of the collectors i.e. a
circular area of radius R.
Discharge = Q =  R2W
and the drawdown at the centre of the collector well becomes
Q WR 2 r
s0  (ln (re / R)  1 ) [Since s0  (ln( e )  1 ) -Derived for the well field]
2T 2 2T R 2
If the collector well can be replaced by, or be imagined equivalent to, a single ordinary well of radius rw
Q
for which s w  ln(re / rw ) [Thiem equation],
2T
Equating sw and s0 gives
r
ln(re / rw )  ln ( e )  1
R 2
rw  0.6R
which implies that a collector well is equivalent to a single ordinary well of radius 0.6 R, although in
practice, rw is more nearly equal to 0.8R from field data. Hence collector wells can be useful for large
pumping capacity in relatively thin aquifers although their capital cost is high.
440
14.8.2 The Method of Images

(Pumping Near Hydrogeological Boundaries)
Hydrogeological boundaries in aquifers are physical boundaries to a groundwater flow system such as
streams (an equipotential boundary) or impermeable barriers (e.g. some faults or an impermeable hard
rock formation). The steady well-flow equations given so far have been developed for aquifers of infinite
extent, and physical boundaries represent finite constraints on such flow systems (Ferris et al., 1962)
However the boundaries can be replaced or represented by an imaginary, equivalent flow system that will
allow the infinite aquifer equations to be used. In other words an image well or wells can be used to
simulate an aquifer boundary, thereby transforming a semi-finite aquifer flow system into an infinite one
in which the radial flow equations can be used.
Example: Pumping well near a stream in a confined aquifer.
Stream
Confined
aquifer
y
PLAN: steam is in y - axis
WELL
X
a
Fig 14.21 Pumping well near a stream in a confined aquifer.

( Source: Ferris et al., 1962)
Consider a semi-infinite aquifer with a constant head or equipotential boundary at the stream.
There is no radial symmetry but the stream can be thought or as a line source as shown in Fig 14.21.
Assumptions
1. The stream stage is constant, so that the interface with the aquifer is a constant head/equipotential
boundary.
2. Both the well and the stream fully penetrate the aquifer. Although in practice, the stream rarely
fully penetrates the aquifer, the consequent errors in the analysis turn out to be relatively small.
3. There is no sealing layer of fine sediment on the streambed (i.e. full hydraulic connection).
4. Pseudo – steady state conditions.
441
The method of images suggests that the stream boundary can be replaced or represented by an image
recharge well, recharging at rate – Q,
at x = -a, y = 0. If the pumping well is y
at x = a, y = 0, pumping at rate +Q as
shown in Fig 14.22. P
The drawdown in the semi-infinite ri
aquifer then equals the drawdown rp y
from the pumped well + the
PUMPING IMAGE RECHARGE
drawdown (or build–up) from the WELL
WELL a x -a
image recharge
Fig.14.22 Pumping and image recharge well
At any point P, from the Thiem equation and superposition, the drawdown s is given by
Image Pump
Q r Q r
s ln ( e )  ln ( e )
2T ri 2T rp
(14.80)
Q  re r  Q r
  ln ( )  ln ( e )   ln ( i )
2T  rp ri  2T rp
Expressing r in terms of x, y, a, from geometry:
{(a  x)2  y 2 }
1
Q 2
s ln
2T {(a  x)2  y 2 }
1
2
If x = 0 in this equation, s = 0 that is, along the y-axis, which is the position of the stream, drawdown is
zero and the line is an equipotential line along which h = h0, a constant  stage in the stream as in Fig
14.23.
The imaginary recharge well therefore produces the same effect as a stream along the Y – axis.
Q -Q
S=0
Image recharge
well
Stream stage
Actual
Drawdown from
pumping
a Position of a
Stream
Fig.14.23 Equivalent hydraulic system of a confined aquifer near a stream

Source: (Ferrris et al., 1962)
442
At s= 0, the pumped well is withdrawing water at the same rate as the image recharge well is supplying it,
hence drawdown is zero.
Since, from the equation already derived for drawdown at any point P,
Q r
s ln ( l )
2T rr
Drawdown at the pumped well, sw is Since rp = rw
ri = 2a – rw and
Q 2a  rw ) Q (2a) Usually rw 2a
sw  s  ln (  ln (14.81)
2T rw 2T (rw )
STREAM
100
90
PUMPING
80
WELL
60
40 IMAGE
RECHARGE
WELL
Fig. 14.24 Equipotentials and streamlines for pumping well near image recharging well
Source: (Ferris et al., 1962)
(Note: The numbers on the equipotentials refer to the potential as a percentage of the total head difference between
the recharging boundary and the well). Fig.14.24 gives a representation of equipotentials and streamlines of
the pumping well near an equivalent recharge well. Thus the resulting flownet shows a single
equipotential line, which is coincident with the axis of the stream. The resultant drawdown of the real
well is given by the algebraic sum of the drawdown of the real well and the build up of the recharge well
as if these wells were located in an infinite aquifer (Todd, 1980)
100
60
40
PUMPING 20 IMAGE PUMPING
WELL WELL
Solid Boundary
Fig.14.25 Equipotentials and streamlines for pumping well near image pumping well
443
Using the same approach it can be shown that an impermeable, solid boundary near a pumping well can
be represented by an image pumping well. If the pumping well is at x = a, y = 0, the Y axis being
impermeable boundary, the image pumping well is at x = -a, y = 0
(Note: The numbers on the equipotentials refer to the potential as a percentage of the “undisturbed” total head at
great distance)
In the same way, Fig 14.25 shows the flownet of the pumping well near an image pumping well. The
solid boundary at the centre represents an equipotential surface and a line of no flow.
Parallel Boundaries
With aquifer boundaries, the image well combinations arising from their analysis rapidly become
complex- An image well corresponding to one boundary will itself require another image well,
corresponding to any other boundary.
In the case of two parallel boundaries, two image planes are considered which leads to an infinite set of
images as shown in Figs14.26a and 14.26b. However, in practice pairs of image wells are added until the
next pair has negligible influence on the sum of all image well effects out to the point..
Barrier Recharge
boundary boundary
I I I I I I I
5 4 2 P.W. 1 3 6 7
2a 2b 2a b b a a 2b 2a 2b 2a
(a) Pumping well between recharge and barrier boundaries
I I I P.W. I I I I
4 5 2 1 3 6 7
2a 2b 2a b b a a 2b 2a 2b 2a
Barrier boundries
(b) Pumping well between two barrier boundaries
Recharging image well (RIW)

Pumping image well (PIW)
Real pumping well (P.W.)
Fig. 14.26 Image well system for pumping well located between parallel boundaries
Note that in the analysis, however, those image wells, whose radii of influence fall outside the boundary
need not be considered in the analysis. The imaginary recharge well therefore produces the same effect as
a stream along the Y- axis.
444
Note the method of images can also be used in the vertical plane.
An Example: A horizontal drain beneath a ponded surface (equipotential boundary) as shown in Fig
14.27.
Z
GSL
d1
X
DRAIN
d2
Fig.14.27 A horizontal drain beneath a ponded surface.
The situation is equivalent to a pumping well near an equipotential boundary in the horizontal plane
provided that the base of the aquifer is a long way below the drain and cannot itself be considered a
boundary i.e. d2>>d1. The image to the drain representing the equipotential, ponded surface is a
“Recharge Drain”, d1 above ground level. The equations for pumping-image recharge well combination
apply but with appropriate parameter changes:
(14.82)
the horizontal
becomes (14.83)
and becomes (14.84)

where:
rd is the radius of the drain.
b becomes the length of drain (perpendicular to the page)
ho is the peizometric head at the ground surface
hz is the piezometric head at the drain surface
445
Example 14.8
Determine the drawdown midway between a borehole and a stream 200 m apart, hydraulically connected
to a confined aquifer, whose thickness is 10.0 m. The borehole is being pumped at a rate of 0.3 m3/hour
and the aquifer has a hydraulic conductivity of 2.0 x 10-6 m/s.
The drawdown at any point p in a confined aquifer due to a pumping well is given as
Q r 
s ln e 
2 bK  r p 
 (1)
where: re= radius of equilibrium or influence or of aquifer

rp = distance from pumping well to point of interest
The Method of Images allows us to represent a hydrogeological boundary by an equivalent system. In this
case the pumping well near the stream is represented by an additional image recharge well shown in
dotted lines at the same distance from the stream ri as shown in the figure below.
The draw down at any point p due the image well is

Q r 
s ln e 
2 bK  ri  (2)
where: ri = distance from pumping well to point of interest
200m 200m
Q Q
Pumping
100m 300m Recharge
rp
ri
Stream
10m
rp A r
i
The drawdown at point A is made up of contribution of the pumping well at rp and recharge well at ri.
By the principle of superposition, we add Equations (1) + (2) as follows:
Q r Q r
sA  ln e  ln e
2 bK rp 2 bK ri
Q r
 ln i
2 bK rp
Given Q=0.3m3/hr, ri=300m, rp=100m, b==10m, K=2.0x10-6m/s
0.3 300
 In
2 x x 2 x10 x10 x60 x60 100
6
sA = 0.729m
446
14.9 Unsteady Flow

14.9.1 Pumping Tests
According to Krusemann and De Ridder (2000), the principle of a pumping test is that if we pump water
from a well and measure the discharge and the drawdown in the well and in piezometer at known
distances from the well, we can substitute these measurements into an appropriate well-flow equation and
calculate the hydraulic characteristics of the aquifer.
Before a pumping test is done, preliminary studies are carried out and information on the geology and
hydrologic characteristics of the source is collected (Johnston, 1983). Geological information includes
information on the subsurface such as lithological and stratigraphic features. Hydrological information
includes aquifer type, confining bed, aquifer boundaries such as bedrock at the sides and lateral recharge
boundaries. Additional information collected includes:
 Data on groundwater flow system

 Any existing wells in the area.
i) Site Selection
The following conditions govern the selection of a new site for a pumping test:
 Steady hydrological conditions which do not change with time
 The site should not be near motorways or railways
 The site should not be in the vicinity of existing wells
 The site should be accessible from the main road.
An existing well can also be used for the test.
ii) The Well

After selecting a site, the drilling operations start. A well essentially consist of an open-ended pipe,
perforated or fitted with a screen in the aquifer to all water to enter the pipe, and equipped with a pump to
lift the water to the surface. A typical design of a well consists of:
a) Well diameter:
Normally a large diameter is not needed for a pumping test. For shallow wells in particular, the
diameter of the well can be small.
b) Well depth:
The depth is normally determined from the log of an exploratory borehole or from logs of nearby
existing wells if any.
c) Well screen:
The length and depth of the screen is determined by the depth at which the coarsest material
materials are found. This is normally determined by examining well logs. The screen should be
slotted or perforated with no more than 30-40% its circumference to keep the entrance velocity
low i.e. less than 3cm/s in order to limit friction losses.
d) Gravel pack:
A gravel pack is usually an artificially-graded coarser material, which is fitted surrounding the
well so as to retain much of the aquifer material that would otherwise enter the well. The
thickness is normally in the range of 8-15cm. Gravel pack material should be clean and smooth
rounded.
447
e) The pump:
Various pumps can be fitted to the well to lift the water to the surface. It should be able to pump
water continuously for a period of at least a few days. The pump should also be able to produce
good and measurable draw downs of at least 100 or 200mm.
iii) Piezometers
Piezometers are fitted near the pumping well to measure the drawdown as a result of pumping. A
piezometer (Fig 14.28) is an open-ended pipe placed in a borehole that has been drilled to the desired
depth in the ground. Placing piezometers between 10 and 100m from the well will normally give reliable
data in most cases. For thick aquifers, the distances should be greater say between 100 m and 250 m or
more from the well.
This is the instrument used to determine the value of h at some point P in a groundwater reservoir.
Piezometers can also be installed in groups to determine directions of groundwater flow.
Fig.14.28 Hydraulic head h, pressure head r, and elevation z for a field piezometer .
iv) Measurements
Two kinds of measurement are done during a pumping test:
 Measurements of the water levels in the well and the piezometers
 Measurements of the discharge rate from the well
v) Duration of pumping
There is no set time for how long a pumping test should take. A pumping test should continue long
enough to determine the effects of the proposed pumping.
However, the following guidelines can be used.
Steady state conditions:

