Faculty of Engineering

School of Civil and Environmental Engineering

Course Notes
Water Resources Engineering CVEN3501
Groundwater Resource

Lecturer: A/professor Martin Andersen

Notes based on material from Prof Ian Acworth and Dr Gabriel Rau
Contents

1 Groundwater resources and flow equations 1

1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.3 Physical Properties of Water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.4 Porosity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.5 Water Potential and Hydraulic Head . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.6 Groundwater Movement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.7 Groundwater Transport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.8 Groundwater Flow Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
1.9 Steady-State Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
1.10 Finite Difference Solution Methods to the Steady-State Flow Problem . . . . . . . . . . . 21
2 Aquifer properties and borehole construction 28
2.1 Storage Properties of Porous Rocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.2 Effective Stress and the impact of dewatering . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.3 Well Drilling Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2.4 Cable Tool Drilling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2.5 Rotary Drilling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.6 Drilling Fluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2.7 Well Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
2.8 Gravel Packs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
2.9 Well Completion and Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
2.10 Field preparations for a pumping test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
2.11 Flow measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
2.12 Pumping Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
3 Pumping test analysis and occurrence of groundwater 56
3.1 Mathematical model of flow to a well . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
3.2 Analytical solution to the radial flow equation . . . . . . . . . . . . . . . . . . . . . . . . . 57
3.3 Pumping test interpretation using graphical methods . . . . . . . . . . . . . . . . . . . . . 58
3.4 Groundwater in Australia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
3.5 Groundwater in sedimentary basins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
3.6 Groundwater in fractured rocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
3.7 Groundwater in Tertiary palaeochannels . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
3.8 Groundwater in unconsolidated surficial deposits . . . . . . . . . . . . . . . . . . . . . . . 78
3.9 Groundwater on oceanic islands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
3.10 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
3.11 Connectivity of Surface Water and Groundwater . . . . . . . . . . . . . . . . . . . . . . . 84
3.12 Interested in further groundwater studies? . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Bibliography 95

i
Lecture 1

Groundwater resources and flow

equations

1.1 Introduction

In this course you will be introduced to the fundamentals of groundwater flow theory and aspects of the
field application of this theory. There will be three workshops. Attendance is compulsory. An assignment
will be set in the first week of the segment and will be due in two weeks. The assignment will be carried
out individually and will require you to develop a simple groundwater flow model using a finite difference
solution to the equation describing steady-state groundwater flow. The assignment can be completed
using a spreadsheet.

Aspects of this course are also covered in the third year geotech subject. Often, there are small differ-
ences in terminology used - so be careful.

1.1.1 Why Study Groundwater

The study of groundwater is important in the following areas:

• Water Supply
• Dewatering
• Contamination remediation

Hydrogeology is the name normally given to the study of groundwater, although it is sometimes referred
to as Geohydrology.

Domenico and Schwartz (1997) provide the following definition of hydrogeology:

Hydrogeology is the study of the laws governing the movement of subterranean water, the
mechanical, chemical, and thermal interaction of this water with the porous solid, and the
transport of energy and chemical constituents by this flow.

Hydrogeology can be studied at the postgraduate level at UNSW where many of the aspects included in
the definition by Domenico and Schwartz (1997) are investigated in detail.

Check out www.connectedwaters.unsw.edu.au for course details and groundwater news and research.

1
Lecture 1 Groundwater resources and flow equations

Figure 1.1: Multiple users at a village well

1.1.2 The Hydrogeological Cycle

Some 97% of all the fresh water (unfrozen) that is found on the planet is stored underground. This
amounts to approximately 8 million km3 of groundwater. At least 1,500 million people depend upon this
reserve for drinking water. Figure 1.1 graphically shows the dependence on groundwater for supply in
some areas of the world! Groundwater has been abstracted from wells for at least the past 5000 years
with the oldest well in the world claimed to be a 5600 year old structure made from wood in Zhejiang
Province, China (Jiao, 2007).

Groundwater reserves are recharged for the most part by rain that infiltrates through the soil to pass
below the root zone of vegetation. These reserves are occasionally augmented by streams and rivers
which lose water through their beds to underground strata. Once underground, the water flows at rates
ranging from more than 10 m per day to as little as 1 m per year, until it reaches an outlet. This may take
the form of a spring or a system of small seepages that keep rivers flowing during droughts. In some
places, the groundwater discharge is sufficient to support a river. Where groundwater discharges but
cannot be drained away, a salt lake will often form, particularly if the discharge occurs in a hot dry area.
The salt lakes which are a common feature in western New South Wales and Northern Victoria occur
where groundwater in the Murray-Darling Basin aquifers discharge.

The time scales of groundwater flow are extremely long. Water in the Great Artesian Basin is older
than 1 million years, but ages of 30,000 years are more common. The often great age of groundwater
means that it can often only be considered as a renewable resource at time scales of thousands of
years. Groundwater which becomes contaminated by any of a large number of processes, will naturally
be regenerated, if we can wait around long enough!

Occasionally, geological events trap groundwater underground, cutting it off from both a source of sup-
ply and from its outlets. Climatic change may also deprive aquifers of any means of recharge, as
has happened under a number of regions which are now desert but which were formerly much wetter.
Groundwater reserves which are no longer replenished are called fossil water.

Water Resources Engineering CVEN3501 2

Lecture 1 Groundwater resources and flow equations

Figure 1.2: The hydrogeological cycle

The schematic from the UNEP shown in Figure 1.2 illustrates some of the processes occurring in the
hydrogeological cycle.

A summary of the impact of groundwater abstraction is given by UNEP for the USA. The problems iden-
tified by UNEP are a useful indication of the sorts of problems associated with groundwater development
world wide.

• Groundwater resources under the United States are estimated at 125, 000km3 , roughly the equiv-
alent of the flow of the Mississippi River for 200 years.
• Groundwater supplies more than half of US drinking water, and 96% of the drinking water con-
sumed in rural areas.
• Some 30% of the groundwater used in the United States for irrigation comes from one aquifer - the
Ogollala aquifer, running from Dakota to Texas.
• Dams along canals in Southern Florida are needed to prevent saltwater contaminating the Bis-
cayne aquifer which supplies drinking water to 3.5 million Floridians.
• Groundwater abstractions have lowered the ground by up to 3 metres in the Houston-Galveston
area, giving rise to coastal and inland flooding. Land in California’s San Joaquin Valley has sunk
by as much as 9 metres since the 1920’s.
• Over-abstraction caused water levels in the Northern Midwest aquifer system to fall by more than
300 m. Many Chicago suburbs switched to water from Lake Michigan and the water table has now
risen by 75 m since 1985.
• Radioactive waste from a former nuclear production facility near Richmond, Washington, has con-
taminated parts of the aquifer below.
• Some three million people are supplied from a three-layered aquifer below Long Island - but the
top layer is now contaminated by oil and petrol, septic tank effluents and pesticides.

1.1.3 References

These notes are intended to be sufficient for your CVEN 3501 studies. A detailed bibliography is included
in the notes. A good general reference for those who may be more interested is supplied by Fetter (2001)

Water Resources Engineering CVEN3501 3

Lecture 1 Groundwater resources and flow equations

in his book Applied Hydrogeology. The text by Domenico and Schwartz (1997) provides a more robust
coverage that is a good starting point for further work in this area.

1.1.4 Units

Reference to water flow through porous media is found in texts derived from a wide variety of disciplines
from geology through the agricultural sciences to civil engineering. With so many different professions
involved it is not surprising that a varied terminology has evolved and, in particular, the use of many
different units.

The International System (SI) of units will be used in this subject. Note that much groundwater litera-
ture comes from USA where US imperial units are frequently used rather than SI units. An excellent
web site covering the basis of the SI system and giving definitions and dimensions of all units is at
http://physics.nist.gov/cuu/Units/index.html.

Examples of SI Units useful in groundwater studies are given in Table ??.

Table 1.1: SI Quantities

Quantity Unit Name Unit Symbol Expression in SI
derived units
Area square meter m2
Volume cubic meter m3
Velocity meter per second m s−1
Acceleration meter per second per second m s−2
Frequency hertz Hz
Density, (mass density) kilogram per cubic meter kg m−3
Momentum Kilogram meter per second kg m s−1
Moment of inertia kilogram meter squared kg m−2
Force newton N kg m s−2
Moment of force (torque) Nmeter Nm
Pressure, stress pascal Pa (N m−2 ) kg m−1 s−2
Viscosity: kinematic meter squared per second m2 s−1
Viscosity: dynamic pascal second Pa s kg s−1 m−1
Surface tension newton per meter N/m kg s−2
Energy, work, quantity of heat joule J (N m) kg m2 s−2
Power, radiant flux Watt W (J/s) kg m2 s−3
◦
Temperature degree Celcius C
Heat flux density watt per square metre W/m2
Thermal conductivity watt per metre kelvin W/m K
Heat capacity joule per degree kelvin J/K
Quantity of electricity C As
Electric potential volt V W/A
Electric field strength volt per meter V/m
Electric resistance ohm Ω
Electric conductance siemens S A/V
Electric capacitance farad F A s/V
Magnetic flux weber Wb Vs
Inductance henry H
Magnetic flux density tesla T W/m2
Magnetic field strength ampere per meter A m−1

1.2 Definitions

The following definitions may be useful.

Water Resources Engineering CVEN3501 4

Lecture 1 Groundwater resources and flow equations

Figure 1.3: Aquifer Types

Aquifer - a saturated permeable geologic unit that can transmit significant quantities of water under
ordinary potential gradients.
Aquiclude - a saturated geologic unit that is incapable of transmitting significant quantities of water
under ordinary potential gradients.
Aquitard - a saturated geologic unit that is capable of transmitting water under ordinary hydraulic gra-
dients but not in sufficient quantities to allow completion of production wells within them.
Unconfined aquifer - An aquifer which is open to atmospheric pressure
Confined Aquifer - An aquifer which occurs between two aquicludes.
Water Table - the water surface in an unconfined aquifer
Piezometric Surface - the height to which water rises in a stand-pipe (piezometer) above the top of a
confined aquifer.

These concepts are illustrated in Figure 1.3.

1.3 Physical Properties of Water

A brief review for revision purposes of the physical properties of water.

1.3.1 Viscosity

The resistance to the movement of one layer of fluid over another adjacent layer is ascribed to the
viscosity of the liquid. The dynamic viscosity of a fluid is the property that allows fluids to resist relative
motion and shear deformation during flow. The more viscous a fluid the greater the shear stress at any
given velocity gradient. According to Newton’s law of viscosity,
dv
τ =µ (1.1)
dy
where:

τ is the shear stress,

Water Resources Engineering CVEN3501 5

Lecture 1 Groundwater resources and flow equations

Table 1.2: Properties of Water at 1 Atmosphere

Temperature Density Dynamic Viscosity Compressibility

◦
C [kg m−3 ] ρ [kg m−1 s−1 ] µ [m2 N −1 ] β

0 999.8395 1.787 × 10−3 5.01 × 10−10

4 999.9720 1.567 × 10−3 -
10 999.6996 1.307 × 10−3 4.78 × 10−10
15 999.0996 1.139 × 10−3 -
20 998.2041 1.002 × 10−3 4.58 × 10−10
25 997.0449 0.8904 × 10−3 4.57 × 10−10
30 995.6573 0.7975 × 10−3 4.46 × 10−10
40 992.2158 0.6529 × 10−3 4.41 × 10−10
50 988.0363 0.5468 × 10−3 4.40 × 10−10
60 983.1989 0.4665 × 10−3 4.43 × 10−10
70 977.7696 0.4042 × 10−3 4.49 × 10−10
80 971.7978 0.3547 × 10−3 4.57 × 10−10
90 965.3201 0.3147 × 10−3 4.68 × 10−10
100 958.3637 0.2818 × 10−3 4.80 × 10−10

dv/dy is the velocity gradient, and

µ is the dynamic viscosity.

The units of dynamic viscosity µ are centipoise N sm−1 × 10−3 .

In many problems involving viscosity, the magnitude of the viscous forces compared to the magnitude
of the inertia forces is of interest. Since the viscous forces are proportional to the viscosity µ, and the
inertia forces are proportional to the density ρ, the ratio µ/ρ is often involved. This ratio is called the
kinematic viscosity, with units of ms /s.

Viscosity is also temperature dependent as it is a function of the number of hydrogen bonds existing.
Values of viscosity for a range of temperatures are given in Table 1.2. The value of fluid viscosity
influences the rate of fluid movement through a porous medium.

1.3.2 Fluid Pressure

The fluid pressure p at any point in a standing body of water is the force per unit area which acts at
that point. Under hydrostatic conditions, the fluid pressure at a point reflects the weight of the column of
water overlying a unit cross-sectional area around the point.

p = ρgψ (1.2)

where:

ρ is density,
g is the acceleration due to gravity (9.80m/s2 )
is the height of the water column above the reference plane.

The units of pressure are pascals Pa or kgm−1 s−2 .

It is possible to express pressure relative to absolute zero pressure, but more commonly it is expressed
relative to atmospheric pressure. In the latter case, it is called gauge pressure , as this is the
pressure registered by a pressure gauge with atmospheric pressure as the reference.

Water Resources Engineering CVEN3501 6

Lecture 1 Groundwater resources and flow equations

1.3.3 Compressibility

The compressibility β of a fluid reflects its stress-strain properties. Stress is the internal response of a
material to an external pressure. For fluids, stress is imparted through the fluid pressure. Strain is a
measure of the linear or volumetric distortion of a stressed material. When a fluid is stressed, the strain
occurs as a reduced volume (and increased density) under increasing fluid pressures.
d
Compressibility is defined as strain/stress dσ . It is the inverse of the modulus of elasticity. Units of
compressibility are m2 /N .

For a given mass of water

dρ/ρ
β=− (1.3)
dp

The compressibility of water has a very low value but is still dependent upon temperature. Values for a
range of temperatures are given in Table 1.2.

The compressibility of solids is usually designated by α where the units are the same as for fluids.
Values of α are several orders of magnitude greater than those for β. This inequality is of importance
when aquifer confined storage is calculated.

1.3.4 Summary of Water Properties

A summary of the physical properties of water are given in Table 1.3.

Table 1.3: Summary of Water Properties

Property Value Units

Liquid Density 998.0 kg/m3

Solid Density 910.0 kg/m3
Vapour Density 1.73 × 10−2 kg/m3
Heat of Fusion 334 kJ/kg
Heat of Vapourisation 2.45 × 103 kJ/kg
J
Specific Heat 1.0 kg K
Relative Permitivity 80 -
W
Thermal Conductivity 1.0 mK
2
Kinematic Viscosity 1.0 m /s
Surface Tension 72.7 × 10−3 J/m2

1.4 Porosity

Water is stored beneath the ground in the pore space of unconsolidated material and rocks. Two types
of porosity are recognised:

Primary porosity due to the properties of the soil or rock matrix.

Secondary porosity due to secondary solution or fracturing of the rock mass.

Various texture and porosity relationships were described by Meinzer (1923) and shown in Figure 1.4.

In Figure 1.4 six examples of porosity development are given, viz:

Water Resources Engineering CVEN3501 7

Lecture 1 Groundwater resources and flow equations

Figure 1.4: Relation between texture and porosity

a Well-sorted sedimentary deposit having high porosity.

b Poorly-sorted sedimentary deposit having low porosity. The smaller grains block the pore space
between the bigger grains.
c Well-sorted sedimentary deposit whose porosity has been diminished by the deposition of mineral
matter (shown black) in the interstices. The Hawesbury Sandstone is a good example.

d Highly porous deposit but with very low hydraulic conductivity because the pores are not connected.
e Limestone rock rendered porous by solution.
f Granite rendered more porous by fracturing.

A range of possible porosity values is shown in the Table 1.4 below.

Table 1.4: Range of Porosity Values

Material φ(%)

Unconsolidated deposits
Gravel 25 - 40
Sand 25 - 50
Silt 35 - 50
Clay 40 - 70
Rocks
Fractured basalt 5 - 50
Karst limestone 5 - 50
Sandstone 5 - 30
Limestone, dolomite 0 - 20
Shale 0 - 10
Fractured crystalline rock 0 - 10
Dense crystalline rock 0-5

Water Resources Engineering CVEN3501 8

Lecture 1 Groundwater resources and flow equations

1.5 Water Potential and Hydraulic Head

1.5.1 Introduction

Water potential plays a similar role in water flow theory to the role of temperature in heat flow problems
or voltage in electrical circuit theory. Water flows in response to gradients in water potential. When water
potential is uniform across a boundary, no water will flow, even though water content may be different
on the two sides of the boundary. Hydraulic head is a particular representation of the potential energy in
the system and can also be considered as a water potential.

The water potential is the potential energy per unit mass of water in the system, compared to that of
pure, free water at atmospheric pressure. The units of potential are Joules/kg.

Note that you have used units of energy per unit weight in Geotechnics. The diversity of approaches for
describing common properties is annoying but something you need to be aware of!

The total water potential is the sum of several component potentials:

 s
m
Φ = Φm + Φv + Φp + Φg (1.4)
s2

where:

Φm is the matric (m) potential,

Φv is the velocity (v) potential,
Φp is the pressure (p) potential,
Φg is the gravity (g) potential

In general, only two or three of these potentials require consideration in any particular problem. For
studies of saturated groundwater flow, the matric potential can be ignored. For studies in the unsaturated
zone, the matric (instead of the pressure) potential is important. The velocity potential is included to allow
comparison with Bernouli’s Equation, but as groundwater flow is very slow, it is normally insignificant.

1.5.2 Pressure Potential

The fluid pressure potential is calculated from:

p
Φp = = gψ (1.5)
ρw

where ψ is the depth of the water column above the reference point.

1.5.3 Gravitational Potential

The gravitational component of the water potential is fundamentally different to any of the other compo-
nents since it is the result of ”body forces” applied to the water as a consequence of its position in the
earth’s gravitational field. The gravitational potential is calculated from:
Φg = gz (1.6)

where z is the distance from the point of measurement to a reference level where Φg is taken as zero.
The reference level is usually taken at mean sea-level by hydrogeologists and the soil surface or the
water table by soil physicists.