 For leaky aquifers, steady state is usually reached after 15 to 20 hours of pumping
 In confined aquifers, it is good practice to pump for 24 hours
 In unconfined aquifers, because the cone of depression expands slowly, a longer period is
required say 3 days
It is not absolutely necessary to continue pumping until a steady state has been reached, because methods
are available to analyze unsteady-state data.
vi) Carrying out the test

a) Obtain the groundwater levels in the test well and the piezometers or nearby wells before
pumping.
b) Start pumping the well and keep the pumping rate constant throughout the test. Take
measurements of the pumping rate.
448
c) Monitor and measure the groundwater levels in the well and the piezometers at specified
intervals of time, say every hour or every two hours. After a longer time of pumping, the time
interval can be increased.
d) Measure groundwater levels accurately - noting the exact time you made the measurement.
Battery powered groundwater probes can be used that have a light or alarm which goes off when
it’s dipped into water.
e) After pumping stops, keep monitoring groundwater for a period similar to the pumping test, or
until water levels have recovered to the pre-test level.
f) Record all the measurements including the date, location and site diagram.
When carrying out a step drawdown pumping test, the number of steps should be at least four. The
other procedures should be basically the same.
vii) Classification of Boreholes

The Directorate of Water Development has provided a classification of boreholes based on their yield
as indicated in Table 14.2
Table 14.2 Classification of Borehole Yields (DWD)
Yieldm3/hr Classification Potential

0-0.5 Poor Poor
0.5-0.7 Moderate Moderate
0.7-1.2 High Good
> 1.2 Very high Very good
14.9.2 Derivation of the Equations

The equations for discharge from a well were developed on the assumption that the flow was steady. The
steady flow conditions occur in a well when it is pumped at a constant rate and the cone of depression
does not change with time. The steady flow condition would occur only when the pumping is continued
for a long time. In practice, the steady flow condition rarely exist because, as the water is pumped out
from the well at a constant rate the storage within the aquifer is reduced. Consequently, the head goes on
decreasing. However, the rate of decline of the head decreases as the area of influence of the well
expands. This causes the variation of the cone of depression with an increase in time, resulting in
unsteady or transient flow towards the well. Thus in unsteady flow, the cone of depression does not
remain stationary when the discharge is constant.
The following assumptions are made for unsteady flow:
1. The aquifer is homogeneous and isotropic. It is on uniform thickness and has an infinite areal
extent.
2. The piezometric surface is horizontal before the pumping is started.
449
3. The well is pumped at a constant discharge rate.

4. The well is of small diameter such that the storage within the well is negligibly small.
5. There is no recharge to the aquifer or leakage from the aquifer.
6. The aquifer starts dewatering as soon as the piezometric surface drops. In other words, the storage
in the aquifer is released instantaneously.
7. Darcy’s law is applicable.
Fig 14.29 shows a well fully penetrating a confined aquifer of thickness b. Let us consider flow through
an annular cylinder of soil with radius r and thickness dr, at a distance of r from the centre of the well.
From the principle of continuity of flow, the difference of the rate of inflow and the rate of outflow from
the annular cylinder is equal to the rate of change of volume of water within the annular space. Thus:
v
Q1  Q2  (14.85)
t
Where Q1 is the rate of inflow, Q2 is the rate of outflow and v / t is the rate of change of volume (V)
within the annular space.
The slope of hydraulic gradient line (i.e. the piezometric surface) at the inner surface is h/ r , where h is
the height of piezometric surface above the impervious stratum. Therefore, the slope of the hydraulic
 h    2h  
gradient line at the outer surface is equal to     2  dr 
 r   r  
Now by Darcy’s law, Discharge = K x i x (area of flow)
Therefore, Inflow, (14.86)
Outflow (14.87)
Now from the definition of the storage coefficient S, is the volume of water released per unit surface area
per unit change in the head normal to the surface.
Therefore, change in volume V  S  2πr  dr dh

V h
Therefore,  S  2πr  dr (14.88)
t dt
where t is the time since the beginning of pumping.
Substituting the values from Equations (14.86), (14.87) and (14.88) in Equation (14.85)
or
Dividing by Kb (2r) dr throughout and neglecting the higher order terms
(14.89)
Substituting the transmissivity T for Kb
450
Equation (14.89) is the basic equation of unsteady flow towards the well and it applies to both confined
and unconfined aquifer flow. In the equation, h is the head, r is the radial distance from the well. S is
storage coefficient for confined flow (or S is the specific yield, Sy, for unconfined flow), T is
transmissivity, and t is the time since the beginning of pumping.
Pumped Well
Q
GS
h h h2
Initial Piezometric surface Slope = Slope =  dr
r r r 2
Instantaneous
Drawdown curve
h1
Q2 Q1
b
r dr
Annular Ring
SECTION
Q2 Q1
dr r
Fig.14.29 Unsteady flow in a confined aquifer.
  2h 1  h   S  h 
 r 2  r  r    T  t  (14.90)
    
(T/S) is sometimes known as , the Hydraulic Diffusivity,

thus
451
  2h 1  h   1  h 
 r 2  r  r    α  t 
    
14.9.3 Pumping Test Analysis Methods

The following methods are commonly used for the determination of aquifer constants S and T in the field
by conducting pumping tests.
1. Theis method
2. Cooper –Jacob’s method
3. Chow’s method
In all the three methods, besides the pump well (or test well), only one observation well is required. The
water from the pump well is pumped out at a constant rate. Q. The drawdown in the observation well is
measured in different instants of time t. The radial distance r of the observation well from the pump well
is also measured. The aquifer constants S and T are then determined from the measured values of Q, s, t
and r by different methods as explained below.
14.9.4 Theis Equation.

Theis (1935) obtained the solution of Equation 14.90, based on an analogy between groundwater flow
towards a well and the flow of heat towards a sink. The heat is assumed to be removed from the sink at a
uniform rate, just as the water is removed from the well at a constant rate.
Assuming that the well is replaced by a mathematical sink of constant strength, and considering the
following boundary conditions,
h = H for t = 0 (H= piezometric head)
and h approaches H as r approaches  for t 0
Theis obtained the solution of Equation 14.90 as
Q  eu du
4πT u u
s (14.91)
where s is the drawdown at a radial distance r from the well after time t. The drawdown is equal to
(H – h ). Q is the constant discharge, and u (Boltzman variable) is a function given by
r 2S r 2
u  (14.92)
4Tt 4αT
T
where α 
S
Equation (14.91) is known as non-equilibrium equation or Theis Equation. The integral in this equation is
called the exponential integral. This integral is usually denoted as W(u). Thus

 eu 
W (u)     du (14.93)
u u 
Where W(u) is called the Well function.
Q
Hence Equation (14.91) can be written as s  W(u) (14.94)
4πT
452
Applications
From Equation 14.94 and from Equation 14.92, for a given aquifer, the terms in the parenthesis are
constants. From these equations, it can be construed that the relationship between W (u) and u must be
similar to the relation between S and (r2/t) and when these are plotted on the same scale double log plot,
the resulting curves will be of the same shape but horizontally and vertically offset by
Q 4T
constants and . Making use of this similarity, Theis suggested a graphical method of
4 πT S
superposition for the determination of S and T. The following procedure is used.
Draw a curve between W(u) and 1/u on a log paper using the tabulated values given in Table 14.3. The
curve is known as the reverse type curve as shown in Fig 14.30.
Fig.14.30 Type curve and measured values

(Source: Namutete, 2006)
i). Values of drawdowns are plotted against values of (t/r2) (Fig. 14.31)) on logarithmic paper both
scales for the same size as for the type curve. The curve is called the data curve.
ii). The observed time-drawdown data are superimposed on the type curve, keeping the coordinate axes
of the two curves parallel and adjusted until a position is found by trial whereby most of the plotted
points of the observed data fall on a segment of the type curve.
iii). Any convenient point is selected, and the coordinates of this match point are recorded. With values of
W(u), 1/u, s, and t /r2 determined, S and T can be obtained from Equations. 14.94 and 14.92.
 Q 
Thus T  W(u) (14.95)
 4πs 
 4T 
and S  2  u … (14.96)
r /t
453
2
A plot of Drawdown s against Log (t/r ) (DWD13287)
50
40
30
s (m)
20
10
0
0.0001 0.001 0.01 0.1 1
t/r2 (min/m2)
Fig.14.31 Plot of drawdown (s) against log (t/r2).

(Source: Namutete 2006)
Note:
1. Early part of observed data may not be well fitted by the theoretical Theis equation (i.e. for small
values of t and near the well) since it assumes Q starts as a step function with time (at t =0) and then
stays constant (pumping rate will actually vary as pumps adjust to changing head, as drawdown
increases) and since it assumes immediate release of water from storage (in practice, there if often a
delay before water arrives in the well).
2. If u>0.01, log type curve is fairly flat and matching is difficult; other solution techniques may
have to be used.
3. Remember the other assumptions inherent in the method
 Aquifer has infinite areal extent, is isotropic, homogeneous and has uniform thickness, at
least in area influenced by pumping tests
 Before pumping, initial piezometric surface is horizontal
The results from this technique are remarkably robust but if one or more of the underlying assumptions
are seriously violated, discrepancies in pumping test results may be explained.
Example 14.9
The following readings were taken in an observation well 39 m from a well pumped at a constant rate of
1.25 m3/min. Find the transmissivity and storage constant for the aquifer (after Dake, 1983).
Time (h) 0.3 1.6 3.1 5.2 11 23 56

Drawdown (cm) 9.9 31.2 42.0 55.4 75.1 95.2 126.0
r2/t (m2/h) 5070 951 491 293 138 66 27
u 0.01 0.02 0.05 0.1 0.2 0.5 1 2 5
W(u) 4.04 3.35 2.47 1.82 1.22 0.56 0.22 0.05 0.001
454
The s vs. r2/t plot superimposed on the W(u) vs. u curve is shown in Fig 14.32. The identified common
point has the following coordinates:
W(u) = 2.62, u = 0.045 and r2/t = 329 (m2/h), s = 0.53 (m)

r 2S 329 S
From u  , 0.045 
4Tt 4 T
From s 
Q
W u  , 0.53 
1.25x60 2.62
4 T 4 T
Thus T = 29.50 m2/h and S = 0.016
r 2/t (m2/hr)
2 3 4 5
10 10 10 10 10
10
10
Drawdown s (m)
1.0
W(u) - u curve
Common 1.00
Point
0.1
W(u)
0.10
0.01
2 3 4 5
10 10 10 10 10
r 2/t (m2/hr)
0.01
0.01 0.10 1.00 10
u
2
Fig 14.32 Plot of s vs. r /t superimposed on the W(u) vs. u plot
14.9.5 Cooper –Jacob’s method
The Well Function W(u) in Equation 14.93 can be expressed in a convergent series as
 u2 u3 u4 
W (u)   0.5772  loge (u)  u    (14.97)
 2x2! 3x3! 4x4! 
Where factorial 2! = 2 x 1, and 3! = 3 x 2 x1 and so on.

Wenzel (1942) tabulated the values of W(u) for different values of u as shown in the Table 14.3.However;
some approximations can be made for easier evaluations.
If u < 0.01, the first two terms are sufficient to evaluate the equation:
455
W (u)   0.5772  loge (u)

Q
 s    0.5772  loge (u) 
4πT (14.98)
2
r
u is small if r is small and / or time t is large, since u  e.g. The error in s is< 6% for u < 0.10
4αT
Table 14.3 Values of well function W (u) for different values of u.
u 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0

x1 0.29 0.049 0.013 0.0038 0.0011 0.00036 0.00012 0.000038 0.000012
-1
x10 1.82 1.22 0.91 0.70 0.56 0.45 0.37 0.31 0.26
-2
x10 4.04 3.35 2.96 2.68 2.47 2.30 2.15 2.03 1.92
-3
x10 6.33 5.64 5.23 4.95 4.73 4.54 4.39 4.26 4.14
-4
x10 8.63 7.94 7.53 7.25 7.02 6.84 6.69 6.55 6.44
-5
x10 10.94 10.24 9.84 9.55 9.33 9.14 8.99 8.86 8.74
-6
x10 13.24 12.55 12.14 11.85 11.63 11.45 11.29 11.16 11.04
-7
x10 15.54 14.85 14.44 14.15 13.93 13.75 13.60 13.46 12.34
-8
x10 17.84 17.15 16.74 16.46 16.23 16.05 15.90 15.76 15.65
-9
x10 20.15 19.45 19.05 18.76 18.54 18.35 18.20 18.17 17.95
-10
x10 22.45 21.76 21.35 21.06 20.84 20.66 20.50 20.37 20.25
-11
x10 24.75 24.06 23.65 23.36 23.14 22.96 22.81 22.67 22.55
-12
x10 27.05 26.36 25.96 25.67 25.44 25.26 25.11 24.97 24.86
-13
x10 29.36 28.66 28.26 27.97 27.05 27.56 27.41 27.28 27.16
-14
x10 31.66 30.97 30.56 30.27 30.05 29.87 29.71 29.58 29.46
-15
x10 33.96 33.27 32.86 32.58 32.35 32.17 32.02 31.88 31.76
(Source: Wenzel, 1942)
Cooper and Jacob (1946) noted that for small value of r and large volumes of t, the value of u is relatively
small. Consequently, the series terms in Equation 14.97 become negligibly small after the first two terms
and can be neglected.
From Equation 14.98

s    0.5772  loge (u) 
Q   r 2S 
s   0.5772  loge  
4πT   4Tt  
Q   r 2S 
or    log (1.7810)  loge  
 4Tt  
e
4πT 
Q  2.25 Tt 
 loge  2 
4πT  rS 
456
2.30 Q  2.25 Tt 
or s log10  2  (14.99)
4πT  rS 
Equation 14.99 shows that the drawdown (s) varies linearly with logarithm of time t because all other
terms are constant for an observation well.
Application
For the determination of the aquifer constants S and T, the following procedure is used.
 A plot of drawdown s versus the logarithm of t (straight line) is made (Fig 14.33).
 The value of change in the drawdown ( s) over the one log cycle is determined from the best fit
straight line.
Fig.14.33 Plot of drawdown against Log of time.

Source: (Namutete, 2006)
The value of T is determined from as follows:

2.30 Q
s1  s2  log10 (t 2 / t1 )
4πT
2.30 Q
Or s  log10 (t 2 / t1 ) , (for t2/t1 = 10.00 = one log cycle, log10 (t2/t1) = 1.0)
4πT
2.30 Q
Therefore, T
4π( s)
For the determination of S, extend the best-fit straight line to intersect the abscissa at t= t0. It may be
noted that t0 is the time up to which the drawdown remains zero.
From the Equation above, substituting s = 0 and t = t0
2.30 Q  2.25 T t 0 
0 log10  
4πT  r 2S 
457
 2.25 T t 0 
Therefore, log10    0  log10 (1)
 r 2S 
or (14.99a)
Knowing the value of T from the equation above, the value of S can be obtained.
This method should be used only for values of u less than 0.01 to avoid large errors.
Advantages of Cooper-Jacob equation
The Cooper-Jacob’s equation has the following advantages over Dupuit equilibrium equation
1. The assumption of steady-flow condition is not necessary.
2. A shorter period of pumping is required in the test.
3. Only one observation well is necessary, besides the pump well
4. The value of storage coefficient (S) can be determined in addition to that of T.
Example 14.10
Groundwater investigations were carried out at a borehole in Nkozi, Mpigi District, Uganda. An
observation well located 10m from the borehole was pumped at an average discharge of 2.38m3/h for 24
hours. The drawdowns at particular time intervals are given in the table below. Estimate the
transmissivity and storativity of the aquifer using Jacob’s method.
t (min) 2 5 14 35 80 150 270 360

Drawdown s(m) 5.11 11.55 21.64 29.87 39.28 42.71 45.82 48.04
t (min) 480 660 780 900 1020 1200 1440

Drawdown s(m) 54.97 54.98 55 55 54.99 55 55.01
The data is plotted with the drawdown as the ordinate and the log of time since pumping begun as the
abscissa as shown below:
458
Transmissivity:
Refer to the figure above
For the cycle Δt = 10-100min;
∆s = 38 - 20 = 18m
Q = 2.38m3/hr = 6.611 x 10-4 m3/s
2.3Q 2.3  6.611104

T   6.722 106 m 2 / s  0.581m 2 / day
4s 4   18
Storativity:
The borehole is 10m from the observation well so r = 10m and to = 0.68min
2.25Tt o 2.25  6.720  10 6  0.68  60

S 2
 2
 6.169  10 6
r 10
14.9.6 Chow’s Method

Chow’s method (1952) is a general method which can be used for all values of u. The method is the most
accurate of the three methods. Chow introduced a new function F (u), which is related to the well function
W (u) by the expression.
W(u)eu
F(u) 
2.30
Fig.14.34 shows the plot between F(u) and W(u) for different values of u. The values of u are written
along the curve
Fig.14.34 Plot of F(u) against W(u).
459
The following procedure is used for the determination of S and T

i). Plot the values of the drawdown s on an arithmetic scale against the time t on logarithmic scale on a
semi-log (Fig.14.35). Draw a smooth curve through the plotted points.
Fig.14.35 Plot of drawdown against Log of Time.

ii). Select an arbitrary point P on the curve and find out its co-ordinates s and t.
iii). Draw a tangent to the curve at P and determine the change in drawdown ( s) per log cycle of time
iv). Compute the value of F(u) from the following relation:
s
F(u)  (14.100)
Δs
where: s is the drawdown corresponding to point P.
v). For the value of F(u) computed in step (iv), determine the values W(u) and u from Fig. 14.34.
Alternatively, the values of W(u) and u may be determined by the following method without using
Fig. 14.34, provided that F(u) > 2.0.
It can be proved that for small values of u,
W(u)  2.30 F(u) (14.101)
2.30 Q 2.30 Q
T or Δs 
4π(Δs) 4πT
Q
From Equation 14.94, s W(u)
4πT
s W(u)
Therefore, F(u)  
Δs 2.30
Or W(u)  2.30F(u)
The value of W(u) is found from Equation 14.101 for the value of F(u) computed in step (iv). Then
the value of u can be found from the value of W(u) from Table 14.3.
vi). Compute the values of S and T from the values of W(u) found in step (v) from Equations 14.95 and
14.96.Thus
460
 Q  4 Tu
T   W(u) and S
 4πs  (r 2 / t)
There are two other common methods that will be considered here. These are the Recovery Test discussed
in Section14.9.9 and the Step drawdown Test discussed in Section14.9.11.
14.9.7 Radius of Influence

The Theis equation and its approximation given above predict that theoretically the cone of depression is
instantaneous, extending to infinite radius. However, in practice there is a radius of influence, beyond
which drawdown can be considered negligible at any given time t, although re (radius of influence will
increase with t), cf Section 14.7.3).
At any time t, let re = radius at which drawdown s = se, any arbitrary small value
Q
Since s W(u) (Theis Equation)
4πT
se 4πT
W(u)  at the radius of inf luence
Q
 s 4 πT 
 u  W 1  e   C , a constant, since se has been given a fixed arbitrarily small value
 Q 
[W-1 designates the inverse of W; i.e. “the value of u for which the value of the Well function W is, in
 se 4 πT 
this case,   ”]
 Q 
re2
Since u at the radius of influence from Equation (14.92)
4αT
4T
re   4αC 2 t
1 1 1 1
2
 ( .C) 2 t 2
S (101a)
and from Equation 14.99a at any value t
(101b)
which shows that re is proportional to t
1. The radius of influence develops rapidly at first, then more slowly as pumping continues.
2. re is large in aquifers with small S (storage coefficient) than in those with larger S. since S tends to be
smaller in confined aquifers than in unconfined aquifers, re tends to be larger in confined aquifers.
Equation (101b) has been used successfully as shown by (Moh et al., 1996) and several others.
14.9.8 Pseudo–Steady State

The Theis equation predicts that drawdown will change continuously with time, but in practice, there will
be an area near that can be assumed to be a steady state’ after pumping has continued for a reasonable
time.
461
For small r and/or large t, when u < 0.01, as shown previously, the Theis equation can be approximately
by:
Q r2
s   0.5772  loge (u)  where u 
4πT 4αT
For 2 observation Wells at r1, r2
Q
Drawdown at r1, s1   0.5772  loge u1 
4 πT
Q
and s2   0.5772  loge u2 
4 πT
Q Q  1 1 
 s1  s2  h2  h1    loge u1  (loge u2 )    loge  loge ) 
4 πT 4 πT  u1 u2 
loge  2  2  loge  2 
Q r 2Q r

4 πT  1 r 4 πT  r1 
loge  2   Thiem equation for steady flow
Q r

2πT  r1 
Therefore, under the assumed condition that u< 0.01, the approximate version of the Theis equation
reduces to the Theis equation already derived, implying that a ‘steady’ or Pseudo-steady state has been
reached. If u is then taken as <0.01 at r2 radial distance,
r2
u  0.01 can be used as a criterion for steady state is being established.
2
4αT
Q r2 r2
s where u  
2 2
Using the Theis equation W(u)
4πT 4αT 4Tt
The objective is to find values of transmissivity T and storage coefficient S, for an aquifer into which a
well has been sunk together with one or more observations wells (piezometer holes).The method is to
apply a stress to the aquifer (by pumping) and by analyzing the subsequent response (series of values of s,
drawdown and t, time), T and S can be evaluated.
The following solution techniques or methods of analysis (to determine S and T) employ the assumptions
implicit in the Theis Equation (14.94), which include:
i. Confined aquifer (although also used for unconfined aquifer if drawdown<< saturated thickness
ii. Unsteady state(drawdown is changing with time)
iii. Water removed from storage is discharged instantaneously with decline of head well diameter is
small (i.e. No storage in the well which can be represented as a ‘line sink’)
iv. When pumping starts at t = 0, discharge instantaneously reaches the pumping rate Q.
462
14.9.9 The Recovery Test