Water Resources Engineering CVEN3501 9

Lecture 1 Groundwater resources and flow equations

1.5.4 Velocity Potential

The velocity potential arises from the energy per mass which the water has as a result of its motion
(kinetic energy). The velocity potential is calculated as
v2
Φv = (1.7)
2

where v is the velocity of the groundwater. The velocity potential is normally several orders of magnitude
less than the pressure or gravity potentials, as the velocity of groundwater is generally very slow.

1.5.5 Hydraulic Head

In many problems involving prediction of flow in fully saturated aquifers, the water potential reduces to
the sum of the pressure and gravitational potentials,
Φ = Φp + Φg (1.8)

From Equations 1.5 and 1.6, the total potential can therefore be written as
p
Φ= + gz (1.9)
ρw

If the density is constant, then ρpw may be substituted by gψ, where ψ is the height of the water column
above the reference point and z is the distance from the reference point to the datum plane.

The gravitational constant g does not vary significantly with surface location and therefore, as long as
density is constant, Equation 1.9 can be divided by g to give:
h=ψ+z (1.10)
where:

Φ
h is the hydraulic head g,

ψ is the pressure head,

z is the elevation head.

The dimension of these head quantities is length. This fundamental head relationship is basic to an
understanding of saturated groundwater flow.

1.5.6 Systems of Units for the Total Soil Water Potential

There are several alternative systems of units in which the total potential and its components can be
described, depending upon whether the quantity of pure water is expressed as a mass, a volume or a
weight. These alternative systems can readily confuse, and a summary table is provided (Table 1.5).

1.5.7 Hydraulic Head at the Water Table

A useful measure of hydraulic head in a flow system is the hydraulic head which exists at the water table
or at the top of a water column in a piezometer (stand pipe). At these locations, ψ = 0 and h = z.

The hydraulic head at the water table or in a stand pipe can be established by simple measurements:

Water Resources Engineering CVEN3501 10

Lecture 1 Groundwater resources and flow equations

Table 1.5: Systems of Units for the Water Potential

Description Symbol Name Dimensions SI Unit

L2
Energy per mass Φ Water Potential T2
J/Kg
Energy per volume p Water pressure M
LT 2
N/m2
Energy per weight h Hydraulic head L m

Figure 1.5: Hydraulic Head, Elevation Head and Pressure Head

1. Measure the depth to the water surface from the ground surface. This quantity is often referred to
as the standing water level (SWL). It can be measured using a dip-meter or contact gauge of some
form.

2. Measure the elevation of the ground surface above the regional datum.
3. Calculate the value of h by subtracting the SWL from the surface elevation to give the elevation of
the water surface.

Question 2 in the first workshop is based on this concept.

The relationships between hydraulic head, elevation head and pressure head are shown for a piezometer
and a flowing well in Figure 1.5.

1.6 Groundwater Movement

1.6.1 Hydraulic head measurement using a piezometer

The basic device for the measurement of hydraulic head is a tube or pipe in which the elevation of
the water level can be determined. In the laboratory, the tube is a manometer, in the field the pipe is

Water Resources Engineering CVEN3501 11

Lecture 1 Groundwater resources and flow equations

called a piezometer. It must be open at the base, so that water can flow into the pipe, and open to the
atmosphere at the top.

The point of measurement is at the base of the piezometer and not at the level to which water rises
inside the pipe. The open area section of the piezometer base should therefore be kept restricted to be
representative of the head in the zone of interest. The simple standpipe piezometer in which the depth to
water is measured by a tape has been replaced to some extent by more complex designs incorporating
pressure transducers, pneumatic devices, and electronic components, allowing the remote collection
and recording of data. The availability of data at frequent time intervals and in a digital manner allows
much improved analysis of hydrogeological parameters.

1.6.2 Equipotentials

In two dimensions, lines can be drawn connecting points of equal hydraulic head. These are similar to
contours on the ground surface which define the surface topography.

In three dimensions, a surface of equipotential is created which is not possible to represent in two-
dimensions, other than to take a cross-section and to assume that there is not significant variation in the
third dimension.

Vertical cross-sections are commonly used in geotechnical work to calculate flow through embankments.
Horizontal cross-sections are often used in groundwater resources. These are also known as potentio-
metric or piezometric maps.

1.6.3 Piezometric Maps

Field measurements of the hydraulic head at a number of different locations may be used to construct a
piezometric map. If the flux of groundwater can be approximated as horizontal, and vertical components
neglected, then measurements of hydraulic head in piezometers or boreholes can be used to construct
equipotentials.

1.6.4 Gradient of the Hydraulic Head

Water moves in response to a hydraulic gradient. If there is no hydraulic gradient the water will remain
stationary. The hydraulic gradient can be simply calculated as the change in hydraulic head between
two points divided by the distance between the two points.

The hydraulic gradient is often calculated in the x-y plane and referred to as the horizontal hydraulic gra-
dient. In many important situations, the hydraulic gradient varies vertically. That is, if you could measure
the hydraulic head at different depths below the same location, the hydraulic head would change.

In a recharge condition, where water is entering the groundwater system, the vertical hydraulic gradient
is negative (z taken as positive upwards). The head will decrease as depth increases.

Under discharge conditions, where water is leaving the groundwater system, the vertical hydraulic gra-
dient is positive. The head will increase as depth increases.

Piezometers are usually installed in groups designed to investigate either the horizontal or vertical move-
ment of groundwater. The hydraulic gradient, either in the horizontal or vertical direction, can be estab-
lished from measurements on a group of piezometers as shown in Figure 1.6.

Mathematically, the gradient of the hydraulic head can be written as

dh
∇h = (1.11)
ds

Water Resources Engineering CVEN3501 12

Lecture 1 Groundwater resources and flow equations

Figure 1.6: Measurement of Horizontal and Vertical Hydraulic Gradients

where:

s is a distance parallel to the direction of ∇h.

If the value of ∇h is zero, then there will be no movement of water irrespective of the value of hydraulic
conductivity. The direction of groundwater flow is a function of the hydraulic conductivity and ∇h. For an
isotropic medium, ∇h is perpendicular to the equipotential and groundwater flows in the direction of ∇h.

1.6.5 Flow Nets

A flow net is a two-dimensional device constructed to describe groundwater flow in either an x-y plane
(horizontal) or an x-z plane (vertical).

Two sets of orthogonal curves can be constructed which describe the flow system. A set of equipotentials
defines the hydraulic head distribution and a set of flow lines, at right angles to the equipotentials (at
least in an isotropic domain). Flow nets are useful tools to help characterise a flow system. Fetter
(2001) gives examples of vertical flow nets which can be used to demonstrate the occurrence of vertical
groundwater flow in local flow systems containing recharge and discharge.

A vertical flow net is shown in Figure 1.7 based upon Hubbert (1940).

Water Resources Engineering CVEN3501 13

Lecture 1 Groundwater resources and flow equations

Figure 1.7: Vertical flow net showing piezometric levels

1.6.6 Measurement of Head at Flowing Wells

Where a head measurement is to be made at a flowing well, the well must be capped and an accurate
pressure gauge installed. The hydraulic head may then be derived from:

p
h=z+ (1.12)
ρg

where:

z is the distance to the reference plane

ρ is the density of the water

g is the acceleration due to gravity (9.80 m s−2)

p is the gauge pressure in Pascals

1.6.7 Factors Affecting the Hydraulic Head

There are many different factors which can cause the hydraulic head to change. Changes in the volume
of water stored in an aquifer produced by an imbalance between recharge and abstraction will cause
long term fluctuations in groundwater l evels. Examples from the Murray Area demonstrate this feature.
Fluctuations over a shorter time period are caused by the aquifer response to rainfall recharge events.
In these examples the change in level is seen in response to aquifer recharge which is related to rainfall
but is also a function of soil type and degree of saturation. There are other factors related to changes in
the stress on the aquifer caused by variations in atmospheric pressure or changes in water pressure.

A list of possible causes which can produce a change in water level is given below:

• Groundwater Recharge
• Evapotranspiration
• Tidal Effects near Oceans
• Earth Tide Effects

Water Resources Engineering CVEN3501 14

Lecture 1 Groundwater resources and flow equations

• Atmospheric Pressure Effects

• Earthquakes
• Groundwater Pumping

• Deep Well Injection

• Artificial Recharge
• Agricultural Irrigation

• Geotechnical Drainage

Water Resources Engineering CVEN3501 15

Lecture 1 Groundwater resources and flow equations

Figure 1.8: Darcy’s Experiment

1.7 Groundwater Transport

1.7.1 Early Experiments

The French hydraulic engineer Henry Darcy, published a report on the water supply in the city of Dijon in
1856. In this report, Darcy (1856) described a laboratory experiment that he had carried out to analyse
the flow of water through sands. The results of this experiment can be generalised into the empirical law
upon which much of the later development of the science of groundwater is based.

Darcy’s experimental apparatus may be generalised as shown in Figure 1.8.

A circular cylinder of cross section A is filled with sand, stoppered at each end, and fitted with inflow and
outflow tubes and a pair of manometers. Water is allowed to flow through the cylinder until such time as
all the pore space is filled with water and the inflow rate Q is equal to the outflow rate.

An arbitrary datum is selected at elevation z = 0 , the elevations of the manometer intakes are z1 and z2
and the elevations of the fluid levels are h1 and h2 . The distance between the manometer intakes is ∆L.
The specific discharge (υ, or also written as q in some texts) (or volume flux per unit area, also referred
to as the Darcy velocity) through the cylinder is defined as:
Q
υ= (1.13)
A
where:

υ is the specific discharge or volume flux

If the dimensions of Q are [L3 T −1 ] and those of A are [L2 ], then υ has the dimensions of a velocity
[L/T].

Darcy’s experiments show that when ∆L is held constant:

υ ∝ ∆h (1.14)
and, when ∆h is held constant
1
υ∝ (1.15)
∆L

Water Resources Engineering CVEN3501 16

Lecture 1 Groundwater resources and flow equations

where:

∆h is the head difference h1 − h2 between the manometer off takes, and

∆L is the difference between the manometer off takes.

The proportionality constant in these relationships was defined as the hydraulic conductivity.

Darcy’s Law can be stated as:

∆h
υ = −K (1.16)
∆L

or in differential form
dh
υ = −K (1.17)
dL

where:

h is the hydraulic head

dh/dL is the hydraulic gradient (water potential gradient)

An alternative form of Darcy’s law can be obtained by substituting for specific discharge as,
dh
Q = −K A (1.18)
dL
This form of the equation is particularly useful in groundwater resource analysis where the fluid moving
through the porous medium is clean water. Further development of the analysis is required before the
equation can be used to study mixtures of fluids and gas or imiscible fluids.

1.7.2 Intrinsic Permeability

It is clear that the proportionality constant in Equation 1.17 or Equation 1.18 must be a function of the
porous medium, in that the porosity, the degree of interconnectivity of pores, the grain size and sorting
of the material will all have a bearing on its value. It is also evident that the viscosity and density of the
fluid must influence this value. For instance, it is intuitively obvious that the rate of movement of a thick
oil through a permeable medium will be slower than that of water, or gas, through the same medium.

Experiments have been carried out with ideal porous media consisting of uniform glass beads of a known
diameter (d) and fluids of different density and viscosity to better resolve the nature of the proportionality
constant.

The following relationships have been observed:

υ ∝ d2 (1.19)

υ ∝ ρg (1.20)

υ ∝ µ−1 (1.21)
dh
Together with the original observation that υ ∝ − dL these relationships lead to a new definition of
Darcy’s Law:

Cd2 ρg dh
υ=− (1.22)
µ dL

Water Resources Engineering CVEN3501 17

Lecture 1 Groundwater resources and flow equations

The parameter C is a dimensionless constant of proportionality. For real soils it must include all the influ-
ence of other media properties which affect the flux, such as grain size and distribution, grain sphericity
and roundness and the nature of the packing.

Comparison with the original Darcy equation shows that:

Cd2 ρg
K= (1.23)
µ

In this equation ρ and µ are properties of the fluid and Cd2 is a function of the medium alone. If we
define lower case k as
k = Cd2 (1.24)

then

kρg
K= (1.25)
µ

The parameter k is known as the intrinsic permeability. The term intrinsic permeability rather than
permeability is used, as in some earlier texts, hydraulic conductivity was also referred to as the coefficient
of permeability or simply, permeability. The term permeability is still in general use but restricted to
the physical properties of the medium. The intrinsic (or specific) permeability is widely used in the oil
industry where the existence of gas, oil and water in multiphase flow systems makes the use of a fluid-
free conductance term of use. Petroleum engineers have defined the darcy as a unit of permeability.

1.7.3 Specific Discharge and Average Linear Velocity

The units of specific discharge υ in Equation 1.17, were noted to be velocity. It is important to realise
that the specific discharge does not represent the linear velocity at which water moves through the pore
space. It is an apparent velocity, representing the velocity at which water would move through the aquifer
if it were an open conduit. The cross-sectional area of the open pore space in the aquifer is much smaller
than an open conduit of the same dimensions, therefore the real fluid velocity through the pore space
will be significantly higher. The average linear velocity may be calculated from Equation 1.26.
Q K dh
ul = = (1.26)
φe A φe dL
where:

ul is the average linear velocity in the direction of L

φe is the effective porosity

The effective porosity is less than the total porosity as some water remains trapped in dead-end pore
space and adheres to the grain surfaces. The total porosity for Botany sand is approximately 0.40, the
effective porosity is approximately 0.36.

1.8 Groundwater Flow Equations

1.8.1 Introduction

Darcy’s Law provides a model to describe the flow of water in a saturated medium. If the hydraulic head
can be measured in the field, then we can predict the direction and rate of movement of groundwater if

Water Resources Engineering CVEN3501 18

Lecture 1 Groundwater resources and flow equations

Figure 1.9: Elemental Control Volume

we have a value of hydraulic conductivity. What is now required is a more rigorous model of groundwater
flow.

1.8.2 Conservation of Fluid Mass

In words, the conservation of fluid mass can be stated as

mass inflow rate - mass outflow rate = change in mass storage with time

in units of mass per time.

This statement can be applied to a domain of any size, but consider the unit volume of porous media as
shown in Figure 1.9 where ∆x = ∆y = ∆z = 1. Such an element is usually called an elemental control
volume. The size of this control volume is chosen to be greater than the representative elementary
volume wherein the macroscopic average velocity holds as an accurate representation of the fluid flow.

The law of conservation of mass for steady-state flow through a saturated porous medium requires that
the fluid mass flux into any elementary control volume must be equal to the fluid mass flux out of the
control volume - assuming that there is no change in the quantity of mass stored in the control volume.

From Figure 1.9 the mass inflow rate through the face ABCD is Jx , where J is the mass flux (kg/m2 /s).

In general, the mass outflow can be different from the mass inflow rate. The mass outflow rate through
the face EFGH and can be written as  
∂(Jx )
Jx + (1.27)
∂x
where the second term represents the change in mass flux in the x direction.

The net outflow rate is then the difference between the inflow and outflow rates. The net outflow rate
through EFGH is then
∂(Jx )
− (1.28)
∂x

Making similar calculations for the remaining faces of the cube, the net outflow through CDHG is

∂(Jy )
− (1.29)
∂y

and through BCGF is

∂(Jz )
− (1.30)
∂z

Water Resources Engineering CVEN3501 19

Lecture 1 Groundwater resources and flow equations

Adding these components of the mass flux in the three co-ordinate directions gives the net mass outflow
rate through all the faces:  
∂(Jx ) ∂(Jy ) ∂(Jz )
− + + (1.31)
∂x ∂y ∂z

1.8.3 Steady State vs Non-Steady State

If the components of Equation 1.31 sum to zero - the system is said to be in steady state. This does not
mean that there is no flow of water through the system, simply that the same amount of water that enters
the REV leaves the REV. There is no change in the amount of water stored in the REV. All groundwater
systems can be considered to be in steady state if the time interval is long enough (possibly hundreds
of years).

If the components of Equation 1.31 do not sum to zero, then there has to be some change in the
quantity of mass (water) stored in the aquifer. The system is said to be in non-steady state and to solve
the equation we need to consider the ways in which the water can be stored in the aquifer. This will be
returned to in detail later. Many groundwater systems are in non-steady state over the short term.

1.9 Steady-State Flow

If we recognise that J = ρυ (mass flux is the product of the volume flux times density) then we can write
Equation 1.31 for the steady state condition as:
 
∂(ρυx ) ∂(ρυy ) ∂(ρυz )
− + + =0 (1.32)
∂x ∂y ∂z

If we assume that the fluid density does not change spatially, the density term on the left-hand side of
the equation can be taken outside of the partial differential operators so that Equation 1.32 becomes:
 
∂(υx ) ∂(υy ) ∂(υz )
−ρ × + + =0 (1.33)
∂x ∂y ∂z

Clearly ρ 6= 0 and therefore the terms in Equation 1.33 can then be further simplified. Substitution
of Darcy’s Law for the υx , υy , and υz components yields an equation for steady-state flow through an
anisotropic saturated porous medium as:
     
∂ ∂h ∂ ∂h ∂ ∂h
Kx + Ky + Kz =0 (1.34)
∂x ∂x ∂y ∂y ∂z ∂z

The expression is now positive because the specific discharge terms are negative in Darcy’s equation.
Water moves from higher energy to lower energy, yet the coordinate system is in the reverse direction,
hence the negative sign.

For an isotropic medium Kx = Ky = Kz , and if the medium is also homogeneous then K(x, y, z) =
a constant. The equation then reduces to
      
∂ ∂h ∂ ∂h ∂ ∂h
K + + =0 (1.35)
∂x ∂x ∂y ∂y ∂z ∂z
or
∂2h ∂2h ∂2h
 
K + 2 + 2 =0 (1.36)
∂x2 ∂y ∂z
Equation 1.36 reduces to Laplace’s Equation when the hydraulic conductivity is homogeneous and
isotropic which, as K 6= 0, gives :
∂2h ∂2h ∂2h
+ 2 + 2 =0 (1.37)
∂x2 ∂y ∂z

Water Resources Engineering CVEN3501 20

Lecture 1 Groundwater resources and flow equations

The representation of the water flux as shown in Equation 1.37 forms the basis of groundwater modelling
for steady state flow and is the basis behind the assignment. The solution to this equation describes the
value of the hydraulic head at any point in a three-dimensional flow field.