At the end of a pumping test, when the pump is stopped the water levels in pumping and observation
wells will begin to rise and this is referred to as recovery of groundwater levels. Measurements of
drawdown below the original static water level (prior to pumping) during the recovery period are known
as residual drawdowns.
If a well is pumped for a known period of time and then shut down, the drawdown thereafter will be
identical to one if the discharge had been continued and a hypothetical recharge well with the same flow
were superimposed on the discharging well at the instant the well is shut down.
From the above principle Theis showed that the residual drawdown s1 is given as:
(14.102)
r 2S r 2S
where: u and u 1 
4Tt 4Tt 1
For r small and t1 large, the well functions can be approximated by the first two terms of W(u) so that
(14.103)
1 1
Thus a plot of residual drawdown s versus the log of t/t forms a straight line (Fig.14.36) where:
where:
and ∆s1 is the slope.
Fig. 14.36 Plot of the residual drawdown against log(t/t1)

Example 14.11
The following drawdown recovery data was obtained during the testing of a borehole, located at Nkozi in
Mpigi District, Uganda after pumping for 24 hrs at an average discharge of 2.38m3/h. Estimate the
transmissivity using the Jacob Recovery Method.
463
Solution
From the data given Column 3 is added to the table by recording the time since pumping began. The
fourth column is added by calculating the ratio of total time since pumping started to the time since
pumping stopped.
Time since Residual Time since

pumping stopped t' drawdown pumping
(min) s'(m) started t (min) Ratio t/t'
1 52.22 1441 1441
3 47.84 1443 481
5 44.65 1445 289
7 43.29 1447 207
9 42.79 1449 161
12 40.94 1452 121
16 35.44 1456 91
20 30.9 1460 73
30 23.21 1470 49
40 18.62 1480 37
60 13.52 1500 25
80 10.75 1520 19
100 9.01 1540 15
150 6.95 1590 11
240 5.89 1680 7
The plot of the residual drawdown as the ordinate and the ratio t/t' as the abscissa, is drawn as shown in
the figure below.
Δs
’
Δt
From the above plot the residual drawdown in a log cycle is obtained and applied in Jacob’s Recovery
formula to estimate the transmissivity as shown below.
For the cycle t= 100-1000min;

464
Q = 6.611x 10-4 m3/s

∆s′ = 57-32 = 25m
2.3Q 2.3  6.611104

T   4.84 106 m 2 / s  0.418m 2 / day
4s '
4    25
14.9.10 Applicability of different Methods

Tindimugaya (2008) used seven different pumping test analysis methods (Theis, Cooper- Jacob, Theis
Recovery, Hantush, Neumann and PTFIT Program) to determine the aquifer hydraulic properties in
Wobulenzi, central Uganda and Rukungiri, western Uganda). In Wobulenzi pumping test data from
weathered and fractured bedrock, gave high variability in the estimates for transmissivity and storage
coefficient. This shows that estimates of the hydraulic properties are sensitive to the method employed.In
contrast, for the alluvial aquifer in Rukungiri the estimates of the hydraulic properties were fairly
consistent and are not very sensitive to the method employed. This suggests the necessity to have
knowledge of the aquifer geometry and groundwater flow in order to select the appropriate pumping test
solution. The results also suggest that the commonly used analytic solutions for confined and leaky
aquifer conditions (Theis, Cooper Jacob and Hantush methods) overestimate transmissivity values for the
leaky fractured bedrock , because they underestimate leakage from overlying layers.
14.9.11 The Step-drawdown Test

During a step-drawdown test, the pumping rate is increased at regular intervals for short time periods. The
typical step-drawdown test lasts between 6 and 12 hr, and consists of three to five pumping rates as
shown in Fig 14.37. The well is initially pumped at a low rate of 6 m3/s until the drawdown within the
well essentially stabilizes. The discharge is then increased to 12 m3/s and then to 18 m3/s in subsequent
steps.
1
0.9 18 m3/s
0.8
Drawdown (m etres )
0.7
0.6
12 m3/s
0.5
0.4
6 m3/s
0.3
0.2
0.1
0
0 50 100 150 200 250 300
T im e (m inute s)
Fig.14.37 Plot of drawdown against time: Stepdrawdown pumping test.
465
Because step-drawdown pumping tests are typically much shorter than constant-rate pumping tests,
transmissivity and storativity values are not as accurate for these tests. When the discharge is divided by
drawdown in a pumping well (sw/Q), the specific capacity of the well is obtained and this is a measure of
the productivity of the well. The larger the specific capacity, the more productive the well is. For a given
discharge, the specific capacity of the well is assumed to be constant. A reduction in specific capacity
may be caused by the lowering of the groundwater level in an unconfined aquifer, or to an increase in
well loss associated with clogging or deterioration of the well screen.
Step-drawdown tests can be used to determine the variation in specific capacity of a well and different
values are plotted against discharge. This is illustrated in Example 14.12 and Fig 14.39.
14.9.12 Step Drawdown Pumping Test Analysis

Assumptions
i. The basic assumption in all step-drawdown test analysis is that the increase in discharge rate at the
beginning is equivalent to a new pump in the well with a discharge rate equal to the rate of increase
in discharge.
ii. The second assumption is that the total drawdown in a step-drawdown test is equal to sum of the
individual drawdowns caused by the theoretical pumps responsible for each discharge in the test.
iii. Except in the first step of the test, this superposition of effects of incremental changes in discharge
rate ignores a basic assumption underlying all well tests. The piezometric surface should be at rest
and in the horizontal plane prior to pumping.
Jacob (1948) noted that part of the loss of head or drawdown occurs in the aquifer formation, where
energy is expended in overcoming frictional resistance of the soil against slowly moving water. A smaller
part of the head loss occurs as water moves at relatively high velocities through the screen and upward
inside the casing to the intake of the pump. This head loss approached some higher power approaching
two and is written as follows:
(14.104)
Where B is the resistance of the formation; this factor represents the total hydraulic resistance of the
formation from the well face to some distance, where the head drop is virtually zero and where radial
motion has not yet began. The resistance B increases with time as the ever expanding area of influence of
the well expands, while C is the well loss factor as shown in Fig 14.38. The total drawdown consists of
both the formation loss and the well loss. According to Rorabaugh (1953) the aquifer losses are laminar
flow, while the well losses are turbulent flow. Bierschenk and Wilson (1964) simplified equation (14.104)
and divided by Q. This becomes:
(14.105)
The arithmetic plot of specific drawdown (sw/Q) against a discharge rate, will therefore, be a straight line
with a slope of C and an intercept of B. The total drawdown for each discharge rate is then the sum of the
incremental drawdown at that step plus all previous steps.
The total drawdown sw at the well can also be written for the steady-state confined case
(14.105a)
where: n is a constant greater than one C is a constant governed by the radius, construction, and condition
of the well. For simplicity, let
466
so that
(14.106)
Q
Ground Surface
Original piezometric surface
Formation loss BQ sw Drawdown Curve

n
Well loss CQ
Confined aquifer
Fig 14.38 Formation loss and well loss in a pumping test.
Brereton’s (1979) method is also based on Jacob’s approximation of the Thesis well equation for u < .01.
For transient flow in a confined aquifer, in the absence of any well losses, or in an unconfined aquifer
where the drawdown sw is very small compared to H the saturated thickness, the time drawdown
relationship is:
(14.107)
or (14.107a)
where:
and
The term alogbt is equivalent to the aquifer loss coefficient B and is time dependant. Substituting
alogbt into Equation (14.104) gives:
(14.107b)
which also provides the drawdown at each step as a function of discharge and the aquifer and well loss
coefficients respectively.With further analysis and taking account of the duration of pumping in each step
and by considering the drawdown at the beginning of the test step gives an equation of the form:
467
S m  a[Qm log bt  Dm ]  CQ 2 (14.108)

where: Sm is the drawdown for each step m; Qm is the pumping rate during each step and Dm is a
function of the drawdown at the beginning of the particular step.
This equation can be solved analytically by taking the slope of the time drawdown curve when data is
plotted relative to the origin of the step on semi-logarithmic graph paper. From the relationship:
(14.109)
where ΔSos is the drawdown per log cycle. The term a can be calculated, thus T for each step may also
be estimated and then T for whole test calculated. Equation (14.108) can be rewritten as:
S m  aDm
 a log bt  CQm
Qm (14.110)
S m  aDm
A plot of Qm against Qm will give a line of slope C and intercept B = alogbt. Knowing a and t
allows b to be determined.
An interactive computer model was developed (Rugumayo, 1988) to analyse pumping test data, using the
Brereton method. It estimates aquifer coefficient B, well loss coefficient C, and coefficients a and b. It
can also be used predict drawdown at a specified pumping rate, after a specified pumping period and
estimate the efficiency of the well.
In comparison with the Bierschenk and Wilson method, Brereton treats each step as an individual event,
with its own behaviour and response to pumping, and obtains it own value for T and B and C whereas
Bierschenk and Wilson uses the drawdown at the end of each cycle, after which B and C are estimated.
Brereton’s method in addition provides better prediction results than other methods, like Bierschenk and
Wilson or Rorabaugh.
Example 14.12
Data from a step drawdown pumping test were obtained as follows;
Step sw(m) Q (m3/day) (sw/Q)x10-3

(day/m2)
1 0.615 84.4 7.29
2 1.260 151 8.34
3 1.380 200 9.15
4 2.410 250 9.64
Estimate i) the formation coefficient and ii) the well loss coefficient using the Bierschenk and Wilson
method.
468
Solution
The data provided in column 2 is the drawdown at the end of each pumping cycle and according to
Bierschenk and Wilson, the drawdown at the end of each cycle is the sum of the incremental drawdowns
of previous steps plus that step and thus Equation 105 applies.
The specific drawdown (sw/Q) is calculated and is shown in the fourth column.
The plot of specific drawdown (sw/Q) against discharge Q is drawn and shown in Fig.14.39.
Fig.14.39 Specific Capacity (Sw/Q) against Discharge Q.
From the plot the slope is calculated and the value of B aquifer coefficient is obtained as
14.5x10-6day2/m5, or in SI Units, 108, 242 s2/m5.
From the intercept the value of C well loss coefficient is also obtained as 6.1x10-3 days/m2, or in SI Units,
527 s/m2.
14.10 Some Recent Groundwater Studies in Uganda

Some recent studies were carried out in Hoima, Mpigi and Wakiso districts in Uganda whose relative
location is shown in Fig. 14.40.
The Aquitest Program was to determine the sustainability of groundwater as a source of water
(Rugumayo and Kibera, 2004) for Bugahya County in Hoima District in western Uganda. It was based on
pumping test data from 12 boreholes.
It was observed from the lithological profiles that the aquifer is generally overlain by a depth of relatively
impermeable material of clay at average depth of 20 m implying that the aquifer in Bugahya County is a
confined one. A sample of some of the lithological profiles, are shown in Fig 14.41. Fig 14.42 gives a
map showing the borehole locations, equipotentials and the direction of groundwater flow.
Three methods were used to determine the aquifer parameters, namely the Theis, the Cooper- Jacob and
the Theis Recovery. The results showed a comparable range for the specific parameters of transmissivity
T, hydraulic conductivity K and storativity S. A map showing groundwater yield distribution in Bugahya
County was developed and is shown in Fig 14.43.
In a study (Kigobe and Kizza, 2006) geostatistical methods were employed to estimate unsampled areas
and measure the hydraulic parameters of transmissivity T and storativity S, in Mpigi and Wakiso districts,
where there is limited availability of data. The results showed that the underlying aquifer systems are
mainly semi-confined and leaky, with low hydraulic conductivity. Maps showing hydraulic and
transmissivity of the area were developed and are shown in Fig.14.44a and Fig 14.44b respectively.
469
In a study (Kigobe and Kakimpa, 2007), a groundwater conceptual model using MODFLOW was
developed by integrating and coupling a physically based distributed model and geographic information
systems, to simulate potential decline in hydraulic equipotentials due to the different levels of abstraction
in the vicinity of Inner Murchison bay of Lake Victoria. The results showed the variation in head travel
distances, over 10 years and the delineation of well head protection zones or well capture zones.
Furthermore, they also showed that the wells collect water from the offshore extents of the aquifer other
than the bay.
Fig 14.40 Map of Uganda showing the relative locations of Bugahya County and Mpigi and Wakiso Districts
470
KEY
Clay
Weathered rock
Shale
Granite
Sand/gravel
Basalt
Fig. 14.41 Samples of lithological profiles of boreholes in Bugahya County.
471
Fig.14.42 Borehole locations, equipotentials and groundwater flow in the model domain.
Fig.14.43 Groundwater Yield Distribution in Bugahya County, Hoima District.
472
32
N
WAKISO
DWD 17054 â â
DWD 17345 DWD 17056

â
DWD 17446 â
MPIGI â DWD 17351
DWD 17349
Kampala City
â
DWD 17350
â DWD 17448 â â DWD 16272
DWD 17352 â â
DWD 17445 ââ DWD 17053
DWD 16273 DWD 17353
â â DWD 17447
T [sq m/d]
0 0
0.0 - 0.5
0.5 - 1.0
1.0 - 1.5
1.5 - 2.0
2.0 - 2.5
2.5 - 3.0
3.0 - 3.5
3.5 - 4.0
4.0 - 5.0
Well
LAKE VICTORIA
#
L.Victoria
32
Fig.14.44a Transmissivity for Mpigi and Wakiso districts.
32
N
WAKISO
DWD 17054 â â
DWD 17345 DWD 17056

â
DWD 17446 â
MPIGI â DWD 17351
DWD 17349 DWD 17350

â â DWD 17448 â
DWD 17352 â â â DWD 16272
DWD 17445 ââ DWD 17053
DWD 16273 DWD 17353
â
â DWD 17447
K [m/d]
0 0 0.003 - 0.02
0.02 - 0.04
0.04 - 0.06
0.06 - 0.08
0.08 - 0.10
0.10 - 0.12
0.12 - 0.14
â Kvalues
L. Victoria
32
Fig.14.44b Hydraulic conductivity for Mpigi and Wakiso districts.
In a study (Mangeni and Katashaya, 2006) analysed data from the National Oceanic and Atmospheric
Adminstration, Advance Very High Resolution Radiometer (NOAA, AVHRR) infra-red images to
investigate the groundwater situation of Lake Victoria. Surface temperature maps were produced using
the Split Window Technique (SWT). The surface temperature maps were then analysed for warm and
cold season patterns in order to infer groundwater inflow into Lake Victoria. The results showed
warm/cool season patterns indicating groundwater inflow into the lake. These results were in good
agreement with the IAEA isotope study to improve the Lake Victoria water balance.
473
The direction of groundwater flow was confirmed in research (Mwebaze, 2007), while using Tritium
isotopes to estimate groundwater mean residence times and stable isotopes and hydrogeochemistry to
assess surface – groundwater interactions. The study also noted few cases of surface groundwater mixing
in the vicinity of the lake. Stable isotopes and hydrogeochemistry showed that groundwater in the basin
was being recharged by rainfall and the larger portion was in the wet seasons of September – November
and March - May. Furthermore, the analysis of groundwater mean residence time, three major water ages
were identified. These were; i) less than 10 years (recent recharge); between 30 and 50 years, (modern
recharge) and above 60 years (sub modern). It was established that recent and modern groundwater
accounted for more than 60% of the samples analysed for residence times. In a study on groundwater-
surface interactions it was found that regional changes in groundwater storage indicated by borehole
hydrographs at 10 stations in Central Uganda correlate well (r2=0.6) with large scale satellite observations
(GRACE) of changes in total water storage (Owor et al., 2008). Their daily observations of the hydraulic
gradient between groundwater and Lake Victoria at Entebbe confirm that groundwater primarily
discharges into the lake.
Some of the challenges experienced by the management of groundwater resources in urban areas have
been well documented (Tindumugaya, 2005). These include the lack of sewerage systems which have led
to the use of on-site sanitation systems, like septic tanks and pit latrines that have caused contamination of
groundwater resources in many areas. Substantial efforts are required to improve groundwater
management and governance in urban areas of Uganda in form of groundwater regulation and monitoring
and assessment of groundwater resources. This will require the involvement of urban water authorities
and water users. An appropriate institutional framework for groundwater resources management in the
urban areas should be developed reinforced by an adequate awareness raising programme on water
resources management to enable the water users appreciate the benefits of groundwater protection.
An example of this is in South Africa, where in order to promote the sustainable use of groundwater, a
Multi Criteria Decision Analysis (MCDA) tool was developed (Pietersen, 2006) and modelled based on
the National Water Act 1998. The key sequential steps included; i) facilitating democratic decision
making models, ii) involvement of appropriate users in technology selection, iii) development of an
appropriate tariff structure with the users, iv) delineating protection zones, v) developing regional
groundwater flow conceptual model parameters, vi)develop artificial water harvesting systems, vii)
considering water conservation and demand side management, viii) developing proactive monitoring
systems for drought and finally, ix) implementing a water services programme at community level.
14.10.1 Groundwater Recharge

Groundwater Recharge can be defined as the entry of water into the saturated zone of water available at
the water table surface, together with the associated flow away from the water table within the saturated
zone.
These can be quantified by the use of a hydrologic balance equation or water balance equation. Let us
consider a watershed where there is no external flow and the surface water and groundwater divide
coincide, then the water balance equation for the annual period can be written as:
P = Q + E+ΔSS+ΔSG (14.111)
Where P is the precipitation, Q is the runoff, E is the evapotranspiration and ΔSS, the change in storage of
the surface-water reservoir, and ΔSG, the change in storage of the groundwater reservoir (both saturated
and unsaturated) during the annual period.
If we average over many years of record, it can be assumed that ΔSS = ΔSG= 0 and the above equation
becomes
P = Q+E (14.112)
474
where: P is the average annual precipitation, Q the average annual runoff and E, the average annual
evapotranspiration. If we consider a watershed with a recharge area as shown, the hydrologic budget
equation can be rewritten as:
P = QS+R+ER (14.113)
Where, QS is the surface water component of average annual runoff, R the average annual groundwater
recharge and ER, the average annual evapotranspiration from the recharge area. In the development and
sustainable use of groundwater resources, the knowledge recharge is very essential and especially since,
over the two decades, there have been increased investments by government and nongovernmental
organizations in Africa, in providing potable water. A study of groundwater recharge estimates (Taylor,
1996) was carried in the Aroca catchment of central Uganda, north of Lake Kyoga, using the soil
moisture balance method and supported by stable isotope data and groundwater flow modelling. The
recharge values average 200 mm a year and are more dependent on the number of heavy (more than 10
mm a day) rainfall events, than the total annual volume of rainfall (Owor et al., 2008).The stable isotope
data also show that recharge occurs during the heaviest rainy season and is derived directly from
infiltration of rainfall.
They further demonstrated that recharge occurs in most years but varies according to fluctuations in
annual precipitation and that recharge pulses occur twice each year and respond to the bimodal
characteristics of rainfall. In the Nyabisheki catchment the recharge estimates were 34 mm a year. More
recent studies (Kigobe, 2008) estimated groundwater recharge using a semi distributed hydrological
model (SWAT soil water assessment tool), applied to Mpologoma catchment. The model was calibrated
against observed streamflow for the 1970-1974 period. The results in terms of simulated streamflow,
recharge and groundwater flow showed reduced groundwater storage in the valleys and are controlled by
the two recharge (rainfall) seasons per year.
In a study of Mukono District near the shores of Lake Victoria in Uganda (Murangira, 2007) receives
substantial amounts of rainfall 1566 mm/year and has high recharge potential and is greater than the
public water demand, implying that aquifers are not yet over exploited by current abstraction projects.
Furthermore, the area is largely covered by weathered crystalline rocks which form the fracture aquifers
and Quaternary deposits that form the confined porous aquifers of gravel formation, which are the source
of water supply for the communities. He noted that the groundwater in Mukono Distrct flows in two
directions; towards Lake Kyoga and towards Lake Victoria and there exists a groundwater divide between
the two basins.
14.10.2 Groundwater Quality

Water quality involves the study of all substances occurring as contaminants of pure water. Groundwater
is primarily a source of supply for:
I. Human consumption (portable supply)

II. Agriculture (livestock, irrigation)
III. Industry (cooling water, manufacturing etc.)
Therefore, depending on the use, the chemical quality of groundwater is of basic importance. The quality
of groundwater is however generally good enough for most purposes with little treatment. Running costs
are therefore low for groundwater as a source of supply. The quality of groundwater concerns the
engineer in three broad aspects:
I. Its suitability as a source of water at particular points

II. How engineering works or anthropogenic interference may affect that quality
475
III. How the natural quality of the water can be used in groundwater exploration and development.
Suitability of groundwater as a source of water
The suitability of groundwater as source of water is judged based on:
I. Public water supply taking in consideration public health. This includes biological/organic/trace
element analyses to high accuracy.
II. Agricultural supply which includes livestock and crops (irrigation). Stringency of criteria varies
with the situation (e.g. Boron level critical for various crops, sodium critical for clay suspension.
III. Industrial supply: Different
Engineering works and their effects on groundwater quality
These are important in so far as treatment may be required to restore the water as a particular source of
supply in relation to:
i. Agriculture: quality is affected by interference from, e.g.