1.10 Finite Difference Solution Methods to the Steady-State Flow

Problem

1.10.1 Introduction

The steady-state flow equation can be solved using finite difference methods.

The finite difference methods are fundamentally very simple. The approximation of the continuous vari-
able leads directly to a set of simultaneous equations without the need to introduce further mathematical
complications.

The geometry of a two-dimensional solution is maintained in the computer code or set out using a spread
sheet. This means that the approach is ideally suited to a 2-dimensional representation of the aquifer.
Two-dimensional arrays are used to store aquifer properties and heads which closely resemble the field
situation. The extension to three dimensions comes naturally as additional layers of information are
added.
2
Finite difference representations of the first order ∂h ∂ h
∂x and second order ∂x2 partial derivatives need to
be developed. The finite difference approximation to the smooth and continuous function can best be
appreciated in terms of gradients (Feynman et al., 1969).

∂2h
1.10.2 FD Discretisation of ∂x2

The use of the finite difference method allows the discretisation of the continuous partial differential
equation. Instead of values of head specified at all points in the x − y − z space, the space is covered by
a two-dimensional grid and the head only calculated for nodes of the grid. If it is necessary to examine
a particular area in detail, then the mesh is increased in density in the region of interest. Conversely, if
large areas of the problem domain are of little interest, then a coarse mesh can be chosen in this area.

As the mesh is orthogonal there are some limitations in choosing the most appropriate mesh and this is
one of the weaknesses of this method.

For a given portion of an aquifer, the smooth hydraulic head variation can be described by a number of
straight line segments as shown in Figure 1.10.

The accuracy with which the smooth function can be represented is determined by the number of straight
line segments into which it is divided. Approximate values of the gradients (first derivative) at intermedi-
ate points in Figure 1.10 can be calculated using the point centered grid system as
 
dh hi+1 − hi
= (1.38)
dx a ∆xa

and

 
dh hi − hi−1
= (1.39)
dx b ∆xb

Where the suffixes xa and xb signify intermediate positions on either side of node 0. Values of the

Water Resources Engineering CVEN3501 21

Lecture 1 Groundwater resources and flow equations

Figure 1.10: Continuous function and finite difference approximation

second differential can be written as

dh dh



dx a − dx b
 2     
d h d dh 2 hi+1 − hi hi − hi−1
= = = − (1.40)
dx2 0 dx dx 0.5(∆xa + ∆xb ) ∆xa + ∆xb ∆xa ∆xb

If the mesh spacing is regular and ∆xa = ∆xb = ∆x then the expressions simplify to:
d2 h hi+1 − 2hi + hi−1
= (1.41)
dx2 ∆x2

The approximation introduced by the finite difference approach can best be examined by considering
the Taylor series expansion of the terms.
∆x2 d2 h ∆x3 d3 h ∆x4 d4 h
       
∆x dh
hi+1 = hi + + + + + ··· (1.42)
1! dx 0 2! dx2 0 3! dx3 0 4! dx4 0

∆x2 d2 h ∆x3 d3 h ∆x4 d4 h

       
∆x dh
hi−1 = hi − + − + − ··· (1.43)
1! dx 0 2! dx2 0 3! dx3 0 4! dx4 0

Summing:

∆x2 d2 h ∆x4 d4 h
   
hi−1 − 2hi + hi+1 =2× +2× + ··· (1.44)
2! dx2 0 4! dx4 0

Therefore, dividing by ∆x2 gives:

 2 
∆x4 d4 h
 
d h 1
= (hi−1 − 2hi + hi+1 ) − + ··· (1.45)
dx2 0 ∆x2 12 dx4 0

4
 
d4 h
The terms − ∆x
12 dx4 are called the truncation error. If two solutions are obtained, the second having
0
a mesh interval equal to half that for the first solution, the truncation error for the second solution is
only one quarter of that of the first. As a result the second solution will be more accurate. For certain
problems the truncation error is zero; then the finite difference solution is identical with the theoretical
solution.

Water Resources Engineering CVEN3501 22

Lecture 1 Groundwater resources and flow equations

1.10.3 FD Representation of Laplace’s Equation in 2D

The 3D representation of steady-state groundwater flow (Equation 1.34) and now including a flux term
(R) that could account for rainfall recharge or discharge to springs, or groundwater abstraction was:
     
∂ ∂h ∂ ∂h ∂ ∂h
Kx + Ky + Kz +R=0 (1.46)
∂x ∂x ∂y ∂y ∂z ∂z
The sign of the flux term (R) will depend upon whether we are dealing with positive flow into the model
(rainfall) or negative flow out of the model (spring flow or groundwater abstraction).

If there is no flow in the vertical direction then the last differential term (in z) in this equation disap-
pears. However, we can not ignore the aquifer thickness b and it is necessary to multiply the hydraulic
conductivity by this thickness to derive a new parameter T , the transmissivity. This is the equivalent of
integrating over z. The equation representing horizontal flow in a 2-dimensional aquifer in steady-state
is then:

   
∂ ∂h ∂ ∂h
Tx + Ty +R=0 (1.47)
∂x ∂x ∂y ∂y
where:

L2
T is the product of the hydraulic conductivity and the aquifer thickness. It has dimensions of T .

If the transmissivity is homogeneous and isotropic it can be removed from the partial derivatives, viz:

∂2h ∂2h
 
T + 2 +R=0 (1.48)
∂x2 ∂y

There may still be a constant flux into and out of the elementary volume. This will appear as a separate
term in Equation 1.48.

If the constant flux term is zero, Equation 1.48 will collapse to the Laplace Equation in 2-D.
∂2h ∂2h
+ 2 =0 (1.49)
∂x2 ∂y

However, if the constant flux term R is non-zero, the transmissivity does not dissapear and we will have:

∂2h ∂2h R
+ 2 + =0 (1.50)
∂x2 ∂y T

We can first concentrate on the differential terms in Equation 1.50. Consider a finite set of points ar-
ranged orthogonally on a regular mesh as shown in Figure 1.11.

The space between the points in the -x- and -y- directions is constant. The smaller this distance be-
comes, the better the accuracy of the approximation. However, as the number of mesh points and
equations increases, the solution time also rapidly increases. To locate any point in the grid we specify
an integer ordered pair (i, j), relative to the pair (1, 1) located in the upper left-hand corner of the grid.
The finite difference expressions can be written in terms of the nodes hi−1,j , hi,j , hi+1,j by taking a
2
central difference approximation to ∂∂xh2 at the point x0 , y0 . These are obtained by approximating the first
derivative at x0 + ∆x ∆x
2 , y0 and at x0 − 2 , y0 , and by then obtaining the second derivative by taking a
difference between the first two differences at those points. That is,

hi+1,j −hi,j hi,j −hi−1,j

∂2h ∆x − ∆x
= (1.51)
∂x2 ∆x

Water Resources Engineering CVEN3501 23

Lecture 1 Groundwater resources and flow equations

Figure 1.11: Finite difference grid - showing the 5-point FD star operator

which may be written as

∂2h hi−1,j − 2hi,j + hi+1,j

2
= (1.52)
∂x ∆x2

Similarly, the relationships for the approximation in the -y- cartesian coordinate direction can be written
as:
∂2h hi,j−1 − 2hi,j + hi,j+1
2
= (1.53)
∂y ∆y 2

The two components are added together to represent Laplace’s equation in two dimensions. For a
regular grid such that ∆x = ∆y, the finite difference expression for the point i, j simplifies to

∂2h ∂2h hi−1,j − 2hi,j + hi+1,j + hi,j−1 − 2hi,j + hi,j+1

2
+ 2 = =0 (1.54)
∂x ∂y ∆x2

Multiplying through by ∆x2 gives:

hi−1,j + hi,j+1 + hi+1,j + hi,j−1 − 4hi,j = 0 (1.55)

and solving for hi,j gives:

hi,j = 0.25 × (hi−1,j + hi,j+1 + hi+1,j + hi,j−1 ) (1.56)

This is equivalent to saying that the head at a central node may be approximated by taking the average
of the four surrounding nodes. Note that for the node i, j the diagonal nodes hi−1,j−1 , hi+1,j−1 , hi−1,j+1
and hi+1,j+1 do not contribute to the approximation.

Water Resources Engineering CVEN3501 24

Lecture 1 Groundwater resources and flow equations

Figure 1.12: Workshop Problem

The finite difference method can be used to calculate the hydraulic head distribution for any size model.
As an example, write the finite difference expressions for the heads at nodes 2, 2, 2, 3, 3, 2 and 3, 3 in
Figure 1.12.

Now, if we include the constant flux term and rearrange the terms we get:

∂2h ∂2h hi−1,j − 2hi,j + hi+1,j + hi,j−1 − 2hi,j + hi,j+1 R

+ 2 = =− (1.57)
∂x2 ∂y ∆x2 T

Solving this for the unknown head at node (i,j) gives;

R∆x2
 
hi,j = 0.25 × hi−1,j + hi,j+1 + hi+1,j + hi,j−1 + (1.58)
T

but remember that the sign of the additional flux term will be determined by whether it represents flow
into the model (positive) or flow from the model (negative)

1.10.4 Iterative Solution Methods

The simplest iteration method is to assume some starting point for each of the four unknown heads in the
above problem and to apply the finite difference approximation to each unknown head in turn, repeatedly
sweeping through the mesh until the change in head at each unknown is less than some predetermined
acceptability criteria. This method was developed by Southwell and is often referred to as the relaxation
method because for each step one of the errors is relaxed.

If the iteration index is referred to as m, then the formula to be used for a regular mesh in which ∆x = ∆y
can be written as:
hm m m m
i−1,j + hi,j−1 + hi+1,j + hi,j+1
hm+1
i,j = (1.59)
4

Water Resources Engineering CVEN3501 25

Lecture 1 Groundwater resources and flow equations

Figure 1.13: Conceptual model and the equivalent mathematical model of a simple groundwater flow
system

Use of this formula at each node in succession is known as the Jacobi Iteration Method.

A more efficient method is to sweep through the mesh in an ordered manner from left to right and top to
bottom, using the most recently computed heads at the left and top in the computation of the new head
at the next iteration for node (i, j). This method of solution is known as the Gauss Seidel Method. This
method allows the use of two newly computed values in the iteration formula, viz:

hm+1 m+1 m m
i−1,j + hi,j−1 + hi+1,j + hi,j+1
hm+1
i,j = (1.60)
4

The use of newly computed head values when ever possible makes the Gauss-Seidel method more
efficient than Jacobi iteration.

1.10.5 Treatment of Boundary Conditions

A conceptual model and the related mathematical models of flow in a 2-D aquifer are shown in Fig. 1.13
(Wang and Anderson, 1982).

The mathematical problem is specified by Equation 1.47 inside the flow domain represented by the
model. To complete the mathematical specification, the conditions around the boundary and at the start
of the simulation must be specified.

Two basic conditions can exist around the boundary, viz:

• A known flow (flux) across the boundary, or

• a fixed head on the boundary.

A specified flow across the boundary is referred to as a Neumann boundary co ndition. The most com-
mon boundary condition of this type is a no-flow boundary condition. Consideration of the basic Darcy
equation indicates that no flow will occur when the gradient of the head across the boundary is zero.

∂h
=0 (1.61)
∂x

Water Resources Engineering CVEN3501 26

Lecture 1 Groundwater resources and flow equations

A specified head along a boundary is known as a Dirichlet boundary condition. Specification of the
Dirichlet condition is achieved by setting the head to a known value on the boundary and not allowing
it to be changed during the solution procedure. Implementation of a fixed head condition is significant
in groundwater modeling as it creates an unlimited source of water into the aquifer, no matter what
hydraulic gradient occurs.

Water Resources Engineering CVEN3501 27

Lecture 2

Aquifer properties and borehole

construction

2.1 Storage Properties of Porous Rocks

2.1.1 Introduction

If the flow system is not in steady state and water is taken into or released from the elementary volume,
then we need to consider how this occurs. What will be the impact on the hydraulic head if a certain
quantity of water is added to or taken from the volume. Mathematically, the change can be represented
by:

∂(ρυx ) ∂(ρυy ) ∂(ρυz ) ∂(ρφ)

+ + = (2.1)
∂x ∂y ∂z ∂t
Equation 2.1 states that the change in storage is described by the way in which the porosity (φ) and
density (ρ) of the water change as a function of time. The full analysis of this is complicated by the
presence of a deforming aquifer skeleton.

Presumably the hydraulic head will increase as water is taken into storage and will fall as it is taken out of
storage. The change in mass flux in the elementary volume should be proportional to the rate of change
of hydraulic head with time, mathematically:

∂(ρυx ) ∂(ρυy ) ∂(ρυz ) ∂h

+ + ∝ (2.2)
∂x ∂y ∂z ∂t

The constant of proportionality is known as the specific storage (Ss ) and after substituting Darcy’s equa-
tion for the volume flux we can write the relationship as:
     
∂ ∂h ∂ ∂h ∂ ∂h ∂h
Kx + Ky + Kz = Ss (2.3)
∂x ∂x ∂y ∂y ∂z ∂z ∂t

where:

Ss is specific storage

If the hydraulic conductivity is homogeneous (does not vary in x-y-z) we can take the K out of the partial

28
Lecture 2 Aquifer properties and borehole construction

differentials, cross multiply and write:

∂2h ∂2h ∂2h Ss ∂h

+ 2 + 2 = (2.4)
∂x2 ∂y ∂z K ∂t

2.1.2 Specific Storage for a confined aquifer

The Specific Storage Ss of a confined aquifer is defined as the volume of water that a unit volume of
aquifer releases for a unit decline in hydraulic head. The water that is released from storage under
conditions of decreasing h is produced by two mechanisms:

• the compaction of the aquifer caused by increasing σe , and

• the expansion of the water caused by decreasing pressure.

The first mechanism is controlled by the aquifer compressibility α, while the second is controlled by the
fluid compressibility β.

The specific storage Ss is the sum of these two factors, viz:

Ss = ρw g(α + φβ) (2.5)

Typical values of Ss range from 10−4 to 10−7 .

For example if a cubic metre of aquifer has a specific storage of 10−6 then one milliliter of water will be
released from storage due to a one metre decline in head. It is evident that very considerable volumes
of material require depressurising to obtain useful yields of water.

2.1.3 Storage in an unconfined aquifer (Specific Yield)

The volumetric expansion of water and the compaction of the aquifer skeleton also occur in an uncon-
fined aquifer when water is removed by pumping. The quantity of water removed is negligible compared
to the water obtained by physically draining the pore space.

The storage term for unconfined aquifers is known as the specific yield Sy .

The concept of specific yield can be shown in terms of moisture contents in the unsaturated zone.

The curve T1 in Figure 2.1 represents the moisture stored above the water table at time t1 , and the
curve t2 the moisture stored at some later time (lower water level). The effect of lowering the water table
causes the moisture stored in the hatched part of Figure 2.1 to be released as drainage. The rate at
which this drainage occurs differs for different soils. In general, coarse soils will drain quickly whereas
fine soils may take many days to drain.

Delayed drainage is a concept developed to account for the slower release of moisture from soils during
pumping tests and is directly related to the rate at which water drains from the soil.

The specific yield will only be equal to the porosity for very clean coarse gravels. For finer material, a
proportion of the water contained at saturation is held back in the soil by surface tension effects. The
quantity is sometimes referred to as specific retention where the sum of the specific yield and the specific
retention is equal to the porosity.
φ = Sy + Sr (2.6)
Field capacity moisture content is analogous to the amount of water contained in the soil when the
specific yield has drained away.

Water Resources Engineering CVEN3501 29

Lecture 2 Aquifer properties and borehole construction

Figure 2.1: Specific Yield

The specific yields of unconfined aquifers are much higher than the storativities of confined aquifers. The
usual range of Sy is 0.01 - 0.30. Values of specific yield for a range of materials are given in Table 2.1.

Table 2.1: Specific yields for various geologic materials

Material Specific Yield

Coarse gravel 23
Medium gravel 24
Fine gravel 25
Coarse sand 27
Medium sand 28
Fine sand 23
Silt 8
Clay 3
Fine grained sandstone 21
Medium grained sandstone 27
Dune sand 38
Loess 18
Peat 44

2.2 Effective Stress and the impact of dewatering

Assume that a stress is applied to a unit mass of saturated sand. There are three mechanisms by which
a reduction in volume can be achieved:

• By compression of the water in the pores

• By compression of the individual sand grains

Water Resources Engineering CVEN3501 30

Lecture 2 Aquifer properties and borehole construction

Figure 2.2: Stress distribution about a horizontal plane within the earth

• By a rearrangement of the sand grains into a more closely packed configuration.

The first mechanism is controlled by the compressibility of water, β. The second is negligible, that is,
the individual soil grains are practically incompressible. The concept of effective stress is involved in the
third mechanism.

Consider the stress equilibrium on an arbitrary plane through a saturated formation at depth as shown
in Figure 2.2.

In Figure 2.2 σT is the total stress (pressure) acting downward on the plane and is due to the weight of
overlying rock and water. This stress is carried in part by the granular skeleton of the porous medium
and in part by the fluid pressure p of the water in the pores.

The portion of the total stress which is carried by the granular skeleton is called the effective stress -
σe . It is this stress which is actually applied to the grains of the porous medium and rearrangement of
the soil grains and the resulting compression of the granular skeleton is caused by changes in effective
stress, not by changes in the total stress. The two are related simply by

σT = σe + p (2.7)

or in terms of the changes

dσT = dσe + dp (2.8)

In most cases, the removal of groundwater does not change the total stress applied to the top of an
aquifer. The weight of rock and water overlying each point remains essentially constant through time. In
such cases dσT = 0 and
dσe = −dp (2.9)

If the fluid pressure increases, the effective stress decreases by an equal amount.

For cases where the total stress does not change with time, the effective stress at any point in the
system, and the resulting volumetric deformations there, are controlled by the fluid pressures at that
point.

Since p = ρw gψ and ψ = h − z changes in the effective stress are governed by changes in the hydraulic
head at that point.
dσe = −ρgdψ = −ρgdh (2.10)

The analysis of effective stress is important in the calculation of slope stability, the response of aquifers
to changes in atmospheric pressure and in subsidence studies.

Water Resources Engineering CVEN3501 31

Lecture 2 Aquifer properties and borehole construction

2.2.1 Effect of Groundwater Conditions on Slope Stability in Soils

The stability of a soil slope is preserved by ensuring that the pore water pressure is always less than the
effective stress component in the soil. The soil will effectively ’float’ if this occurs. Where the soil occurs
on a slope, and a potential sliding plain exists, then slope failure will occur as a result.

It is possible to calculate the pore water pressure from a measurement of the hydraulic head from the
relationship:
p = ρg(h − z) (2.11)
where:

p is the pore water pressure,

ρ is the water density,

g is acceleration due to gravity,
h is hydraulic head, and
z is the elevation of the piezometer intake above the datum (the elevation head)

If an analysis is carried out on a slope at the bottom of a valley, which may be a groundwater discharge
location, then a simple analysis using a local flow net will underestimate the pore water pressure and
may not accurately predict failure on the slope.

2.2.2 Compressibility of a Porous Medium

The compressibility of a porous medium is defined as

−dVT /VT
α= (2.12)
dσe

where:

VT is the total volume of a soil mass and,

dσe is the change in effective stress.

The soil compressibility α is usually determined from the slope of a strain-stress plot obtained by labora-
tory tests (Freeze and Cherry, 1979). The compressibility values for clays under loading and unloading
can differ by an order of magnitude. For sands the ratio is closer to one. For clays that have considerable
ranges of α values, the compression which occurs under compaction is largely irreversible.

Typical values of compressibility are given in Table 2.2 (Domenico and Schwartz, 1997).

To illustrate the nature of the deformations that can occur in compressible aquifers, consider an aquifer
of thickness b. If the weight of overlying material remains constant and the hydraulic head in the aquifer
is decreased by an amount −dh, the increase in effective stress dσe is given as ρgdh and the aquifer
compaction is then
db = −αbdσe = −αbρgdh (2.13)

The minus sign indicates that the decrease in head produces a reduced thickness b. The α term is
the aquifer compressibility. In general it is a variable quantity with α = αx,y,z . The compaction of the
aquifer itself is relatively small. Greater compaction occurs in the silty clay (aquitard) material between
the sands.

Water Resources Engineering CVEN3501 32

Lecture 2 Aquifer properties and borehole construction

Table 2.2: Values of Compressibility

Material α
(m2 /N )

plastic clay 2.0 × 10−6 to2.6 × 10−7

Stiff clay 2.6 × 10−7 to1.3 × 10−7
Medium-hard clay 1.3 × 10−7 to6.9 × 10−8
Loose sand 1.0 × 10−7 to5.2 × 10−8
Dense sand 2.0 × 10−7 to1.3 × 10−8
Dense, sandy gravel 1.0 × 10−8 to5.2 × 10−9
jointed rock 6.9 × 10−10 to3.3 × 10−10
sound rock < 3.3 × 10−10
Water at 25 ◦ C 4.8 × 10−10

2.2.3 Land Subsidence

When groundwater is abstracted from an unconsolidated sedimentary sequence, typical of any recent
accumulation of sediments, significant ground subsidence has been noted. This subsidence has aver-
aged one metre every three years for a 35 year period (1935 - 1970) in several groundwater basins.
The occurrence is well documented at Long Beach in California, where groundwater and oil have been
pumped from aquifer systems; in the San Joaquin Valley (San Francisco Silicon Valley) and in Mexico
City.

Subsidence is also well documented at Morwell in Victoria.

In unconsolidated formations, significant quantities of water exist in the aquitards such that abstraction
from the high hydraulic conductivity aquifers almost acts as a drain to the larger quantities of more slowly
released water in the aquitards. As the groundwater is released from the aquitards, compaction of the
clays occurs causing subsidence.

The consolidation of the clays leads to a rearrangement of the clay plates and an expulsion of the water
contained in the clay porosity. This process is irreversible!

Water Resources Engineering CVEN3501 33

Lecture 2 Aquifer properties and borehole construction

2.3 Well Drilling Methods

2.3.1 Hand Dug Wells

Hand dug wells have been a major source of water for thousands of years. Recent reports of archae-
ological investigations in Shanghai indicate wells dating back to 4000 years BC. Hand dug wells still
supply a large proportion of the world population.

As the depth to water increased, various well drilling methods have been developed. Water bores are
constructed in materials ranging from hard basalt to unconsolidated sands and gravels. A wide variety
of drilling methods has been developed to cater for this range of conditions. The cable tool and direct
rotary methods are described in these notes.

Well construction comprises five stages:

1. drilling,
2. installation of the casing,
3. placing the screen and gravel pack,
4. grouting to provide sanitary protection, and

5. development of the well to provide sand-free operation.

In cable-tool drilling, the first two stages are performed simultaneously.

2.4 Cable Tool Drilling

2.4.1 Introduction

The cable-tool percussion method was developed by the Chinese and has changed little in the last 4000
years. With better engineering and manufacturing skills, the materials used have evolved, but the basic
method of raising and letting fall a heavy weight suspended by a cable has remained.

Figure 2.3 shows a schematic of a Ruston Bucyrus cable-tool rig.

In cable tool drilling, the drill bit is attached to the lower portion of the weighted drill stem that, in turn, is
attached by means of a rope socket to the rope or cable. The cable and drill stem are suspended from
the mast of the drill rig through a pulley (Fig. 2.4).

The cable runs through another pulley that is attached to a spudding beam. The beam is powered by
the rig engine and moves up and down, the bit is alternately raised and dropped by this action. Given
time, this method can penetrate all geological formations. A full string of cable tool drilling equipment
consists of five components:

1. drill bit,
2. drill stem,

3. drilling jars,
4. swivel socket, and
5. cable.

Water Resources Engineering CVEN3501 34

Lecture 2 Aquifer properties and borehole construction

Figure 2.3: Schematic of a cable tool rig

Water Resources Engineering CVEN3501 35

Lecture 2 Aquifer properties and borehole construction

Figure 2.4: Cable tool drilling rig.

Water Resources Engineering CVEN3501 36

Lecture 2 Aquifer properties and borehole construction

Each of the tools is an integral part of the cable-tool system. The drill bit is usually a massive steel bit
with a chisel built into the base. Water flow channels are formed on either side of the bit to allow the
slurry to pass away from the cutting face. The drill stem adds weight to the bit and also increases the
length of the assembly. This increased length helps to keep the hole straight.

The drill jars consist of a pair of linked steel bars. When the bit is struck, it can be freed most of the time
by upward blows of the free-sliding jars. The stoke of the drilling jars is 230 mm to 450 mm. The swivel
socket connects the string of tools to the cable. The socket also transmits the rotation of the cable to the
drill it. In this way, each subsequent stoke of the drill bit is rotated from the previous stroke, and keeps
the hole circular.

The wire cable that carries and rotates the drilling tool is called the drill line. It is 16 mm to 25 mm
in diameter and is a left-hand lay cable that twists the tool joint on each upstroke to prevent it from
unscrewing. The cable is suspended over the top of the mast, down to the spudding sheave on the
walking boom (Fig. ref.

The drill bit crushes and breaks material at the base of the hole in consolidated formations, or loosens
material in unconsolidated formations. Water mixes with the crushed material to form a slurry. In dry
conditions, water is added to the hole to help make a slurry.

The drill bit should strike the bottom of the hole at the extreme (elastic) limit of the cable so that it
immediately snaps upward. In this manner, the base of the hole is given a sharp blow which causes
maximum damage to the formation. The rig is protected from vibration by the inclusion of a shock
absorber in the mounting of the drill line.

Sample is recovered to the surface for logging (Fig 2.5).

When the drilling progress is slowed by the presence of too much slurry, the cuttings are removed from
the bottom of the hole by a bailer. This is generally a hollow metal cylinder fitted with a flap valve at the
base. The drilling tools are removed from the hole and the bailer run in on a separate line.

Tools for removing lost materials or components of the drill string from the base of the hole have been
developed for cable tool methods. The process is referred to as fishing.

2.4.2 Drilling Consolidated Formations

In consolidated materials it is possible to drill without a supporting casing. This method is called open
hole drilling. In these formations the drill bit is effectively a crusher.

2.4.3 Drilling Unconsolidated Formations

In unconsolidated materials, a combination of pushing a casing into the ground and bailing the cuttings
from the inside can be used. First, a casing is driven into the ground using a piling technique, then
the unconsolidated material is loosened by drilling and removed with a bailer. The depth of drilling is
determined by the friction on the outside of the casing. For this reason, a number of casing reductions
are often used, with the smaller casing telescoped inside the larger casing. The maximum diameter and
weight of casing is determined by the strength of the drilling rig. This determines the maximum depth of
drilling.

Water Resources Engineering CVEN3501 37

Lecture 2 Aquifer properties and borehole construction

Figure 2.5: Geologist examining and logging the drill returns

Water Resources Engineering CVEN3501 38

Lecture 2 Aquifer properties and borehole construction

2.4.4 Advantages and Disadvantages of Cable Tool Drilling

Advantages

The following advantages of the cable tool method are given by Driscoll (1986).

1. Rigs are relatively inexpensive

2. Rigs are simple in design and easy to maintain
3. Rigs have low energy requirements
4. The borehole is stabilised throughout the operation.
5. It is possible to recover reliable samples from the base of the hole throughout the drilling operation,
unless heaving of formations is encountered.
6. Wells can be drilled in areas where makeup water supplies are limited
7. Wells can reconstructed with minimal contamination
8. The driller maintains intimate contact with the drilling process and the materials encountered by
keeping a hand on the drilling cable
9. The rig can be operated by one person, although a helper is usually available to assist
10. The rig can be assembled in areas where space is limited, or carried into inaccessible terrain
11. Rigs can be operated under all temperatures and weathers
12. Wells can be drilled through formations where lost of circulation is a problem
13. Wells can be bailed at any time to get an idea of the formation yield

Disadvantages

Some disadvantages of the cable tool method include:

1. Penetration rates are relatively slow

2. Casing costs are higher because heavier wall or larger diameter casing may be required
3. It may be difficult to remove construction casing in some formations without access to heavy lifting
jacks.

2.5 Rotary Drilling

2.5.1 Introduction

Rotary drilling methods were developed for the oil and minerals industries and have been adapted for
the groundwater industry. A number of variations exist, including:

• Direct mud rotary,

• Reverse circulation mud rotary,
• Air-flush rotary, and

Water Resources Engineering CVEN3501 39

Lecture 2 Aquifer properties and borehole construction

Figure 2.6: Rotary Drilling Rig

• Down-hole air hammer.

The direct rotary drilling method was developed to increase penetration rates over those available by
cable tool drilling. The borehole is advanced by rotating a cutting bit at the base of the hole. The bit
breaks the formation and the cuttings are removed by a drilling fluid which is pumped down the inside of
the drill pipe and returns to the surface up the hole outside the drill pipe. The bit is attached to the rig by
lengths of drill pipe which transfer the rotation from the surface to the bit.

2.5.2 Drilling Equipment

The basic components of a rotary drilling rig are shown in Figure 2.6and a rig deployed to site is shown
in Figure 2.7.

Rotation of the drill string is achieved by one of two methods:

• Through a rotary table which transfers torque to the drill string through a special square cross

Water Resources Engineering CVEN3501 40

Lecture 2 Aquifer properties and borehole construction

Figure 2.7: Rotary rig at the Cattle Lane investigation site on the Liverpool Plains

section of drill pipe (the kelly) which is always a the top of the drill string. As the hole is progressed,
the kelly slides through the rotary table.
• By using a top drive assembly. Typically, a hydraulic drive is attached to the top of the drill string
and moves downward as the hole is advanced.

2.5.3 Drill Bits

The drill bit can be changed depending upon the formation. A drag bit is used in soft formations or a
tricone roller bit used in hard formations. The drill bit slowly wears away and needs to be replaced. The
replacement is achieved by pulling all the drill pipe sections out of the hole, changing the bit, and putting
all the drill pipe back again. This can be a time consuming exercise. New rigs are specifically designed
to minimise the time taken to run the drill rods into the hole. Spare drill bits are shown at a drilling site in
Figure 2.8.

2.6 Drilling Fluids

2.6.1 Introduction

The use of drilling fluids in the construction of bores is a large field in its own right. The subject is covered
in detail by Driscoll (1986).

The first drilling fluid utilised in rotary drilling was water. The entrainment of natural clay particles pro-
duced a mud and gave rise to the term drilling mud. Later on, specific clay compounds were deliberately
added to the mud to control the mud properties. The technology has developed to a point where many
different compounds can be added to the fluid to control its physical and chemical characteristics.

Water Resources Engineering CVEN3501 41

Lecture 2 Aquifer properties and borehole construction

Figure 2.8: Drill-bits for a rotary drilling rig

2.6.2 Types of drilling fluids

Water based fluids are used in the water-well drilling industry. Water-based drilling fluids consist of the
following phases:

• a liquid phase,
• a suspended-solid (colloidal) phase, and

• cuttings entrained during the drilling.

Air-based drilling fluids may consist of only air, but more commonly have a surfactant added to create a
foam.

There are many different additives commonly used, including:

• flocculants to precipitate clay particles,

• dispersant (thinning agents),

• weighting materials to increase the fluid density,

• corrosion inhibitors,
• lubricants,

• bactericide, and
• lost-circulation materials.

Water Resources Engineering CVEN3501 42

Lecture 2 Aquifer properties and borehole construction

The exact drilling fluid system chosen for a given project will depend upon the geology of the area and
the type of materials to be drilled. In general, air-based systems are used with rotary-hammer rigs
to drill through hard-rock formations, while water-based fluids with clay or polymer additives are used
in unconsolidated or soft sedimentary formations. Clean water is used for reverse circulation drilling
equipment.

2.6.3 Functions of a drilling fluid

The primary functions of a drilling fluid are :

• The removal of drill cuttings from beneath the drill bit to the surface. The rate at which cuttings are
removed depends upon the viscosity, density and up-hole velocity of the drilling fluid, and the size,
shape and density of the cuttings.
• The drilling fluid preserves the stability of the borehole wall and prevents expansion of swelling
clays. Water based fluids must provide a positive pressure to the borehole walls to prevent caving
and collapse of the formation. If collapse does occurs, the drill bit and string can become trapped
and the hole abandoned.
• The drilling fluid serves to both cool and lubricate the drill bit.
• The correct drilling fluid can control losses of fluid to the formation by creating a highly impermeable
cake of drill mud on the formation wall. The invasion of drilling fluids into the formation must
however, be minimised, as this material has to be removed from the formation during the well
development process.
• The physical characteristics of the drilling fluid require adjustment so that the drill cuttings drop out
readily at the surface and are not returned down the hole by the fluid circulation.
• The drill fluid returns the cuttings to the surface where the geologist may be trying to form a
geological log of the formations penetrated. The fluid should act to disguise the true nature of the
cuttings as little as possible.
• Lastly, the fluid should be able to suspend the cuttings in the borehole if the circulation of fluid
is halted for any reason. This ability is important as the settling to the bottom of the hole of the
cuttings in a deep bore can easily trap the drill string.

In many cases the drilling fluid will be unable to satisfy all the above criteria and continuous monitoring
of the physical properties of the fluid is advisable. The drilling fluids are commonly mixed in mud pits
dug close to the drill rig. Figure 2.9 shows a typical drilling site with mud pits dug around the rig to form
the mud flow system.

2.7 Well Design

2.7.1 Introduction

The objective of good well design is to install a well into the drilled borehole that has the minimum
possible head loss occurring as water passes from the formation into the well. Aspects of good well
design are shown in Fig. 2.10.

2.7.2 Introduction

The well screen serves to support the formation and allow water to pass from the formation to the well.
Important screen functions and criteria include:

Water Resources Engineering CVEN3501 43

Lecture 2 Aquifer properties and borehole construction

Figure 2.9: A typical drilling site showing the drill pipes stacked in the foreground, the rig in the back-
ground and the mud pits. The rig is tilted so that geophysical logging can be carried out in the borehole
to obtain lithological data for the design of the abstraction bore.

• The screen should be capable of easy development

• there should be a minimum tendency for incrustation

• there should be a low head loss through the screen

• there should be a large percentage of open area
• the slots should have no tendency to clog

• the screen material should be resistant to corrosion

• the screen column should have sufficient compressional and collapse strength

Well screens can be made by cutting slots in existing plastic or steel casing, or by assembling plastic
segments to create a screen, or from specially manufactured wire-wound screens. An example of a wire
wound screen is shown in Figure 2.11 after Driscoll (1986). The choice of screen material will depend
upon the chemistry of the well water, the depth of the well and the purpose the well is constructed for.

2.7.3 Slot Size

To achieve the best well efficiency, the percentage open area of the well screen should be equal to or
exceed the porosity of the aquifer material. Water flows more evenly, with less head loss, through a
screen with a high percentage open area.

Screens with large open areas and low entrance velocities are less likely to suffer encrustation due to
the small pressure drop through the screen.

Screens are available in two basic sizes relative to the size of the casing. Telescopic screens are
designed to be installed inside the casing. Pipe-size screens are installed at the end of the casing and
must therefore be installed as part of the casing string.