 Animal waste
 Fertilizers
 Insecticides/ herbicides
ii. Waste disposal

 Septic tank
 Sewage, sludge lagoons
 Rubbish tips
 Urban runoff
 Industrial waste
 Mining (acid mine drainage, tailings ponds)
 Cemeteries
iii. Sea water intrusion in coastal areas.
Natural quality of groundwater
The Natural chemical analysis of groundwater can be used as a means of identification. Spot sample
analyses from boreholes, maps of certain ion concentration levels, etc can be of use in exploration,
directing good targets for water supply. Chemical analyses also give clues as to the geological source of
the water (e.g. in mine dewatering or in underground flooding). In other words, the chemical identity of a
given groundwater can be useful in tracing its source and movement (e.g. groundwater tracing in Karst
terrain, tracing pollutant sources, tracing salinity movement in seawater intrusion). Conversely, a
groundwater may be imprinted with a readily identifiable character (dyes or radioactive tracer) in order to
monitor its movement.
In a study (Nyende et al., 2007), the hydro-chemical evaluation of groundwater from the regolith and
basement aquifers were undertaken in Pallisa District and the groundwater samples analysed for minor
and major elements. This was to enable the quality of groundwater presently being pumped from the
fractured bedrock aquifer to be compared to the quality of water that could be expected from the
development of the regolith aquifer. The results showed that electrical conductivity values (and thus
476
degree of mineralisation) are significantly lower for springs than for dug wells which in turn are clearly
lower than boreholes .On the other hand, boreholes clearly show the highest degree of mineralization,
with relative enrichment of nearly all tested elements. The above results however, do not apply to nitrates
and suggest that non-geological factors influence nitrate concentration. It was also observed that the
boreholes situated in these aquifers represent a regime, where there is calcium enrichment. Furthermore,
both the regolith and bedrock groundwaters are dominated by the carbonate ions. The overall
groundwater quality assessment shows that the boreholes in Pallisa District are safe for drinking purposes
except for a few areas of high iron concentration.
In a study on phosphorus transport in groundwater of peri-urban waters of shallow wells, in the Bwaise
suburb of Kampala (Kulabako et al., 2007) showed that there were seasonal variations of the phosphorus
content that corresponded to rainfall events. The high phosphorous content was linked to waste disposal
animal rearing activities, highly porous vadose zone and poor drainage conditions in the area resulting in
flooding. Transport mechanisms of phosphorous appeared to be a combination of adsorption, precipitation
leaching from the soil media and colloids, with leaching playing a more significant role, especially in the
wet season.
14.11 Geophysical Methods

Geophysical surveys can be used to study many subsurface geologic and hydrologic problems especially
in the determination of the availability of water. In some cases hand augured (shallow wells) or deep
boreholes provide better information. Sometimes a combination of the drilling and geophysical
measurements may provide an optimal solution.
These surveys require skilled personnel to conduct them and make interpretations and usually computer
programs are used for the data analysis. Good interpretation of the data requires knowledge and
experience of both geology and geophysics. Geophysical surveys are of two types; surface and borehole.
(USGS, 1980)
14.11.1 Surface Geophyisical Methods

Surface geophysical surveys fall into three main categories namely; i) seismic ii) gravimetric and iii)
electrical resistivity. The first two methods are briefly discussed, while third is more detailed because of
its applicability.
i) Seismic – The basis of this method is the timing of artificially generated pulses of elastic energy
propagated through the ground and picked up by electromechanical transducers (geophones) operating as
detectors. These geophones respond to the motion of the ground and their response is transformed into
electrical signals and recorded on magnetic tape. The geophysicist is interested in the propagation
velocity and the geometry of the propagation path, which will guide him on the availability of rock. The
determination of these parameters is complex.
ii) Gravimetric – Gravimetry is the geophysical measurement of the acceleration due to gravity. It is
based on two laws of physics; The Law of Universal Gravitation and the Newton’s Second Law of
Motion. In gravimetric studies, the local vertical acceleration is measured. The maximum gravity
difference between the Earth’s normal field and that actually measured on the surface and corrected for
altitude and latitude is called the gravity anomaly and it reflects lateral density variations in rocks to
depths of several kilometres.
iii) Electrical Resistivity. In groundwater investigation resistivity is controlled more by porosity, water
content and water quality than the resistivities of the rock matrix. The higher the porosity of saturated
477
rock, the lower is the resistivity. The electrical properties of most rocks depend on the amount of salinity
and distribution of water in the rock. Saturated rocks have lower resistivity than unsaturated and dry
rocks. Clays and conductive minerals reduce the resistivity. For example, due to the good electrical
conductivity of groundwater the resistivity of a sedimentary rock is much lower when it is waterlogged
than in the dry state.
The electrical resistivity method measures changes in the apparent electrical resistivity with depth and the
information obtained builds up a geo-electric model across the site. Thus variations in the electrical
properties of water bearing strata are obtained.
An electric current either DC or very low frequency AC is introduced into the ground by means of two
outer metal stakes or electrodes. Variations caused by hydrogeological conditions affect the subsurface
current flow and alter the electric potential patterns associated with them. These patterns are measured by
a second inner pair of electrodes in line with the first. If the resistivity is uniform everywhere in the
subsurface zone beneath the electrodes; an orthogonal network of arcs will be formed by the current and
equipotential lines as shown in Fig. 14.45.
Current Current
source meter
Volt meter
C1 P1 P2 C2 Surface
Current
Current flow
through earth Voltage
Fig.14.45 Electrical Resistivity operating principles.

Source: (Lowrie, 1997)
where: C1 and C2 are current electrodes and P1 and P2 are potential electrodes
For a semi-infinite medium, which is the simplest earth model, and with both current and potential point-
electrodes placed at the Earth’s surface,
(14.114)
where: V is the electric potential at any point 

I is the current
AM is the distance on the Earth’s surface between the positive current electrode A and the
potential electrode M.
The value of the apparent resistivity measured in the fields is a function of several variables; the electrode
spacings, the geometry of the electrode array, and the true resistivities and other characteristics of the
subsurface materials, such as layer thicknesses, angles of dip, and anisotropic properties (Lowrie, 1997).
478
iv) Electrode Configuration

The two most commonly used configurations for the four electrodes are the Wenner and the
Schlumberger arrays as shown in Fig 14.46. In the Wenner Array, the current and potential electrode pairs
have a common mid-point and the distances between the electrodes are equal. In the Schlumberger Array
the current and potential pairs of electrodes also have a common point but the distances between adjacent
electrodes differ.
Increasing the electrode spacing increases the depth of current penetration and so a depth probe can be
completed in which the variation of apparent resistivity with depth can be measured.
The results are usually plotted as a curve of log apparent resistivity against log of some function of
electrode spacing. The curve can be interpreted to find the absolute resistivity of each ground layer and
the depth from the surface of the interface between each layer. A frequent method of interpretation is by
using standard comparison curves.
Wenner Array Schlumberger Array
a
a a a M N
BA s s
M N B A B
V  s 
2 a2  V
 a  2a a  4
i a i
Fig.14.46 Electrode Configurations for the Resistivity Method
The curve obtained in the field is plotted on log/log graph paper with the same cycle size as the standard
curves which illustrate calculated curves for up to a four layer model earth. By using a simple matching
technique the standard curve nearest to the field curve is found and the values of the absolute resistivities
and interface depths calculated (Lowrie, 1997).
Other Applications in Groundwater Studies

a)Buried stream channels can be mapped accurately by the resistivity method using horizontal profiling,
electrical soundings or both. Lateral profiling furnishes on the presence or absence of shallow buried
stream channels while electrical soundings determine the depth preceding and following the lateral
profiling survey. The Wenner configuration is best adapted for lateral profiling. The assemblage of four
electrodes is displaced stepwise along a profile while maintaining constant values of the inter-electrode
distances corresponding to the configuration employed. The separation of the current electrodes is chosen
so that the current flow is maximized in depths where lateral resistivity contrasts are expected. Results
from a number of profiles may be compiled in a resistivity map of the region of interest. The regional
survey reveals the horizontal variations in resistivity within an area at a particular depth.
b)The depth of the water table overlain and underlain by several layers of different resistivities can be
determined by Vertical Electrical Sounding (VES) by observing the variation of resistivity with depth.
The Schlumberger configuration is most commonly used for VES investigation. The midpoint of the array
is kept fixed while the distance between the current electrodes is progressively increased. This causes the
current lines to penetrate to greater depths depending on the vertical distribution of conductivity. Before
the advent of portable computers two layer cases were interpreted with the aid of characteristic curves.
479
These theoretical curves, calculated for a particular four-electrode array, take into account the change of
depth penetration when current lines cross the boundary to a layer with different resistivity (USGS, 1980).
An example of Sclumberger soundings is shown in Fig 14.47.
c) Limitations
It has been noted that the interpretation of a multilayer sounding curve is not unique, which means that a
given electrical sounding curve can match a number of subsurface distributions of layer thicknesses and
resistivities. Other limitations are inherent in the conventional methods of electrical soundings. This
however, does not mean that the interpretation of electrical soundings is a futile exercise. In practice, it is
by the correlation of different sounding curves, by making crossed soundings, by sounding with different
arrays, by traversing the area with horizontal resistivity profiles, by knowledge of the general geology of
the area, by recognition of the electrical properties of the area that correct interpretation is achieved.
Fig.14.47 Schlumberger resistivity soundings for the Buranga Geothermal Prospect.

Source: (BMZ/GSMD, 2007)
14.11.2 Borehole Methods

Geophysical logs are important means for the hydrologist to acquire data on the groundwater
environment. They often provide the only practical measurement and of the undistributed surface
sediments. They are used to determine geologic correlations, lithology, geometry, resistivity formation,
resistivity factor, bulk density, porosity, permeability, moisture content, specific yield of water bearing
rocks. They can be used to define the source movement and chemical and physical characteristics of
water.
14.12 Isotopes in Groundwater Studies
All chemical elements have two or more isotopes. The formula H2O is a gross simplification from the
structural viewpoint as well the atomic view point. Natural water can be a mixture of the six nuclides as
shown in Table 14.4. Eighteen combinations of H – O – H are possible using these nuclides.
2
H216O,1H218O, 3H217O are some of the examples of molecules that are in water and the most common is
2
H216O.
480
The chemical behaviour of isotopes of the same element is broadly similar since the chemistry of the
atom is influenced by the electronic shell configuration and not the constitution of the nucleus. The
physical properties of different isotopes will be influenced by their nuclear masses.
Absolute ratios of stable isotopes are not normally used to express isotropic compositions of water
samples, but are expressed as deviations from a standard, namely Standard Mean Ocean Water (SMOW)
explained later.
Since the early 1950s naturally occurring isotopes that exist in water in the hydrologic cycle have been
used in investigations of groundwater and surface water systems. Tritium and Deuterium can be
represented as T and D respectively. 3H and 14C are used to guide the age of groundwater. 18O and 2H
serve mainly as indicators of groundwater source areas and evaporation indicators in water resource
studies.
Table 14.4 Isotopes of Hydrogen, Oxygen and Carbon
Isotopes Relative Type

Abundance
1
H Proteum 99.984 Stable
2
H Deuterium 0.016 Stable
3 -15
H Tritium 0-10 Radioactive (half-life 12.3 years)
16
O Oxygen 99.76 Stable
17
18
14
C Carbon <0.001 Radioactive (half-life 5730 years)
a) Carbon-14
Before the thermonuclear explosions in 1953, 14C in the global atmosphere was derived entirely from the
natural process of nitrogen transformation caused by the bombardment of cosmic rays
Based on the law of radioactive decay, which describes the rate at which, the activity of 14C and all other
substances decrease with time and is expressed as:
A= A02-t/T (14.115)
where: A0 is the radioactivity level at some initial time, A the level of radioactivity after time t, and T the
half-life of the isotope. This law when applied with the measurements of groundwater can be used to
estimate the age of groundwater. The age specifically refers to the time that has elapsed since the water
moved deep enough into the groundwater zone to be isolated from the earth’s atmosphere. When water
moves below the water table and becomes isolated from the earth’s CO2 reservoir, radioactive decay
causes the 14C content in the dissolved carbon to gradually decline. To determine useful information from
this type of data, it is necessary to determine the source of the inorganic carbon.
b)Tritium
Both man-made and natural sources contribute to the availability of tritium in waters in the hydrological
cycle. Tritium is produced naturally in the earth’s atmosphere by the interaction of cosmic-ray-produced
neutrons with nitrogen. The first major source of man-made tritium entered the atmosphere through the
initial tests of large thermonuclear devices in 1952. Since that time, the Tritium content in precipitation
481
has been monitored in many locations in the Northern Hemisphere, and fewer in the Southern
Hemisphere. At a given latitude, concentration of tritium in precipitation in sampling stations near the
coast are lower than those inland because of dilution of oceanic water vapour, which is low in Tritium.
Measurements of Tritium concentrations can assist in groundwater investigations. For instance if in a
sample of groundwater from the Northern Hemisphere contain a high concentration of tritium, it is
evident that a large proportion of this water entered the groundwater zone sometime after 1953. Tritium
from detailed measurements can also be used to distinguish different age zones within the modern part of
groundwater flow systems. It can also be used to estimate the time of recharge to a system, infiltration
studies, lakes mixing studies and in oceanography (IAEA, 1968, 1978).
c) Oxygen and Deuterium

Since the early 1950’s when the mass spectrometer became in use it has been possible to make accurate
measurements of isotope ratios. Hydrologists in particular are interested in the ratios that comprise the
water molecule, 18O/16O and 2H/1H. The isotope ratios are expressed in delta (δ) units as per mille (parts
per thousand or 0/00) differences relative to an arbitrary known standard, which has been used for many
years namely the Standard Mean Ocean Water (SMOW);
δ0/00 = [(R – Rstandard)/Rstandard] 1000 (14.116)
where: R and Rstandard are the isotope ratios2H/1H or 18O/16O, of the sample and the standard respectively.
The different isotopes of water have different vapour pressures and freezing points. The two properties
cause differences in 18O and 2H concentrations in various parts of the hydrologic cycle is an example of
isotropic fractionation, which is the process where the isotopic content of a substance changes as because
of evaporation, condensation, freezing, melting, chemical reactions or biological processes.
For instance, when water vapour condenses the rain or snow that forms has higher 18O and 2H
concentrations than the remaining water vapour. This process is repeated many times as part of the
regional and atmospheric circulation systems, to the extent that the content of 18O and 2H at a given
locality, is dependent upon the condensation-precipitation history of the atmospheric water vapour.
Where temperatures are above 50-100oC, like in deep subsurface zones, the 18O and 2H content of
groundwater can be altered significantly, because of being in contact with the host rock. At shallower
depths, the concentration of these isotopes is hardly affected by chemical processes.
Other applications of 18O and 2H are; the determination of areas and sources of recharge; infiltration from
lakes and reservoirs as a source of recharge; tracing the origins of salinity; flow distribution in swamps
and; the origin of geothermal waters; leaking dams; the interaction between aquifers; paleaoclimatic
investigations; dewatering of mines; glaciology and runoff from glaciers; hydrology of permafrost areas
and leach water from land tips (IAEA, 1968, 1978).
Recent studies to assess the interactions between groundwater, surface water and wetlands used isotopes
techniques (Mukwaya and Tindimugaya, 2008). This was based on a groundwater monitoring well
constructed in an alluvial aquifer 140m from Lake Victoria, which showed different responses from Lake
Victoria after July 2005.The well showed a similar response to four other wetlands in the dry and wet
periods, showing that the wetlands and the well are hydraulically connected. The response of the lake
water levels corresponded to two of the wetlands, showing they are hydraulically connected.
14.13 Geothermal Waters

In some countries of the world like USA, New Zealand, Iceland, Kenya and Uganda there exist thermal
springs which discharge water having a temperature in excess of the local groundwater commonly called
hot springs. Most waters from hot springs are highly mineralized and contain mostly meteoric water,
which has been modified in quality by its passage underground. This hydrothermal phenomena, which
involves the release of water and steam is nearly always associated with volcanic rocks and tends to be
concentrated in regions where large thermal gradients occur (Todd,1980). In USA, New Zealand, Iceland
482
and Kenya the geothermal waters have been used to generate considerable amounts of energy for
domestic and commercial use.
Some recent geochemical and isotope geochemical analyses have for the Buranga Prospect in the Semilki
valley adjacent to the Ruwenzori Mountains of western Uganda were conducted by the Geological Survey
and Mines Department. From isotope hydrological investigations as shown in Fig. 14.48, it was
concluded that the source of Buranga thermal water is of meteoric origin. The absence of a positive shift
in δ18O suggests a reasonably high permeability. The recharge cannot come from the immediate
neighbouring Rwenzori mountain due to the heavier isotopic composition of the local rivers. Therefore
the more distant high Rwenzori Mountains were suggested to be favourable recharge areas. The resulting
long flow path is compatible with tritium measurements giving >50 years residence time for the thermal
waters (pre-bomb-values) and a few tritium units for the local river waters (BMZ/GSMD, 2007)
Fig.14.48 Stable isotopic composition of Buranga thermal waters and local river waters
Source: (BMZ/GSMD, 2007)
Fig 14.49a A conceptual model of the Buranga Geothermal Prospect from the south
Source, (BMZ/GSMD, 2007)
483
Fig 14.49b A conceptual model of the Buranga Geothermal Prospect from the north.
Source: (BMZ/GSMD,2007)
Figs14.49a and 14.49b show conceptual models of the Buranga Geothermal Prospect taken from the
south and north directions respectively. The local reservoir feeding the hot springs has temperatures of
120-130oC. The thermal water flows from great depths along the Bwamba fault (scarp), infiltrates the
permeable sedimentary cover and forms a shallow reservoir in the subsurface of Buranga. These hot
springs provide a potential source of electrical energy by driving a conventional steam turbine or a binary
power plant, which can then be used for rural electrification.
14.14 Borehole Drilling

A borehole may be dug, bored driven, jetted or drilled and therefore the borehole drilling methods include
rotary, percussion and other simple drilling techniques (Elson and Shaw, 1996; Ragunath, 2007)
14.14.1 Rotary Drilling
i) Direct Rotary (Hydraulic Rotary)

The direct rotary method involves drilling by continuous rotary action transmitted to a bit by a string of
drill pipe. Circulating fluid (usually a clay-based mud) is pumped down the drill pipe and returns via the
annulus between the pipe and hole carrying the rock cuttings to the surface for disposal. The fluid also
cools and lubricates the bit and stabilizes the hole by hydrostatic pressure and mud (filter cake) build-up
on the hole face to prevent caving.
484
The direct rotary method excels in drilling small to moderate diameter deep holes in sedimentary rock.
However, it may yield minimum data (unless the hole is cored as it is drilled) because of rapid drilling
action and the masking effect of the drilling fluid on geologic and ground-water conditions. Geophysical
logging usually must follow direct rotary drilling to obtain maximum subsurface data.
The advantages and disadvantages are explained below:
Advantages 1. Generally rapid penetration rates in most formations to great depth.

2. Unconsolidated formations supported without inserting temporary tubes.
3. Bit changing infrequent
4. Less down-hole problems when drilling alternate hard and soft strata.
Disadvantages 1. Relatively high cost of equipment – approx. 3-4 times that of cable tool.
2. Risk of ‘mudding-off’ weak aquifers (needs well development).
3. Great difficulty in drilling boulders
4. Possibility of losing all mud into fissures e.g. in limestone
5. Considerable quantities of drilling water required.
6. Difficulty in identifying an aquifer, especially when exploration drilling.
7. Requires foreign exchange for operations and maintenance.
ii) Reverse Circulation (Reverse Rotary)

This method is similar to the direct rotary except that the fluid (usually water) is circulated by gravity
down the annulus and returns within the drill pipe. The fluid level is maintained at ground level to prevent
caving. This required an adequate supply of make-up water.
The reverse circulation method is most useful in drilling shallow – to moderate-depth holes in fine – to
medium-grained, unconsolidated materials. It yields moderate to maximum geologic data because of use
of clear water and minimum grinding of cuttings. However, ground-water conditions may be masked.
This method may not be practical for high-pressure artesian conditions.
Advantages 1. Extremely rapid drilling at large diameters, especially in gravels and soft
sediments.
2. Power is used to remove the formation rather than to crush and reduce it.
3. Rapid production of representative strata samples.
4. No risk of mudding-off the aquifer provided the lagoon is maintained silt
free.
5. Temporary tubes unnecessary below short conductor pipe.
6. A very clean partially developed hole is produced.
Disadvantages 1. Equipment of high capital cost.

2. Very large quantities of clean back-up water should be available.
3. Boulders cannot be drilled unless of smaller diameter than the drill pipe
bore.
iii) Air Rotary

The air rotary method is similar to the direct rotary method, except air (and often a foam-like additive) is
used as a circulating fluid.
The air rotary method is most useful in drilling small diameter holes in hard rock or in arid regions. It
yields minimum geologic data because of maximum grinding of cutting and mixing of cuttings from
485
various depths. However, in hard-rock terrain where water is encountered in fractures or similar openings,
it can give moderate to maximum ground-water data.
Advantages 1. Loss of circulation does not prevent drilling.

2. Impossible to mud-off aquifer
3. Drilling mud not required.
4. Rapid return of cuttings
5. Longer bit life
6. Faster penetration
7. No contamination of cuttings
Disadvantages 1. Inability to support caving formations

2. Efficiency falls with increase in depth below water.
3. Usually desirable to have mud back-up if unsure of strata.
14.14.2 Percussion Drilling

This method drills by lifting and dropping a string of heavy cable-suspended drilling tools. A chisel-like
bit at the bottom of the string of tools breaks or crushes the rock at the bottom of the hole into small
pieces. Cuttings are removed by a bucket-like bailer suspended on a cable.
The cable-tool method is most useful in drilling small – to moderate-diameter, shallow – to moderate-
depth holes in boulder material, highly permeable rock, and where subsurface conditions are diverse. It
yields moderate to maximum geologic and ground-water data because of slow discontinuous drilling
action, lack of a circulating fluid and common use of casing.
Advantages 1. Simplicity
2. Low capital cost and running costs
3. Able to make a hole under almost all conditions
4. Immediate facility for water sampling as drilling progresses.
5. Low demand for drilling water
6. Worldwide tool availability
7. Penetration of aquifer apparent immediately
8. No drilling mud so no blocking the aquifer.
Disadvantages 1. Unconsolidated formation has to be supported with multiple columns of

casing.
2. Slow progress in hard rock.
14.15 Simple Drilling Methods

In many countries, water is obtained from hand pumps installed above shallow (less than 60m deep)
boreholes.
Drilling constraints
Whatever drilling method is used, there are several considerations (Elson and Shaw, 1996), which must
be taken into account:
486
 The amount of energy required to drill is governed by the rock type. Unconsolidated formations
such as sand, silt or clay are weak and much easier to drill than consolidated rocks such as
granite, basalt or slate which are hard, strong and dense.
 For hard rocks, cutting tools will need cooling and lubrication.
 Rock cuttings and debris must be removed.
 Unconsolidated formations will require support to prevent the hole from collapse.
The following low-cost, appropriate drilling methods are described and illustrated below:
i) Percussion drilling
The lifting and dropping of a heavy (50kg+) cutting tool will chip and excavate material from a hole. The
tool can be fixed to rigid drill-rods, or to a rope or cable as shown in Fig 14.50. With a mechanical winch,
depths of hundreds of meters can be reached.
Fig. 14.50 Simple Percussion Drilling

Source: (Elson and Shaw, 1996)
Advantages
 Simple to operate and maintain.
 Suitable for a wide variety of rocks.
 Operation is possible above and below the water table.
 It is possible to drill to considerable depths.
Disadvantages
 Slow, compared with other methods.
 Equipment can be heavy.
 Problems can occur with unstable rock formations.
 Water is needed for dry holes to help remove cuttings.
487
ii) Hand-auger drilling

The cutting tool (known as the auger head) is rotated to cut the ground, and then withdrawn to remove
excavated material as shown in Fig 14.51. The procedure is repeated until the required depth is reached.
This method is only suitable for unconsolidated deposits.
Fig 14.51 Hand Auger Drilling Method.