Water Resources Engineering CVEN3501 44

Lecture 2 Aquifer properties and borehole construction

Figure 2.10: Basic design of an abstraction bore

Water Resources Engineering CVEN3501 45

Lecture 2 Aquifer properties and borehole construction

Figure 2.11: Wire Wound Screen

2.8 Gravel Packs

In formations where the grain size is very uniform, it is necessary to place a graded gravel pack between
the aquifer formation and the screen. If this is not done then the grains in the aquifer will be continuously
pumped through the screen. It will not be possible to produce a sand free water at the surface.

Placement of a gravel pack requires a larger constructed diameter of the borehole. This means a heavier
rig, more drilling fluid, etc. etc. The decision to design the bore using a gravel pack has significant cost
implications. In some formations it is unavoidable.

If the aquifer sands are not well sorted, it is usually possible to development a natural gravel pack by
removing the fine grain size sands during development of the well.

2.9 Well Completion and Development

2.9.1 Introduction

After the installation of the well string into the borehole, two further stages are required. A cement grout
is placed between the formation and the casing to prevent leakage of formation water into the aquifer.
This is particularly important close to the surface. The final stage is well development.

2.9.2 Well Development

If rotary mud-flush techniques have been used, the mud cake that formed against the formation to
stabilise it during construction, now needs to be removed to let water into the bore. This can be achieved
by a number of methods:

• swabbing,

Water Resources Engineering CVEN3501 46

Lecture 2 Aquifer properties and borehole construction

• surging,
• air lifting, or
• adding chemicals

Well development is continued until a clean sand free water is produced at the surface. The bore is then
ready to be pump tested.

2.10 Field preparations for a pumping test

Management of groundwater resources requires an accurate assessment of groundwater availability

and reliable monitoring of the groundwater resource. There are typically 3 aspects considered:

• Groundwater levels,
• groundwater quality and
• abstraction volumes.

but water quality monitoring will not be covered here.

Monitoring implies regular measurement in the field either by field visits or by the installation of appro-
priate recording equipment. A monitoring network of boreholes which are dedicated and specifically
designed for monitoring is often established. The monitoring of water levels in these boreholes is rel-
atively easy and is carried out widely. A fundamental aspect of water resource management that has
frequently been omitted in the past, is the monitoring of abstraction rates. The monitoring of groundwa-
ter quality is far more time consuming, involving time at the borehole site and also significant time in the
laboratory.

2.10.1 Groundwater level monitoring

The basic measurement made in groundwater level monitoring is the distance from the top of the bore
to the water level. This measurement can be made by a number of techniques (see below) and forms
the fundamental monitoring data set for a borehole location. Change in water level is an indication of
change in storage and flux in the aquifer and can provide valuable information for management.

There are five main methods available, viz.:

• dip meter,
• float-type,
• pressure transducer,
• acoustic probes and
• air lines.

Dip meter

A dip (water level) meter is a simple contact gauge that closes a DC electrical current when the tip of the
meter is lowered into the water. A hand held meter is shown in Fig. 2.12 and lowered into a piezometer
in Fig. 2.14.

Water Resources Engineering CVEN3501 47

Lecture 2 Aquifer properties and borehole construction

Figure 2.12: Electrical dip meters Figure 2.14: Measuring the depth to water

Figure 2.13: Data loggers

Water Resources Engineering CVEN3501 48

Lecture 2 Aquifer properties and borehole construction

Figure 2.15: Schematic illustrating a typical Figure 2.16: Combination of fluid conductivity
transducer installation and water level

Float devices

The float rests on the water surface and is suspended on a steel wire over a pulley device connecting to
a pen, and then back down to a counter weight. Movement of the float moves the pen, and if the pen is
arranged to write on a chart attached to a drum, a continuous record of the water level can be made. A
clockwork timing mechanism is used to move the drum. Gears are selected to give one revolution of the
drum per day, per week or per month.

Pressure transducers

A pressure transducer is a device which is installed beneath the water surface in the piezometer and
measures the pressure of the water column above the transducer location. In clean water of a known
density, this is a direct measure of the pressure head ψ. As the water surface rises or falls, the pres-
sure (weight of water) changes and the output from the transducer changes. A schematic showing a
data logger and transducer installation is shown in Fig. 2.15 and a possible transducer combination in
Fig. 2.16.

The pressure sensitive device is a thin film of silicon with strain gauges attached. As the weight of the
water changes, the change is detected by the strain gauges and presented to the data logger as a linear
change in voltage output. It is clear that the pressure transducer must be fixed in place, otherwise the
change in the depth of the transducer will be mistakenly interpreted as a change in water level. It is also
clear that the density of the water must be known, or assumed constant. Variation in density will change
the pressure of the water column and would be interpreted as a change in the water-level measurement.

Transducers may be readily connected to data loggers and it therefore becomes relatively easy to set
up a system designed to record the water level at regular intervals. The memory, clock and power
requirements are all available in standard data loggers. It is possible for approximately $1,000 to $1,500
to purchase self contained systems for water-level measurement.

Transducers can either be sealed so that they measure absolute pressure (DIVER systems) or vented
to the surface so that they measure gauge pressure. The major advantage of the sealed systems is

Water Resources Engineering CVEN3501 49

Lecture 2 Aquifer properties and borehole construction

that the complete unit can be fixed in place below the water surface in a piezometer. A disadvantage
is that the atmospheric pressure must be recorded separately and the impact of atmospheric pressure
changes removed from the water level record. By contrast, a vented system requires that an expensive
cable must be used to vent the back face of the silicon diaphragm to atmospheric pressure. Price (2009)
has given a detailed review of the corrections required.

Acoustic probes

An acoustic probe can be used to transmit an acoustic signal from the surface to the water surface.
The water surface then reflects the signal back to the surface. Acoustic probes can then give a direct
and accurate depth-to-water measurement. The acoustic signal loses definition by scattering down the
hole so that there is a maximum depth at which the devices will work satisfactorily. Perhaps a more
practical restriction is that the bores must have been installed vertically otherwise the reflected signal
is not reflected back to the surface but is scattered against the side of the bore. Acoustic probes are
designed for bores as small as 100 mm and for depths limited to 10 m.

Air lines

Air lines are installed at a known depth beneath the water; and by measuring the pressure of air neces-
sary to discharge water from the tube as a stream of bubbles to the water surface, the height of the water
column above the discharge point can be calculated. This type of application is popular in geotechnical
studies where a permanent installation is required.

2.10.2 Manometer boards

It is possible to obtain a detailed measure of vertcal hydraulic head if a bundled piezometer has been
installed in a bore (Acworth, 2007). Fig. 2.17 shows a board set up at the edge of a creek to investigate
groundwater flow processes. The different levels can be seen on the board where food dye has been
used to colour the water inside the tubes on the board. Fig 2.18 shows the results from a manometer
board set up at David Phillips Field (UNSW Campus) to show the effects of pumping from a bore 12 m
from a bundled piezometer installation.

2.10.3 Accuracy requirements

The accuracy of the hydraulic head measurement will directly determine the accuracy of the analysis
possible. In general terms, a measurement to within 10 mm is sufficiently accurate.

It is necessary that the various measurement devices be checked for accuracy on a regular basis. Tapes
can stretch, transducers can drift and chart recorders can run out of ink!

2.11 Flow measurement

For efficient management of the groundwater resource, metered extraction of large yielding bores, such
as those for town or irrigation should be installed. Generally these meters are of the impeller type and are
installed in the discharge pipe according to the local standards. They are being progressively installed
across Australia. A section of the rising main discharge line is usually chosen to install the meter. It will
normally require a power source and to be installed in a straight line section of the discharge line. The
sensor element can be an in-line impeller or an eletromagnetic sensor.

Water Resources Engineering CVEN3501 50

Lecture 2 Aquifer properties and borehole construction

Figure 2.17: Manometer board installed above a bundled piezometer (Photo RIA)

Figure 2.18: Manometer board results for David Phillips Field

Water Resources Engineering CVEN3501 51

Lecture 2 Aquifer properties and borehole construction

Figure 2.19: Flow monitoring with a weir tank Figure 2.20: In line flow meters (2) are shown
equipped with baffle plates to stabilise the flow. at a pumping test monitoring site in Queensland
The head over the weir is measured with a trans- (Photo: CH)

ducer. (Photo RIA)

Figure 2.21: Orifice meter used to monitor dis- Figure 2.22: Orifice pipe set up on a small dis-
charge during a pumping test. The height that charge line. (Photo CH)
water rises in The manometer tube behind the
orifice is proportional to the discharge. (Photo CH)

In-line meters (Fig. 2.20) can be available during pumping tests and will be replaced by permanent
installations when the water well is commissioned. The size of the meter is determined by the discharge
rate. Measurement of discharge can be by a weir tank (Fig. 2.19) or by a simple bucket and stop watch
or by an orifice meter (Fig. 2.21).

2.11.1 Borehole design for monitoring and sampling

High production boreholes for irrigation and urban supply are being progressively metered throughout
Australia to monitor the amount of groundwater extracted; while bores used for monitoring water levels
and sampling for chemical analysis are usually dedicated for that purpose.

Fully screened wells and open (uncased) boreholes have the disadvantage that they reflect information
over a broad depth range. Samples obtained by pumping are depth-averaged. Samples recovered
from standing water in the borehole using surface activated grab samplers may not be representative
of groundwater in the aquifer, especially if the groundwater quality is stratified. Such samples would be
of lesser value in detailed studies of aquifer behaviour, but would be representative of water abstracted
for supply or for some other beneficial use, or for providing initial broad indications of contamination. It
is possible to use inflatable packers to isolate intervals within open or fully screened bores to provide
more detail on vertical groundwater quality variations. The aim of the packers is to restrict vertical water

Water Resources Engineering CVEN3501 52

Lecture 2 Aquifer properties and borehole construction

Figure 2.23: Preparation of a bundle of mini- Figure 2.24: Installation of the bundle (Photo RIA)
samplers prior to installation inside a hole cased
using a cable-tool rig (Photo RIA)

movement within the borehole casing, and to recover groundwater from that part of the aquifer adjacent
to the packered interval.

It is of course preferable to monitor water level or sample from a narrow zone, especially for detailed
investigations.

Multiport installations and piezometer bundles

Multiport installations or small diameter piezometer bundles provide numerous point measurements
(Fig. 2.23 and Fig. 2.24). The access tube to the inlet is also often of reduced size, minimising the
need for purging. These installations can provide groundwater quality profile data at a very fine-scale,
and are especially useful where stratification or large concentration gradients are expected. Often these
devices use gas-lift or suction (peristaltic pump) to remove water present in access lines and to recover
a sample.

2.12 Pumping Tests

2.12.1 Purpose of Pumping Tests

A pumping test, or pumped well test, may serve two objectives.

Well Test - To provide information about the yield and drawdown of the well. These data can be used
to determine the specific capacity of the well which is a measure of hydraulic efficiency. This type
of test is called a well test, as it is the well rather than the aquifer which is being tested.

Water Resources Engineering CVEN3501 53

Lecture 2 Aquifer properties and borehole construction

Aquifer Test - To provide information from which the hydraulic properties of the aquifer can be deter-
mined. Properly planned and carefully conducted, the aquifer test may provide basic information
sufficient to attempt an inverse solution to the mathematical equation describing groundwater flow.

It would be common practice to carry out these two steps sequentially during the development of a
new abstraction well site. Accurate information on the specific capacity is required to enable a full
specification for a permanent pump. For this reason, much of the early testing is carried out with a test
pump installed in the well. This pump will have a wide range of head-discharge characteristics to allow
it to be to be used in the unknown conditions.

Well tests generally can be completed in a day and, in contrast to aquifer tests, do not require detailed
planning.

An Australian Standard (AS 2368-1990) has been prepared by Standards Australia (1990) for Test Pump-
ing of Water Wells.

2.12.2 Drawdown Measurement

The change in hydraulic head as a result of pumping is referred to as the drawdown.

The static water level is recorded in the pumped borehole and any observation points before the start of
the test and immediately before any change in the rate of pumping. Water level changes vary logarith-
mically with time, therefore measurements of water levels are made with gradually increasing periods
between observations. The Australian Standard recommends the following schedule of measurements:

1. Immediately before the change in discharge occurs,

2. At 0.5, 1, 2, 3, 4, 5, 6, 8, 10, 15, 20, 25, 30, 45, 60, 75, 90 minutes,
3. Every 30 minutes until 8 hours have elapsed
4. Every 60 minutes until 24 hours have elapsed
5. Every 120 minutes until 48 hours have elapsed
6. Every 4 hours until the end of the test.

Figure 2.25 shows a combination of data logging devices incorporated into a pumping test data logger.
Equipment like this can very dramatically reduce the labour associated with running a pumping test, and
ease the data interpretation by recording all the data digitally.

2.12.3 Pre-Test Planning for an Aquifer Test

An aquifer pumping test is expensive and the number of tests carried out are normally strictly limited. All
available information concerning the aquifer, the boreholes and monitoring points should be established
as part of the pre-test planning exercise.

The pre-test study should enable identification of the aquifer type (confined or unconfined), an approx-
imation of the transmissivity, and an identification of the other major fluxes in the aquifer. It is partic-
ularly important to identify the pumping regime of other abstractors within the zone of influence of the
proposed test. The aquifer test characteristics will vary depending upon the available information. A
different testing programme is designed for an aquifer about which little if anything is known, compared
to a supplementary test in a well defined aquifer.

A location for discharge of the aquifer test water is required. It is important that this be some distance
away from the test site to avoid recirculation of the test water. This is not such a problem with a confined

Water Resources Engineering CVEN3501 54

Lecture 2 Aquifer properties and borehole construction

Figure 2.25: Schematic of a pumping test data logger

aquifer but is very important in unconfined aquifer tests. It may be necessary to pump the water several
hundred metres away. If the water is to be added to an existing stream then flow gauging in the stream
is necessary before the test begins and again during the test.

Careful planning is required for the collection of the aquifer test data. Data collection is required very
frequently initially, and this requires two people if a dip meter is used. Personnel requirements can also
be considerable, especially when it is realised that the test measurements are required hourly throughout
the night.

2.12.4 Performance of a pumping test

The measurements to be taken during a pumping test include the following:

• the establishment of a schedule of monitoring points and measurement of the water levels at these
points prior to testing and during testing,

• measurement of atmospheric pressure,

• measurement of rainfall during the test and recovery period
• measurement of the discharge rate,

• measurement of chemical quality of discharge water.

Ideally, an aquifer test should not be started before the already existing water level changes in the
aquifer are known, including both long-term regional trends and short-term variations of the water level.
Therefore, monitoring of the water levels in the available observation points should be carried out for
several days before the test commences. A level at a monitoring point remote from the test should also
be included in the monitoring programme to enable extraneous factors to be identified.

Unless all the data is to be collected using automated equipment, the preparation of measurement forms
are essential prior to testing. The Australian Standard Standards Australia (1990) contains proforma for
testing.

Water Resources Engineering CVEN3501 55

Lecture 3

Pumping test analysis and occurrence

of groundwater

3.1 Mathematical model of flow to a well

The groundwater flow equation represents a mathematical model of the groundwater flow process. For
certain simple situations, the mathematical model, as represented by the equations, can be solved by
analytical methods. The Theis method is an example of an analytical solution to the transient flow
equation (The Diffusion Equation) in two dimensions.

The equation for non-steady state flow of groundwater through a homogeneous medium can be written
using the ∇ representation for the partial derivatives, and dividing through by the hydraulic conductivity
K, gives:
Ss ∂h
∇2 h = (3.1)
K ∂t

This equation is known as the diffusion equation. The solution h(x, y, z, t) describes the value of the
hydraulic head at any point in a flow field at any time. This equation can not be solved using analytical
methods. A full numerical analysis is required using either finite differences or finite elements.

(Theis, 1935) produced an analytical solution to the equation based on a two-dimensional representa-
tion. An essential step in this approach is to reduce the dimensionality from three dimensions to two
dimensions. This is achieved by assuming that:

• that the thickness of the confined aquifer is constant,

• that the aquifer is of infinite extent, and
• that there are no vertical flow components in the aquifer which therefore makes the derivative of
the head with respect to vertical distance (z), equal to zero.

Then, multiplying the top and bottom of the right hand side of Equation 3.1 by the aquifer thickness (b)
gives the transient flow equation in two dimensions as:
∂2h ∂2h S ∂h
+ 2 = (3.2)
∂x2 ∂y T ∂t

where:

T is the aquifer transmissivity - the product of the hydraulic conductivity and the aquifer thickness,
S is the aquifer storage coefficient - the product of the specific storage and the aquifer thickness.

56
Lecture 3 Pumping test analysis and occurrence of groundwater

Figure 3.1: Radial Flow to a Well

3.2 Analytical solution to the radial flow equation

Theis (1935) gave an analytical solution to this equation for radial flow to a well.

It is clear that the flow to a well in a homogeneous aquifer will be radially symmetric and it is therefore
convenient to convert Equation 3.2 to radial co-ordinates, viz.:
∂ 2 h 1 ∂h S ∂h
+ = (3.3)
∂r2 r ∂r T ∂t

where:

p
r is the radial distance from the well r = x2 + y 2 .

The mathematical region of flow is a one-dimensional line through the aquifer, from r = 0 at the well to
r = ∞ at the boundary. A schematic of the flow regime is shown in Figure 3.1.

The initial condition is that

hr,0 = h0 (3.4)
for all of radial distances, and where h0 is the initial water level, or standing water level (SWL).

The boundary conditions assume no drawdown in hydraulic head at the infinite boundary

h∞,t = h0 (3.5)

for all times (t). A constant pumping rate Q at the well is also assumed:

The solution hr,t describes the hydraulic head at any radial distance r at any time t after the start of
pumping.

Theis utilised the heat flow analogy to arrive at an analytical solution to this equation subject to the initial
and boundary conditions presented. His solution written in terms of the drawdown, is
Z ∞ −u
Q e
s = h0 − hr,t = du (3.6)
4πT u u

Water Resources Engineering CVEN3501 57

Lecture 3 Pumping test analysis and occurrence of groundwater

where:
r2 S
u= (3.7)
4T t
The integral in the equation is the exponential integral and is well known in mathematics. Tables of its
values are available or can be readily calculated using a programmable calculator and the approximation
shown below.
u2 u3 u4
W (u) = −0.577216 − ln u + u − + − + ... (3.8)
2 × 2! 3 × 3! 4 × 4!