Advantages
 Inexpensive
 Simple to operate and maintain.
Disadvantages
 Slow, compared with other methods.
 Equipment can be heavy.
 Problems can occur with unstable rock formations.
 Water is needed for dry holes.

iii) Jetting
Water is pumped down the centre of the drill-rods, emerging as a jet. It then returns up the borehole or
drill-pipe bringing with it cuttings and debris. The washing and cutting of the formation is helped by
rotation, and by the up-and-down motion of the drill string as shown in Fig 14.52. A foot-powered treadle
pump or a small internal-combustion pump are equally suitable.
Advantages
 The equipment is simple to use.
 Possible above and below the water table.
Disadvantages
 Water is required for pumping.
 Suitable for unconsolidated rocks only (e.g. sand, silt, clay
488
 Boulders can prevent further drilling.
Fig.14.52 The Jetting Method.

iv) Sludging (reverse jetting)

A hollow pipe of bamboo or steel is moved up and down in the borehole while a one-way valve-your
hand can be used to improvise successfully – provides a pumping action. Water flows down the borehole
annulus (ring) and back up the drill pipe, bringing debris with it as shown in Fig 14.53. A small reservoir
is needed at the top of the borehole for recirculation. Simple teeth at the bottom of the drill-pipe,
preferably made of metal, help cutting efficiency.
Advantages
 The equipment can be made from local, low-cost materials, and is simple to use.
Disadvantages
 Suitable for unconsolidated rocks only.
 Boulders can prevent further drilling.
489
Fig.14.53 The Sludging Process

v) Rotary-percussion drilling
In very hard rocks, such as granite, the only way to drill a hole is to pulverize the rock, using a rapid-
action pneumatic hammer, often known as a ‘down-the-hole hammer’ (DTH). Compressed air is needed
to drive this tool as shown in Fig 14.54. The air also flushes the cuttings and dust from the borehole.
Rotation of 10-30 rpm ensures that the borehole is straight, and circular in cross-section
Advantages
 Drills hard rocks.
 Possible to penetrate gravel
 Fast.
 Operation is possible above and below the water-table.
Disadvantages
 Higher tool cost than other tools illustrated here.
 Air compressor required.
 Requires experience to operate and maintain.
490
Fig.14.54 Rotary- Percussion Drilling.

vi) Rotary drilling with flush

A drill-pipe and bit are rotated to cut the rock. Air, water, or drilling mud is pumped down the drill-pipe
to flush out the debris as shown in Fig 14.55. The velocity of the flush in the borehole annulus must be
sufficient to lift the cuttings.
Fig. 14.55 Rotary Drilling with Flush.

491
Advantages
 Most rock formations can be drilled.
 Water and mud supports unstable formations.
 Fast.
 Operation is possible above and below the water-table.
 Possible to drill to depths of over 40 depths.
 Possible to use compressed air flush.
Disadvantages
 Requires capital expenditure in equipment.
 There can be problems with boulders.
 Rig requires careful operation and maintenance.
14.16 Soils as Drainage Filters

Sometimes when groundwater emerges from a soil through an exposed surface or passes from a soil into a
drainage system, it is likely to carry with it some of the fines of the soil. This phenomenon is called
piping or internal erosion and if it is allowed to continue, it may lead to instability of the soil since the
loss of fines produces zones or pipes of increased permeability and flow rate extending in the upstream
directions. The hydraulic gradient is subsequently increased in the approaches to these permeable zones
and therefore the tendency to rode is increase.
Flow form an open outlet of soil or the side slope of an earth dam may cause surface erosion, which can
lead to instability by weakening the slope surface. These situations are common place in engineering for
safe and durable earthworks and the efficient drainage, the erosion should be prevented.
This erosion can be prevented by the placing of selected natural material at the outflow surface in the
form of a filter and promote the proper functioning of drainage works and wells and especially in cases
where seepage water emerges from soil (Jackson and Dhir, 1988). Examples of typical civil engineering
applications are illustrated in Fig.14.56.
The design of a filter is largely empirical, but must be fine enough to prevent soil particles being washed
away and yet coarse enough to allow the passage of water.
Terzaghi and Peck (1948) developed the following criteria for filter design:
D15 filter > 4 x D15 of base material

D15 filter < 4 x D85 of base material (14.117)
The first equation shows that the filter layer has permeability several times higher than that of the soil it is
designed to protect. The requirement of the second equation is to prevent piping within the filter. The
ratio D15 (filter)/ D85 (base) is known as the piping ratio. The above Terzaghi’s rule is an established
method, which has been extremely useful for several decades. With recent progress in geotechnical
engineering it is now considered as more a conservative approach (Smith, 1990).
Pumping and observation wells used in pumping tests are perforated to allow the ingress of water. When
it is likely that fine material will be washed into the well, it should be surrounded by a layer of filter
material. A common rule for the choice of a suitable material to form such a filter is:
> 1.0 (14.118)
492
The required thickness of a filter layer depends upon the flow conditions and can be estimated with the
use of Darcy’s Law. In addition to the above criteria, the filter material should be well graded with a
grading curve more or less parallel to the base material. Fig 14.57 shows the range of suitable filter soils.
Protected filters are usually constructed in layers each, which is coarser than the one below it and as such,
they can be referred to as reversed filters.
More recently, the design of graded filters has been made more effective by the inclusion of fabricated
materials like synthetic fabric, stainless steel mesh, pins and cables as well as artificial geotextiles. These
geotextiles, have greatly reduced the need for multiple layers of granular materials and are usually cost
effective (Watson and Burnett, 1993).
Fig.14.56 The use of soils as drainage filters.

(Source: Jackson and Dhir, 1988)
493
Fig 14.57 The selection of suitable filter soils.

(Source: Jackson and Dhir, 1988)
Summary
Groundwater is the subsurface water that occurs beneath the water table in soils and geological
formations. This chapter discusses the significance of groundwater, its occurrence in aquifers, Darcy’s
Law and its movement, factors affecting the hydraulic conductivity, the simulation of groundwater under
different conditions of steady flow, unsteady flow, pumping tests and their analysis for estimating
groundwater quantities, groundwater quality and electrical resistivity surveying for assessing groundwater
availability. It also discusses groundwater recharge, isotopes in groundwater detection and assessment,
geothermal resources, drilling methods and soils as drainage filters.
References
1. Arora, S.O., Water Resources Hydropower and Irrigation Engineering, Standard Publishers and
Distributors, 2007, Delhi, India.
2. Bear,J.,Dynamics of Fluids in Porous Media, Elservier, 1972, New York, USA
3. Bierschenk, W.H., Determining Well Efficiency by Multiple Step Drawdown Tests, International
Association of Science Hydrology 1964, 64, pp 493-507.
4. Boast, C.W.and Kirkman, D., Augur Hole Seepage Theory, Soil Science Society America,
Proceedings, 1971, Vol 35, pp 365-373.
5. Bouwer, H., Groundwater Hydrology, McGraw Hill, 1978, New York, USA
6. Bradley, G., Tindimugaya, C., Taylor, R., Relict River Channels in Uganda: Potential Corridors
of Preferential Groundwater Flow and Storage, Proceedings Groundwater and Climate in
Africa, 2008, Kampala, Uganda, Ministry of Water and Environment, University College,
London.
7. Brereton, N.R., Step Drawdown Pumping Tests for the Determination of Aquifer and Borehole
characteristics, Water Research Council 1979, LondonUK.
8. Chow, V.T., On the Determination of Transmissibility and Storage Coefficients from Pumping
Test Data. Trans American Geophysical Union, 1952, Vol 33,pp397-404.
494

Groundwater

Uploaded by

Document Informationclick to expand document information

Copyright:

Available Formats

Groundwater

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Groundwater

Uploaded by

Copyright:

Available Formats

An Introduction to Hydrology and Water Resources Engineering in Uganda

a) Virtually free of pathogenic organisms; Often easily contaminated

b) Effectively self purifying; soil or Treatment usually necessary

e) Relatively constant temperature Variable temperature; can vary

f) Colour, turbidity good Colour, turbidity may change readily,

N.B.: These terms are relative

Aquifers take two basic forms

iv) Confined Aquifer: An aquifer confined between two aquitards

Fig 14.1a A perched water table.

x) Piezometric surface or potentiometric surface:

14.1.1 The Groundwater Column

Fig 14.2 The Groundwater column.

ii) Capillary Zone

iii) Groundwater (Saturated) Zone

Sy= Wy/V (14.2)

α =Wr + Wy where all pores are interconnecting (14.3)

14.2 The Occurrence of Groundwater

14.2.1 Sources of Groundwater

b) Rock types as aquifers

Extrusive: All lavas

ii) Sedimentary rocks

iii) Metamorphic rocks

iv) Superficial deposits

14.2.2 Groundwater Occurrence in Uganda

14.3 Water Movement in Soils (Saturated Flow)

14.3.1 Steady Flow of a Perfect Fluid

Fig.14.4 A control volume with steady flow.

Equating work input in δt by P1 to work expended in δt against P2

for a viscous fluid

14.3.2 Viscous Flow between Parallel Plates

14.3.3 Viscous Flow through a Tube

Normalise by introducing mean hydraulic radius RH

for parallel plates for a tube

So for any shape or cross section, we can write:

If the porosity is n, for a given cross sectional area A of material.

Substitute for RH in equation 14.12,

So define Discharge Velocity = Q = V

Substitute for Vs/As in equation (14.12):

Table 14.1 Typical range of values of K

Soil Type K (m/sec)

14.4.1 Intrinsic Permeability

14.4.2 `Hydraulic Conductivity

14.4.4 Methods of Measuring Permeability

Fig 14.5 Constant Head Permeameter.

From Darcy’s Law, V = Ki But Q = VA

Consider soil at depth x

Pore Pressure  P2 + P1 - P2 .x (14.24)

When Equation. 14.23 equals Equation 14.24 P1 = P2 = Tg (14.25)

i.e. from Equation.14.25 and Equation.14.26

Most sands have T close to 2w -icrit  1

Ground surface Water table

impermeable or highly impermeable layer

Fig 14.7 Auger hole cross section.

iv) Pumping Tests of Wells

14.5 Heterogeneity and Anisotropy of Hydraulic Conductivity

Homogeneity and Heterogeneity

14.5.1 Types of Heterogeneity Formations

a) Layered geological strata