Substituting this function into Equation 3.6 yields:

Q
s= W (u) (3.9)
4πT

A plot of w(u) vs u or W(u) vs 1/u is commonly called a Theis Curve.

If the aquifer transmissivity (T ) and storativity (S) and the pumping rate (Q) are known, it is possible
to predict the drawdown at any distance (r) from a well at any time (t) after the start of pumping. All
that is required is to calculate the appropriate value of (u), use tables to get a corresponding value of
the well function W (u) and calculate the value of the drawdown. Similar relationships can be calculated
for observation points at different radial distances from the pump well. This is a very commonly used
method to estimate drawdown and is the base of workshop question 3.1

Example: A farmer on the River Murrumbidgee flood p lain i s d eveloping fl ood ir rigation of ri ce on
his farm. He has an existing bore that he pumps 250 l/s from a sand and gravel aquifer beneath his
property. The aquifer lies at 80 m depth and is 20 m thick. The hydraulic conductivity (K) of the sand
is 50 m/day and the specific s torage ( Ss ) o f t he a quifer i s 0 .00015. W hat w ill b e t he d rawdown i n the
aquifer 500 m away from the abstraction well after pumping continuously for 100 days?

Firstly, calculate all the quantities in terms of common units. You are given abstraction in l/s and a
hydraulic conductivity in m/day. It is common practice to use units of m/day in groundwater calculations
of this type - so:
The transmissivity T ( T = K times aquifer thickness) will be 50 × 20 = 1000 m2/day
The storage coeficient S (Ss times aquifer thickness) will be 0.00015 × 20 = 0.003
The pumping rate in m3 /day will be 250 × 24 × 60 × 60/1000 = 21, 600
500∗500∗0.003
Substituting in Equation 3.7 for a value of u gives u = 4∗1000∗100 = 0.001875
A value of the well function W(u) can be derived from Equation 3.8 using just the first 3 terms:
W (u) = −0.577216 + 6.279145 + 0.001875 = 5.703804.

The drawdown is then calculated from Equation 3.9

21,600
s = 4×3.141593×1000 5.703804 = 9.804 m

Note that drawdown and pumping rate are related linearly: double the pumping rate, you double the
drawdown.

3.3 Pumping test interpretation using graphical methods

The interpretation of pumping test data is an example of a graphical solution to the inverse problem of
radial flow to a well. Instead of using values of T and S in the Theis method to solve the radial flow
equation in the forward direction to produce values of drawdown at various times and distances; we use
measurements of drawdown at various distances and times from an abstraction borehole to solve the
equation in the reverse direction to derive values of T and S.

Theis devised a superposition solution to this problem using type curves. Taking logs to the base 10 of

Water Resources Engineering CVEN3501 58

Lecture 3 Pumping test analysis and occurrence of groundwater

Equation 3.9, we get:

Q
log s = log + log W (u) (3.10)
4πT
and taking logs to the base 10 for Equation 3.7
r2 4T
log = log + log u (3.11)
t S

r2
Hence, for a constant pumping rate Q, s is related to t in the same manner as W (u) is related to u.

This equivalence allows a graphical procedure to be adopted for the solution of transmissivity and stora-
tivity from pumping test data.

The steps of the graphical procedure are as follows:

1
• Plot the function W (u) against u on log-log paper. This curve is known as a type curve.
t
• Plot s vs t or r2 on a second sheet of log-log paper of the same size and scale as the W(u) vs 1/u
plot.
• Superimpose the field curve on the type curve, keeping the principal axes parallel. Adjust the
curves until most of the observed data points fall on the type curve.
t
• Select an arbitrary match point and read of paired values W (u), 1/u, s and t or r2 at the match
point.
• Calculate values of transmissivity and storativity from the equations above using the values of u,
W (u), t and s derived from the match point.

This is workshop question 3.4.

3.3.1 Jacob Approximation - straight line solution methods for pumping test
interpretation

For values of u less than 0.01, all but the first two terms in the series expansion for the well function
W (u) can be neglected. In the example shown above, we used the first 3 terms - but you can see how
small the contribution of third term was.
The Theis equation can then be approximated without recourse to the well function and tables as:
r2 S
 
Q
s= −0.5772 − ln (3.12)
4πT 4T t

Noting that ln u = 2.30 log u

and − ln u = ln u1

and that ln 1.78 = 0.5772

the equation can be rewritten as

2.30 Q 2.25 T t
s= log (3.13)
4πT S r2

Since Q, r, T and S are constants it is clear that s vs log t should plot as a straight line as long as u is
less than 0.01.

This method can be used to interpret aquifer test observations made at a single observation point for the
duration of the test. A plot of drawdown versus the logarithm of time should yield a straight line if there
are no impacts from boundary conditions or aquifer inhomogeneity.

Water Resources Engineering CVEN3501 59

Lecture 3 Pumping test analysis and occurrence of groundwater

Figure 3.2: Composite Pumping Cones

At time t1 , the drawdown s1 is expressed from Equation 3.13:

2.30 Q 2.25 T t1
s1 = log (3.14)
4πT S r2
At time t2 , the drawdown s2 will be:
2.30 Q 2.25 T t2
s2 = log (3.15)
4πT S r2

It follows that
2.30Q t2
s2 − s1 = log (3.16)
4πT t1
If t1 and t2 are selected one log cycle apart, then log(t2 /t1 ) = 1. Then Equation 3.16 becomes:

2.30Q
∆s = (3.17)
4πT
where ∆s is the drawdown for one log cycle of time. Equation 3.17 can be used to solve for T .
This is question 3.3 in the workshop.

The storativity can be obtained by setting the drawdown to zero in Equation 3.13, such that:

2.30Q 2.25T t
s=0= log (3.18)
4πT Sr2
2.25T t0
This can only occur if Sr 2 = 1. Therefore:

2.25T t0
S= (3.19)
r2

where:

t0 is the intercept of the straight line at zero drawdown, i.e., the time immediately prior to drawdown in
the observation well.

3.3.2 Multiple-Well Systems

The drawdown in hydraulic head at any point in a confined aquifer in which more than one well is pumping
is equal to the sum of the drawdown that would arise if each of the wells were pumped independently.
Figure 3.2 shows the effect of summing the contributions from different drawdown cones.

Water Resources Engineering CVEN3501 60

Lecture 3 Pumping test analysis and occurrence of groundwater

For any system of n wells pumping at rates Q1 + Q2 + . . . + Qn , the arithmetic summation of the Theis
solutions can be used to derive the drawdown at a point.
Q1 Q2 Qn
s= W (u1 ) + W (u2 ) + . . . + W (un ) (3.20)
4πT 4πT 4πT
where

ri2 S
ui is 4 T ti

i is 1, 2, . . . , n
ti is the time since pumping commenced at well i, and
Qi is the pumping rate at well i.

3.3.3 Boundary Conditions

Significant vertical and lateral variations in hydraulic conductivity exist in real aquifers. They must at
some point be terminated by geological changes which form hydrogeological barriers. It is possible
to use the Theis method to approximate the drawdown in boreholes pumping from bounded aquifers.
The cone of depression caused by pumping will extend through the aquifer until the total abstraction is
balanced by recharge or until the boundary of the aquifer is reached. There are two fundamental types
of boundary condition:

• A barrier or no flow boundary, or

• a fixed head or recharge boundary.

The effect of these two boundary condition types on a plot of drawdown against log time are shown in
Figure 3.3 and Figure 3.4.

Recharge Boundaries

The ideal recharge boundary occurs when the cone of depression caused by pumping from a confined
aquifer extends up to a fully penetrating river. The river is considered to provide a constant head bound-
ary as shown in Figure 3.5.

The hydraulic situation can be modelled by using the concept of an imaginary or image well. The
imaginary well is located an equal distance on the far side of the river boundary to the distance from the
real well to the river boundary. The effect of the river on the drawdown in the real well can be calculated
using Theis theory by assuming that the image well is a recharging well and summing the drawdown
components from the real well and the image well as shown in Equation 3.21.
Q Q
s= W (uw ) − W (ui ) (3.21)
4πT 4πT
where

2
rw S
uw is 4T t

ri2 S
ui is 4T t

rw is the distance from the real well to the observation point,

ri is the distance from the imaginary well to the observation point,
t is the time since pumping commenced at the well, and
Q is the pumping rate at the well

Water Resources Engineering CVEN3501 61

Lecture 3 Pumping test analysis and occurrence of groundwater

Figure 3.3: Change in drawdown due to a barrier boundary

Figure 3.4: Change in drawdown due to a recharge boundary

Water Resources Engineering CVEN3501 62

Lecture 3 Pumping test analysis and occurrence of groundwater

Figure 3.5: Recharge Boundary Conditions

Water Resources Engineering CVEN3501 63

Lecture 3 Pumping test analysis and occurrence of groundwater

Barrier Boundaries

Barrier boundaries are perhaps more common than recharge boundaries. The general geological model
and hydraulic model are shown in Figure 3.6.

The approach for analysis is similar to the recharge barrier case except that the image well is considered
to be an imaginary pumping well and therefore the drawdown components add together as show in
Equation 3.22.
Q Q
s= W (uw ) + W (ui ) (3.22)
4πT 4πT
where

2
rw S
uw is 4T t

ri2 S
ui is 4T t

rw is the distance from the real well to the observation point,

ri is the distance from the imaginary well to the observation point,
t is the time since pumping commenced at the well, and

Q is the pumping rate at the well

3.3.4 Summary

It is convenient to summarise the assumptions implicit in the Theis solution to the radial flow equation,
viz.:

1. The aquifer has an infinite areal extent

2. The Aquifer is homogeneous, isotropic and of uniform thickness over the area influenced by the
pumping test.
3. Prior to pumping, the piezometric surface and/or water table surface are horizontal over the area
influenced by the pumping test
4. The aquifer is pumped at a constant discharge rate

5. The pumped well penetrates the entire aquifer and thus receives water from the entire thickness
of the aquifer by horizontal flow
6. The aquifer is confined
7. The flow to the well is in unsteady state, i.e. the drawdown differences with time are not negligible
nor is the hydraulic gradient constant with time

8. The water removed from storage is discharged instantaneously with decline of head
9. The diameter of the pumped well is very small, i.e. the storage in the well can be neglected
10. Water passes from the aquifer through the well with no loss of head.

The large number of assumptions in the Theis analytical method well demonstrates the difficulty of
developing methods for more complicated conditions using analytical methods. However, this has not
stopped the publication of numerous papers in the literature dealing with partial penetration, leakage,
unconfined flow, large diameter wells, etc. and the use of many different sets of type curves.

Water Resources Engineering CVEN3501 64

Lecture 3 Pumping test analysis and occurrence of groundwater

Figure 3.6: Barrier Boundary Conditions

Water Resources Engineering CVEN3501 65

Lecture 3 Pumping test analysis and occurrence of groundwater

3.4 Groundwater in Australia

3.4.1 Rock types

Geology is the science and study of the physical matter that constitutes the Earth. The field of geology
encompasses the study of the composition, structure, properties, and history of the planet’s physical
material, and the processes by which it is formed, moved, and changed. There are 3 basic types of rock
recognised by geologists:

• igneous
• metamorphic
• sedimentary

Igneous rocks

Igneous rocks are formed from molten material that either cools rapidly from the surface eruption of a
volcano to produce a fine-grained basalt; or cools slowly in a magma chamber beneath the earth surface
to form a coarse grained granite rock. The colour of the rock is a function of the chemistry of the magma.
If there is free silica (quartz) then a granite is formed, otherwise a gabbro is formed.

Metamorphic rocks

Metamorphic rocks are rocks that have been heated but not melted to form a magma. There has been
some realignment of crystals and the rock commonly has some foliation but the original structures can
often still be seen.

Sedimentary rocks

Sedimentary rocks are comprised of the products of weathering derived from igneous and metamorphic
rocks. These rocks can be formed from sand and gravel, or from muds and clays or deposited from sea
water as in an evaporite or limestone. Sedimentary rocks may either be unconsolidated as found at the
beach or in flood-plain deposits, or consolidated as found in sandstones or mudstones. Unconsolidated
deposits have the highest porosity and generally contain the most groundwater. Consolidated deposits
have had all of the initial water that filled the pore space during deposition, squeezed out. During his
process there will have been realignment of the sedimentary particles and the deposition of additional
mineral phases derived from percolating groundwater that will further reduce the porosity of the unit.
Simply stated, an unconsolidated sand with high porosity is turned into a sandstone that can be cut with
a rock saw.

3.4.2 Factors affecting groundwater occurrence

There are many factors affecting the occurrence of groundwater. These include the following:

• rainfall
• runoff
• topography

Water Resources Engineering CVEN3501 66

Lecture 3 Pumping test analysis and occurrence of groundwater

• lithology
• structure
• vegetation and
• past climate

Groundwater is held in the pore or fracture spaces in a sediment or rock. Groundwater is recharged from
the surface where there is frequently a surplus of rainfall over evapotranspiration. The surplus percolates
downwards below the soil profile. Some is used by plants during transpiration, some becomes interflow
or bank storage and returns to the surface water system before reaching the groundwater storage and
some finds its way to the aquifers and becomes groundwater recharge. Heavy rainfall may saturate the
soil and fall at such a rate that the soil can not absorb the water. Run off then occurs and water moves
directly to surface drainage lines. Clearly, the amount of vegetation, the topography and the rainfall
intensity all impact on this complex set of processes.

The porosity of the rocks and the type of rock (lithology) will impact on the quantity of water that is stored
as groundwater.

Groundwater is often of significant age. It is not unusual to find groundwater in Australia that is 10,000 to
30,000 years old. In fact, groundwater moving through the Great Artesian Basin takes 1.25 million years
between recharge and discharge. The climate has changed dramatically over these time frames and it
is quite possible that the water in the aquifer is old. Developing this groundwater supply is then akin to
mining the water.

Groundwater is frequently of a very good quality - fit to drink; although, salty groundwater can also occur
and even hypersaline groundwater is frequently found in Australia. This is often the result of low rainfall
and very slow groundwater discharge into a flat locality where the water can not easily escape to the
ocean.

The occurrence of groundwater in Australia can be sub-divided as shown in Figure 3.7 produced by
Geoscience Australia (http://www.ga.gov.au/) (GA). GA have subdivided the continent into 5 aquifer
types:

1. porous, extensive highly productive aquifers

2. porous, extensive aquifers of low to moderate productivity
3. fractured or fissured, extensive highly productive aquifers
4. fractured or fissured, extensive aquifers of low to moderate productivity
5. local aquifers of generally low productivity

The classifications developed by GA are based on how groundwater is found in the various rock types.
There is some overlap here between the hydrogeological and geological classifications. For example,
both a sandstone an a basalt can be predominantly fractured with groundwater only recovered from the
fractures. However, they are completely different rock types and the geologist would never mix the two
together. It may also be easier to group together the groundwaters held in a particular sedimentary
basin rather than describe each aquifer type in that basin separately. There is no simple answer to this
classification problem.

3.5 Groundwater in sedimentary basins

Groundwater occurs in the primary and secondary (fractures) porosity of rocks deposited in large sedi-
mentary basins. The dark blue areas in Fig. 3.7 represent these basins, where the Great Artesian Basin
is clearly the largest. A brief description of three important basins is given below.

Water Resources Engineering CVEN3501 67

Lecture 3 Pumping test analysis and occurrence of groundwater

Figure 3.7: Aquifer types in Australia

3.5.1 The Paleozoic and Mesozoic sediments of the Great Artesian Basin

The GAB is one of the largest groundwater basins in the world. The GAB spans across three states and
the Northern Territory over an area of 1.7 million km2 , (20% of the Australian landmass), and is formed
from middle Triassic to Cretaceous aquifers and associated confining beds. Recharge to the aquifers
takes place on the eastern margins of the basin in outcrop areas, and environmental isotope studies
generally support the contention that this takes place today. Discharge from the basin takes place as
concentrated outflow from springs, from vertical leakage towards the regional groundwater table, inter
basin leakage, and from artificial discharge from artesian and pumped wells (Habermehl, 1983). The
water is predominantly fresh and in most areas is under sufficient pressure that flowing artesian bores
can be constructed. The first artesian bore was constructed at Bourke in 1878.

Approximately 4,770 flowing artesian bores have been constructed to average depths of 500 m up to
depths of 2,000 m into the lower Jurassic aquifer. These bores were often left to flow into bore drains at
the surface and water was used to feed cattle. Unfortunately, the vast majority of this water evaporated
and the overall pressure in the aquifer declined. Today there are about 3,000 free flowing bores, 35,000
sub-artesian bores and more than 600 spring complexes around the margin of the Basin.

The use of bore drains to distribute water and to provide water for stock, while being effective, is very
inefficient. It has been estimated that only 5-10% of the water distributed in this way is actually used;
the rest is lost by evaporation and seepage. An ongoing programme aimed at capping and rehabilitation
of flowing bores was commenced in 1988 and will reduce abstractions by an order of magnitude and
allow more eficient usage of the resource. This programme is funded jointly by the Federal and State
governments and by landowner contribution. The programme is now (2010) more than 50% complete.

There are also 20,000 non-flowing bores, mainly in the higher aquifers, extending only up to several
hundred metres deep. Most shallow bores are found in the central parts of the basin, where the artesian
systems are too deep for economic abstraction. While there are some very deep bores in the central
part, deeper bores are generally occur on the margins of the basin.

Springs discharging around the margin of the Basin (Fig. 3.8) have supported Aboriginal occupation for
thousands of years prior to European settlement.

Water Resources Engineering CVEN3501 68

Lecture 3 Pumping test analysis and occurrence of groundwater

Figure 3.8: The Great Artesian Basin

The GAB is predominantly a confined and closed groundwater basin comprising a complex multi-layered
system separated by largely impermeable units. The aquifers are largely continuous and extend across
the basin as shown schematically in Figure 3.9. The aquifers occur at depths up to 3000 m.

Major issues for the management of the GAB are described by the Great Artesian Basin Consultative
Council (1998) as:

• groundwater pressures have declined in a number of regions in the Basin and 1500 artesian bores
have ceased to flow. Spring flows have declined significantly and some springs have dried up.
• up to 90% of artesian flow is wasted due to the use of open bore drains to distribute water across
properties and the large number of uncontrolled bores. There are about 27 000 km of bore drains
in Queensland, 6 000 km in NSW and about 1 000 km in South Australia. The longest individual
drains exceed 150 km.
• distribution of artesian water by bore drains has negatively impacted on land sustainability and
conservation of biodiversity.
• there is a potential for groundwater contamination with the advent of more intensive farming in the
intake beds.

the discharge point of groundwater flowing in the GAB is frequently a mound spring as seen in Fig. 3.10.

3.5.2 The Tertiary Murray Basin

The Murray Basin occupies an area of about 280,000 km2 , and is composed of Tertiary age sediments
which are flat-lying but increase in thickness towards the centre of the basin. The maximum thickness

Water Resources Engineering CVEN3501 69

Lecture 3 Pumping test analysis and occurrence of groundwater

Figure 3.9: Conceptual model of the multi-layered aquifers in the GAB

Figure 3.10: Mound spring discharge from the GAB. Note the palaeo discharge mound in the background

Figure 3.11: Capped & piped artesian bore at Rockwell station, SE Queensland. The property name
reflects a large well excavated in rock by Aboriginals (though not artesian)

Water Resources Engineering CVEN3501 70

Lecture 3 Pumping test analysis and occurrence of groundwater

Figure 3.12: West to East cross-section through the Murray Basin

is around 640 m. The Basin underlies parts on New South Wales, Victoria and South Australia. A
generalised map and cross section of the basin is shown in Fig. 3.12. The basin contains a number of
aquifers which contain groundwater with quality from good to highly saline, and with varying bore yields.
An important limestone aquifer underlies an area of around 30,000 km2 , which straddles the SA/Victoria
border region. This unit, 50-150 m thick at depths of up to 100 m below surface, provides good quality
water for town, industrial, irrigation and stock supplies. In the Mallee district, silts and clays are found
along with highly saline water.

In the Murray Basin, groundwater generally is under-utilised in all three states. Presumably this is partly
due to the availability of cheaper surface water resources for irrigation in this important agricultural area.
The costs of surface water (in 2009) are reportedly $10-$14 /ML, and for groundwater $15-$30/ML.
Additionally there are capital costs for drilling of up to $25,000 for shallow wells and up to $75,000 for
deeper production wells. However, it is still economic to use groundwater costing $50/ML, although
this is highly dependent on the type of crop. Also, given recent policy changes relating to water pricing
and capping surface water allocations, it seems likely that costs of surface water will rise and combined
use of surface water and groundwater will become increasingly common, given the availability of the
groundwater.

Despite the apparent under-utilisation of the groundwater resource, there is considerable concern that
in some areas local over pumping could take place, giving rise to degradation of the resource, and
this is already occurring. For example, the Angas-Bremer irrigation area on the western flank of the
basin abstracts 25 × 106 m3 /y of groundwater, which is causing ingress of saline groundwater to the
aquifer (Jacobson et al., 1991).

Recent changes indicate that long-term ”mining” of groundwater (abstraction above sustainable levels)
can no longer be tolerated. However, controlled depletion (short-term, reversible lowering of groundwater
levels) is encouraged, and monitoring of groundwater quality is being increasingly used to make sure
that resources are not being degraded (for example by salinity increases) beyond the beneficial use of
the water (ARMCANZ/ANZECC, 1995).

A significant problem has arisen in the Murray Basin from rising groundwater levels which result from
land-use changes over the last 100-150 years, particularly in the southern part of the basin. These

Water Resources Engineering CVEN3501 71

Lecture 3 Pumping test analysis and occurrence of groundwater

Figure 3.13: Early Miocene geomorphology

changes have thrown the hydrological system out of balance. They have greatly increased groundwater
recharge and have caused water logging of the surface in most irrigation areas on the Riverine Plain in
the southern part of the basin. Older irrigation areas (Kerang/Cohuna, Murrumbidgee IA) are reported
to have stable ”groundwater mounds” while newer areas have continually rising levels. Recent reports
suggest that dryland as well as irrigated agriculture has given rise to significant salinisation within the
basin. An estimated 75% of salt loads in streams within the basin are reported to have come from
dryland catchments (Prime Minister’s Science, Engineering and Innovation Council, 1999). There is
considerable debate as to the most appropriate way to manage salinisation within the basin, but it is
clear that this is highly complex and there is no one panacea solution. The complexity and extent of
the salinity problem was thought to require a response at government level during the 1990’s, however,
falling water levels in fractured rock aquifers in the first decade of the 21st century has put this analysis
in some doubt. On the Riverine Plain in Victoria and some alluvial valley systems in New South Wales
(Namoi) there is a double problem with rising water levels in shallow aquifers and rapidly declining levels
in deeper aquifers.

The geomorphology of the Murray Basin during the Early Miocene is shown in Fig. 3.13. The evolution
of the geomorphology in Quarternary time is shown in Fig 3.14.

3.5.3 The Late Paleozoic to Mesozoic Perth Basin

The onshore Perth basin extends from Augusta to just northeast of Geraldton, a distance of 1,000 km,
covering an area of 51,500 km2 . The basin consists of sedimentary rocks of Silurian to early Tertiary
age, overlain by a veneer of Quaternary surficial sediments. Most sediments are Permian-Cretaceous
age, and consist mainly of inter bedded sandstone and shale, sandstone and coal measures. The major
aquifers in the basin are Cretaceous and Jurassic sandstones.

The Perth Basin has been extensively investigated to identify water resources and also for oil and gas
exploration. Fifteen groundwater sources have been identified within the basin, from a variety of aquifers,
including the surficial formation, the Cretaceous Leederville Formation and the Jurassic Yarragadee and

Water Resources Engineering CVEN3501 72

Lecture 3 Pumping test analysis and occurrence of groundwater

Figure 3.14: Quarternary geomorphology of the Murray Basin

Cockleshell Gully Formations. The mostly confined non-surficial systems are recharged either in outcrop
areas where they occur, or by downward leakage from surficial aquifers. It has been estimated that the
volume of groundwater in storage in these systems is very large compared with that renewed each year
through recharge, and that recharge has varied considerably in the last few thousand years. Given the
strategic position of these reserves close to Perth and the numerous other cities and townships along the
Western Australian coast, this has generated considerable debate as to whether such large and valuable
resources should be utilised through controlled and reversible depletion (as in parts of the Murray Basin)
to allow further development of the area (Allen et al., 1992).

3.6 Groundwater in fractured rocks

The distribution of fractured rock provinces in Australia is given in Fig. 3.15. Groundwater in this category
includes water found in the weathered profile of granites and migmatites and the water found in individual
fractures in consolidated sedimentary rocks. An estimated one third of all bores are in fractured rock
systems in Australia, these being important for stock watering and other uses. These bores supply an
estimated 310 × 106 m3 /yr of water, equating to 10% of total abstractions. Areas of fractured rocks are
shown in Fig. 3.15

Generally, individual yields from bores drilled into fractured rocks are much less than those from bores
drilled into consolidated or unconsolidated sedimentary rocks. Also, as a result of their lower porosity,
they generally exhibit a more rapid response to stress (recharge or discharge) and have large seasonal
variation in water levels. The success of a bore drilled into a fracture rock domain is dependent upon the
bore intersecting a water containing fracture.

3.6.1 Groundwater in weathered granites

Groundwater occurs in weathered granites as shown schematically in Fig. 3.16. The water first enters
the rocks along stress-relief fractures or into fractures associated with faults. The rock, which is initially
impermeable, becomes more so as chemical weathering acts upon the minerals present. The end result
is that localised patches of more permeable and porous material develop that can supply yields of up to
5 L/s for several hours. The result of weathering that normally follows fracturing in a granite or gneiss

Water Resources Engineering CVEN3501 73

Lecture 3 Pumping test analysis and occurrence of groundwater

Figure 3.15: Areas of fractured rock and porous aquifer

(high grade metamorphic rock) are shown in Fig. 3.16. The different grades of weathering (Acworth,
1987, 2001) give rise to different porosities.

Recharge is usually local and seasonal fluctuation in water levels is often large, both of which reflect the
heterogeneity of this type of aquifer. Well yields from these are often low (a few L/sec) except in heavily
fractured zones. Groundwater quality is generally good in areas where there is high rainfall, but often
poor in arid areas.

3.6.2 Fractured sandstones

Fractured rock aquifers present unique problems in their investigation, evaluation and management,
largely because of their heterogeneous nature, and the dependence of aquifer properties on fracture
distribution, connectivity and extent.

Fractured rock aquifers are characterised by high spatial variability in hydraulic conductivity, making tradi-
tional hydraulic methods for estimating groundwater flow difficult to apply. The possible range of fractures
and fracture sets in rock are shown in Fig. 3.17. Note that the orientation and degree of interconnection
of fractures is critical in the development of effective pathways through which groundwater can move.
(Cook, 2003) has described the appropriate approach for investigation of fractured rock aquifers.

In Australia, examples of important areas where groundwater extraction from fractured rocks for irrigation
has the potential to stress the groundwater resource include:

• The Clare Valley, South Australia, where expansion of the viticulture industry is increasing demand
at a rapid rate;
• groundwater extraction from the Howard River Basin for Darwin’s water supply may be stressing
phreatophytic vegetation in the area, including rain forests and tourist springs.
• in New South Wales, approximately 30 towns in the New England region and central and southern
tablelands are partially dependent on water supplies from fractured rock aquifers.

Water Resources Engineering CVEN3501 74

Lecture 3 Pumping test analysis and occurrence of groundwater

Figure 3.16: Isometric diagram showing disposition of weathering grades developed in granite

• The Mereenie Sandstone in Central Australia provides water from fractures. The well known water
holes in the McDonald Ranges north of Alice Springs, and the water supply for Alice Springs, come
from this aquifer.

On a smaller scale, fracture patterns are less predictable. In the Clare Valley, groundwater availability
appears to be controlled by closely spaced (<20 m), small aperture fractures which have no obvious sur-
face expression. On a larger scale, it may be possible to improve the success rate of water-well drilling,
but bore siting remains problematic in areas where there are no obvious geological structures. Records
of unsuccessful bores (dry wells or low-yield wells) are not readily available, making reliable statistics
of success rates in different environments impossible to determine. However, in the Archean Shield of
Western Australia, approximately 2600 bores were drilled in 13 districts as part of a Government drought
relief program during 1969-70. The percentage of successful bores (where yield was sufficiently high for
the intended use) ranged from 23.5% in one district to 1% (AWRC, 1975).

Well interference is a commonly reported problem where there is a concentration of groundwater users
within a small region. The low and variable porosity and hydraulic conductivity of fractured rocks means
that drawdown can be very large, even for a low extraction rate. There are numerous reports of extraction
from deep irrigation wells drying up shallower domestic supplies (e.g. Dandenong Ranges, Victoria;
Clare Valley and Mount Lofty Ranges, South Australia). Because of the low storage of many fractured
rock aquifers, seasonal variations in the water table level tend to be very large, even though recharge
may be small. Long-term monitoring may be required to distinguish the effects of well interferences
from natural variations of the water table. Where there is a strong, preferred orientation of fractures, the
draw down cone can be highly anisotropic. Well field design in fractured rock aquifers will thus need to
consider fracture orientations.

In Central Victoria there is fracture flow of cold CO2 -bearing mineral water, principally in the Daylesford-
Hepborn area in Lower Palaeozoic sediments of the Great Dividing Range. These waters effervesce
naturally and are sought for drinking and bathing (Weaver et al., 2006; Shugg and Knight, 1994) as
shown in Fig. 3.18. Similar conditions occur at several locations in NSW with a bore at Ballymore shown
in Fig. 3.19.

Groundwater quality issues are also different in fractured rock aquifers from those in permeable, inter
granular flow systems. High water velocities through fractures mean that contaminants can potentially
travel large distances very quickly. The direction of contaminant migration may be difficult to predict

Water Resources Engineering CVEN3501 75

Lecture 3 Pumping test analysis and occurrence of groundwater

Figure 3.17: Possible fracture configurations and the implications for groundwater supply

Water Resources Engineering CVEN3501 76

Lecture 3 Pumping test analysis and occurrence of groundwater

Figure 3.18: Daylesford Springs capped to allow Figure 3.19: Carbon dioxide rich spring water at
to be controlled. (Photo RIA) Ballymore in NSW. (Photo RIA)

from hydraulic head data, particularly if there is a strong anisotropy in hydraulic conductivity. There
is currently little published information on the movement of contaminants through fractured rocks in
Australia, although it has been highlighted as a major concern in numerous overseas studies.

3.6.3 Groundwater in basalts

Extensive areas of basalt occur widely in the eastern Sates of Australia. These are the remnants of
Tertiary age volcanos which still form areas of significant high ground. Weathering of basalt provides
the best soil available in Australia for agricultural purposes. The presence of basalt lava flows or basalt
intrusives as dykes or sills can also be a major problem where they occur in or above coal measures
that are being mined.

In central to western Victoria is an extensive area of lava plain and valley flows, with over 250 scoria
and lava volcanoes. The unit is known as the Newer Volcanics and has a thickness of up to l20 m. It is
very heterogeneous and anisotropic depending on the occurrence of jointing and vesicular zones, with
bore yields ranging from l l/s to 60 1/s The groundwater salinity ranges from less than 100 mg/l TDS to
8000 mg/l TDS . The lower salinity groundwater occurs where local recharge is high as in the younger
unweathered flows and at volcanoes and scoria cones; however these waters are often characterized by
high nitrate concentrations. Although the groundwater is mostly used for stock and domestic purposes,
it is also used for irrigation and town supplies where yields permit. The volcanic plain itself is dotted with
numerous shallow lakes which act as throughflow lakes. More recently a number of these lakes have
dried up reflecting rainfall decline and the growth in groundwater irrigation. The basalt aquifers in that
part of the volcanic plain in the western suburbs of Melbourne are affected by a number of contaminant
plumes derived from waste disposal from industry in the immediate post WW2 period.

The basalts in Victoria represent the youngest of a number of chains of volcanoes that reaches from
Queensland through New South Wales to Victoria. The basalt lava flow often cools to give hexagonal
cracking that forms characteristic columns (Fig. 3.20).

3.7 Groundwater in Tertiary palaeochannels

The Australian land surface was steadily eroded away during the 65 million years of the Tertiary period.
The rivers associated with this period of erosion formed their own deep channel systems and should be
thought of as palaeo river systems. Later in the Tertiary and into the more recent past, some of these
river systems became infilled with alluvial deposits, or in some cases, by basalt lava flows. During the
Pleistocence (last 1.5 million years) many of these systems were finally covered with a sheet of silt and
clay, efectively hiding the old (palaeo) river channel.

Water Resources Engineering CVEN3501 77

Lecture 3 Pumping test analysis and occurrence of groundwater

Figure 3.20: Hexagonal cracks formed during cooling can often transmit significant quantities of ground-
water (Photo RIA)

Very significant groundwater resources are recovered from these palaeo valleys when boreholes can
be drilled into the maximu depth of the channel. Water for irrigation is currently pumped from many of
the palaeodrainage channels along the Murray and Murrumbidgee catchments and also from channels
close to the some of the tributaries of the Darling, such as the Namoi and Gwyder systems. At Leeton
in the MIA, some bores drilled only to 150 m produce as much as 350 l/s of water with excellent quality
(fluid EC of less than 200 µS/cm). Figure 3.21 gives an example of one such bore where the water is
being used to irrigate rice, and Figure 3.22 shows a typical location of a palaeochannel which produce
such supplies (Lau et al., 1987).

The management of groundwater in Tertiary palaeochannels is a major concern as they represent the
primary water resource for a significant proportion of the irrigated agriculture in Australia. The old river
channel is frequently covered by a layer of silt and clay which makes the location of the channel difficult.
Fig. 3.22 shows the results of collecting all available geological data into a geological model of a Tertiary
channel on the Namoi River in northern NSW.

3.8 Groundwater in unconsolidated surficial deposits

Surficial aquifers occur in alluvium, colluvium, dune sands, calcrete and aeolianite, and are mostly of
Quaternary age. These exist mostly within 100 m of the surface.

The surficial aquifers are highly productive, with about 60% of abstractions in Australia coming from
these, supplying both irrigation and urban supplies (Lau et al., 1987). However, some of these aquifers
are coming under stress through overuse. Water is stored in inter granular spaces in sands, alluvium
and colluvium and gravels. The water levels show marked seasonal fluctuations in response to recharge,
as most aquifers are unconfined and recharge takes place over a significant proportion of the aquifers.
Hence, unlike some of the deeper groundwater systems, annual recharge can be significant when com-
pared to the volume of water in storage.

However, groundwater abstractions are often concentrated close to point of use (a significant advan-
tage of groundwater resources), and as pointed out above, locally abstraction and natural discharge

Water Resources Engineering CVEN3501 78

Lecture 3 Pumping test analysis and occurrence of groundwater

Figure 3.21: Irrigation bore for rice growing in the MIA. (Photo RIA)

Figure 3.22: Geological model of a palaeochannel on the Namoi River in NSW

Water Resources Engineering CVEN3501 79

Lecture 3 Pumping test analysis and occurrence of groundwater

can greatly exceed recharge, giving rise to resource depletion, saline intrusion and knock-on effect on
surface water ecosystems which are dependent on groundwater (lakes, wetlands, swamps etc) or for
base-flow (rivers, streams).

These impacts can be alleviated to some extent by artificial recharge, but it has become increasingly
obvious that careful management is required for allocation of groundwater supplies from these aquifers.

3.8.1 Swan Coastal Plain

Sand aquifers on the Swan Coastal plain contribute over 70% of water supplies, and up to 35% of
drinking water supplies for the city of Perth and associated coastal areas. The aeolian sands and
coastal limestone (aeolianite) provide excellent aquifers which are readily recharged by winter rains.
The groundwater systems also support extensive wetlands and swamplands throughout the plain, and
groundwater discharges to both ephemeral streams along the eastern edge of the plain and to the Swan
river and Moore River on the central and northern part of the plain as well as to the Indian Ocean in the
west.

The pattern of recharge and surface discharge produces groundwater mounds where the phreatic sur-
face is at its highest elevation in the central parts of the plain, and decreases in elevation towards the
discharge points at the east and west margins and into westward flowing rivers and streams which tra-
verse the plain. The undulating surface of the sand dunes intercepts the phreatic (water table) surface in
dune swales, forming wetlands, damp lands and swamps, which are entirely dependent on groundwater
for their existence. Urban and agricultural development on the coastal plain have impacted wetlands
to such an extent that most of these in the urban and associated horticultural areas are eutrophic,
and although there has been considerable effort put into management of water levels through artificial
recharge, conjunctive use of groundwater and surface water, groundwater modelling and monitoring,
water levels have shown a steady downward trend since developments began in the 1950s. Some wet-
lands have disappeared and others have been inundated from land drainage so that fringing vegetation
has died. Phreatophytic vegetation (deep-rooted trees and shrubs which obtain water from the water
table) has also been severely affected in some areas close to water supply well fields.

The situation on the Swan Coastal Plain is by no means uncommon, where exploitation of shallow
fresh groundwater supplies in unconfined aquifers beneath the expanding city not only draws down the
groundwater table, but also gives rise to contamination of water supplies from a wide range of activ-
ities. Perth is relatively unique in that land-use restrictions have been applied to the developing city
suburban areas, to help minimise impacts on groundwater quality. Perhaps the most pressing problem
in areas like the Swan Coastal Plain is our lack of understanding of the dependence of natural ecosys-
tems on groundwater, which limits our ability to manage the resource whilst maintaining environmental
sustainability. It is clear that groundwater on the Swan Coastal Plain maintains a wide range of wet-
lands, supports phreatophytic vegetation, and also now has been identified as a significant driving force
expelling nutrients from anoxic sediments into the Swan River, causing nuisance algal blooms.

3.8.2 Sand dune aquifers of the east Australian coast

There are major aquifer systems in wind blown sands along the east coast of Australia. These include
the Botany Aquifer and the Tomago Sands in New South Wales and Moreton Island, Stradbroke Island
and Fraser Island in Queensland. Fraser Island is the largest sand dune island in the world.

Each of these aquifers systems contains very large quantities of fresh groundwater and supports com-
plex ecosystems associated with the groundwater discharge. Fig. 3.23 shows a major perched ground-
water lake (Brown Lake) on North Stradbroke Island.

Water Resources Engineering CVEN3501 80

Lecture 3 Pumping test analysis and occurrence of groundwater

Figure 3.23: Perched groundwater lake on North Stradboke Island, Queensland (Photo RIA)

3.9 Groundwater on oceanic islands

Groundwater is a vital source of water on many tropical islands. The island of O’ahu (Hawai’i) relies on
groundwater for 90% of its drinking water and about half of all water used for irrigation. During El Niňo
episodes, rains often fail in the Pacific and groundwater becomes the only source of water.

The occurrence of groundwater on oceanic islands depends upon the geology of the island. Volcanic
islands, such as Lord Howe, have groundwater associated with basalt as well as groundwater in calcrete
or sand deposits. In the Pacific Ocean, it is frequently the case that limestone has formed on top of
volcanic lava producing quite a complex aquifer system. Tribble (2008) has reported on a number of
typical occurrences from the Pacific.

The most important concept is that fresh water has a lower density than sea water and will therefore
tend to float on seawater in the aquifer. The interface between the two is susceptible to movement as a
result of groundwater abstraction. This makes the management of groundwater lenses a difficult issue
that requires careful monitoring. Fig. 3.24 shows the relationship between freshwater and groundwater.

Groundwater may occur at high levels in islands (Fig. 3.25) (Tribble, 2008) where lithological changes
give rise to perched aquifers; or may be more extensive at low levels (Fig. 3.26) (Tribble, 2008).

The impact on hydraulic head measurements where salinity varies is discussed in Chapter ?? and spe-
cific geophysical techniques applicable to this problem are presented in Chapter ??. There are signifi-
cant issues with groundwater dependent ecosystems that are also covered in Chapter ??. Groundwater
development may give rise to upconing as shown in Fig. 3.27 (Tribble, 2008).

Water Resources Engineering CVEN3501 81

Lecture 3 Pumping test analysis and occurrence of groundwater

Figure 3.24: Freshwater lens beneath an island

Figure 3.25: Aquifers can be perched well above the regional water level

Water Resources Engineering CVEN3501 82

Lecture 3 Pumping test analysis and occurrence of groundwater

Figure 3.26: Extensive sand dune aquifer systems or systems developed on low permeability layers

Figure 3.27: Upconing of saline groundwater below a pumping well

Water Resources Engineering CVEN3501 83

Lecture 3 Pumping test analysis and occurrence of groundwater

3.10 Summary

Groundwater is a resource which is big, pervasive, vulnerable and enduring. Groundwater issues sur-
round us. They can almost overwhelm us, there is so much to be done to sustain and protect the
resources, and the surface water and ecosystems which they support, and to overcome the problems of
salinity and pollution.

We need to understand the fundamental processes controlling groundwater quantity and quality, and
there is still much to learn. Australian hydrogeologists are developing an understanding of groundwater
processes. We have seen the linkage between recharge rates and salinity, due to irrigation and dryland
clearing. We are still developing an understanding of the dependence of ecosystems on groundwater,
and nutrient and solute attenuation and interactions between groundwater and surface water systems.
The community has a right to expect us to understand these relationships. If we don’t who will?

We need to improve quantitative assessments to allow better resource management: the following would
seem to be most problematic:

• recharge estimation in fractured media

• surface water-groundwater interactions and ecosystem dependence on groundwater
• impacts of land-use change and waste management on groundwater, and
• remediation and restoration of degraded resources
• definition of capture zones for wells and springs
• optimisation of groundwater abstraction to allow resource and environmental sustainability

We are good on measuring concentrations, and poor on measuring mass fluxes to aquifers and assess-
ing ”sustainable yield”. For total aquifer management we will need to develop such a capacity.

Our past failures have not been deliberate. They have been caused by not having a clear understanding
of the groundwater system, groundwater sources, or fluxes, flow mechanisms, discharge points, or rates
of attenuation of contaminants.

We need to educate communities as trustees. The local community is normally the best group to man-
age their shared resource (certainly for unconfined aquifers and regional groups for confined aquifers)
so they can understand their problem and choose their own solutions. It is our role to provide them
with the knowledge they need to make intelligent decisions for themselves. Community participation in
groundwater management is essential.

3.11 Connectivity of Surface Water and Groundwater

3.11.1 Introduction

Winter et al. (1998) from the U.S. Geological Survey published a seminal volume ”Ground Water and
Surface Water: A Single Resource”.
It can be down loaded as a .pdf file from http://pubs.usgs.gov/circ/circ1139/pdf/front.pdf.

Base Flow

Streams in catchments overlying aquifers will be continuously fed with base flow from the aquifer, as long
as the hydraulic head in the aquifer is above the stream bed (Fig. 3.28). These streams are known as

Water Resources Engineering CVEN3501 84

Lecture 3 Pumping test analysis and occurrence of groundwater

Figure 3.28: A gaining stream Figure 3.29: Losing stream

gaining streams. The quantity of base flow will depend upon the hydraulic gradient between the aquifer
and the stream (Winter et al., 1998).

In a gaining stream, the water table slopes towards the stream so that there is a hydraulic gradient
towards the stream. Conversely, in a losing stream, the water table slopes away from the stream. A
stream which is normally a gaining stream may temporarily become a losing stream during a flood. The
water that moves into the aquifer is refereed to as bank storage by hydrologists. This water returns to
the stream after the flood peak has passed and the hydraulic gradient returns to its normal configuration.
Fig. 3.29 shows a losing stream.

Groundwater abstraction close to a river may reverse the normal hydraulic gradient creating a recharge
boundary for the aquifer. Water then flows continuously from the stream to the aquifer and towards the
abstraction bore. This effect is often utilised to filter river water using the channel bank sands, prior to
abstraction. An example of an installation of this type is given by Grischek et al. (1996) in Figure 3.30.

The River Elbe flows over Quaternary deposits consisting mostly of glacial gravels and coarse sands up
to 15m in thickness. Pumping from abstraction bores on the banks induces recharge from the river to
the aquifer.

There maybe more substantial flow into the beds of rivers from groundwater discharge or the springflow
is so large that it forms a stream straight away. The major springs associated with basalts aroun Rotorua
in New Zealand come into this category, as shown in Fig 3.31.

The type of rock in the catchment will have a major impact upon the base flow response in a stream.
If rocks have little porosity and low storage, then the response to rainfall in the stream will be very
spikey. By contrast, rain falling on a limestone aquifer rapidly infiltrates and even major storms will have
a very subdued response in the streams which will continue to discharge water from storage. Data from
Zambia illustrates this difference in response. In Fig 3.32 the hydrographs from a limestone dominated
catchment and a basement dominated catchment are shown. The catchments are adjacent to each
other outside Lusaka.

Water Resources Engineering CVEN3501 85

Lecture 3 Pumping test analysis and occurrence of groundwater

Figure 3.30: River bank abstraction from the Elbe River, Germany

Figure 3.31: Major spring flow through the base of the stream issuing from volcanic rocks around Ro-
torua in New Zealand

Water Resources Engineering CVEN3501 86

Lecture 3 Pumping test analysis and occurrence of groundwater

Figure 3.32: Difference in run-off between a limestone dominated catchment and a basement dominated
catchment

Water Resources Engineering CVEN3501 87

Lecture 3 Pumping test analysis and occurrence of groundwater

Figure 3.33: Water level in lakes and bores in Centennial Park, Sydney

3.11.2 Interaction of Lakes or Wetlands with Groundwater

When lakes and wetlands occur in sands and gravels, they will be inextricably linked to the groundwater
system. The Lachlan Swamps or the ponds in Centennial Park provide good examples in the Botany
Sands Aquifer (Fig 3.33) where flow through the Botany Aquifer in Centennial Park is shown to be
directly related to the presence of ponds. Townley and Trefry (2000) gave a detailed assessment of the
3D head distribution around lakes that can be used to better understand the movement of groundwater
into, through or from lakes. These authors carried out extensive work on the Swan Coastal Plain.

Fetter (2001) provides an indication of the possible relationship between streams, lakes and groundwater
in Figure 3.34.

Wherever groundwater flow directs water close to the surface (Fig 3.35, wetlands are likely.

Winter et al. (1998) give further examples shown in Figs 3.36 and 3.37. As shown

3.11.3 Hyporheic Zone

Some gaining streams have reaches that lose water to the aquifer under normal conditions of stream
flow (Winter et al., 1998). This local recycling of water is considered to be very important in cycling
nutrients from the ground into the stream system. The zone of interchange, which occurs vertically in a
pool and riffle system (Fig. 3.38 and Fig 3.39), or horizontally in a meander loop (Fig 3.40), is called the
hyporheic zone.

3.11.4 Groundwater Discharge to the Ocean

One of the most complex zones where groundwater mixes with surface water, is at the beach. The
USGS Circular 1262 by Barlow (2003) is an excellent summary and can be downloaded from the web at
http://water.usgs.gov/ogw/gwrp/saltwater/index.html.

Fig. 3.41 gives a typical representation of the groundwater seawater interface zone drawn for the Atlantic
margin in the USA, but could equally represent any beach zone adjacent to a non-destructive plate
margin. The east and west coasts of Australia fall in this category.

Water Resources Engineering CVEN3501 88

Lecture 3 Pumping test analysis and occurrence of groundwater

Figure 3.34: Interactions between streams, lakes and groundwater

Figure 3.35: Origin of wetlands from groundwater discharge

Water Resources Engineering CVEN3501 89

Lecture 3 Pumping test analysis and occurrence of groundwater

Figure 3.36: A gaining stream Figure 3.37: Losing stream

Figure 3.38: Schematic of a pool and ripple sys- Figure 3.39: Pool and riffle system in Colorado
tem

Figure 3.40: Meander belt

Water Resources Engineering CVEN3501 90

Lecture 3 Pumping test analysis and occurrence of groundwater

Figure 3.41: Typical groundwater interface with sea water at a passive plate margin beach

The dynamics of the tides and waves at the beach directly impact the discharging groundwater Turner
et al. (1996); Acworth and Dasey (2003) with groundwater being set up at an elevation close to the mean
high tide mark. This observation has a major implication for models of groundwater flow that use the
beach as a constant-head boundary.

Many possible configurations of fresh water and groundwater are possible. The controlling factor will
the the variation in hydraulic conductivity of the sediments. Fig 3.42 shows relationships in the Florida
aquifers (Barlow, 2003).

On a beach scale, there are many important processes that are under investigation, with the chemical
implications of mixing proving particularly interesting. There is now a suspicion that groundwater flux to
the oceans is a major source of contaminant and nutrients. Fig 3.43 shows the dependence of vegetation
on the beach to different salinity water. There area also major implications for the biota.

Water Resources Engineering CVEN3501 91

Lecture 3 Pumping test analysis and occurrence of groundwater

Figure 3.42: Multiple interfaces between fresh groundwater and sea water in Florida

Water Resources Engineering CVEN3501 92

Lecture 3 Pumping test analysis and occurrence of groundwater

Figure 3.43: Beachface dynamics

Water Resources Engineering CVEN3501 93

Lecture 3 Pumping test analysis and occurrence of groundwater

3.12 Interested in further groundwater studies?

If you have got to the end of the notes and find that you are interested in this variable and challenging
area, then consider selecting the elective Groundwater Resource Investigation CVEN4503 in Session 1
every year.

Please note that the course involves a 3-day field trip to the UNSW Farm at Wellington (between Orange
and Dubbo) during the non-teaching week after Easter. Student numbers are limited to 45 - so get in
early if you are interested!

There is no written examination but a major field report (group) using the data you collect to investigate
the groundwater resources on the farm.

You can also download various chapters of the Australian Groundwater School at the same location.

Water Resources Engineering CVEN3501 94

Bibliography

Acworth, R. I. (1987). The development of crystalline basement aquifers in a tropical environment.

Quarterly Journal of Engineering Geology, 20:265–272.
Acworth, R. I. (2001). The electrical image method compared with resistivity sounding and electro-
magnetic profiling for investigation in areas of complex geology - A case study from ground-water
investigation in a fractured rock environment. Exploration Geophysics, 32:119–128.

Acworth, R. I. (2007). Investigation of shallow vertical head variation in an unconfined sandy aquifer
using a multi-channel manometer board. Hydrogeology Journal, 15(7):1291–1306.
Acworth, R. I. and Dasey, G. R. (2003). Mapping of the hyporheic zone around a tidal creek using a
combination of borehole logging, borehole electrical tomography and cross-creek electrical imaging.
Hydrogeology Journal, 11(3):368–377.
Allen, A. D., Laws, A. T., and Commander, D. P. (1992). A review of major water resources in Western
Australia. Technical report, Kimberly Water Resources Development Office.
Barlow, P. M. (2003). Ground Water in Freshwater-Saltwater Environments of the Atlantic Coast. Circular
1262, U.S. Geological Survey, USGS, Reston, Virginia.

Cook, P. G. (2003). A guide to regional groundwater flow in fractured rock aquifers. Technical report,
CSIRO, CSIRO Land and Water, Glen Osmond, SA, Australia.
Darcy, H. (1856). Les fontaines publiques de la ville de Dijon. Victor Dalmont.
Domenico, P. A. and Schwartz, F. W. (1997). Physical and Chemical Hydrogeology. John Wiley and
Sons, second edition.
Driscoll, F. G. (1986). Groundwater and Wells. Johnson Division, St. Paul, Minnesota 55112, second
edition.
Fetter, C. (2001). Applied Hydrogeology. Prentice Hall, Upper Saddle River, New Jersey 07458, USA,
fourth edition.
Feynman, R. P., Leighton, R. B., and Sands, M. (1969). The Feynman Lectures in Physics - Volume I.
Addison-WesleyPublishing Company.
Frank, C. L. and Reilly, T. E. (1987). The effects of boundary conditions on the steady-state response of
three hypothetical groundwater systems - results and implications for numerical experiments. Water-
Supply Paper 2315, US Geological Survey.
Freeze, R. A. and Cherry, J. A. (1979). Groundwater. Prentice-Hall, Inc., Englewood Cliffs, NJ.
Great Artesian Basin Consultative Council (1998). Great Artesian Basin Resource Study Summary.
Grischek, T., Nestler, W., Piechniczek, D., and Fischer, T. (1996). Urban Groundwater in Dresden,
Germany. Hydrogeology Journal, 4(1):48–63.
Habermehl, M. A. (1983). The Great Artesian Basin. In Groundwater and Man - Conference Proceed-
ings, Sydney. Unknown.
Hubbert, M. K. (1940). Theory of ground-water motion. Journal of Geology, 48(8):795–944.

95
Lecture BIBLIOGRAPHY

Jacobson, G., Jankowski, J., and Abell, R. S. (1991). Groundwater and surface water interaction at Lake
George, New South Wales. BMR Journal of Australian Geology & Geophysics, 12:161–190.
Jiao, J. J. (2007). A 5,600-year-old wooden well in zhejiang province, china. Hydrogeology Journal
Geological Science (Czech), 15:1021–1029.
Lau, J. E., Commander, D. P., and Jacobson, G. (1987). Hydrogeology of Australia. Bulletin 227, Bureau
of Mineral Resources, Geology and Geophysics, Australian Government Publishing Office, Canberra.
Meinzer, O. E. (1923). The occurrence of groundwater in the United States, with a discussion of princi-
ples. Water-Supply Paper 489, U.S. Geological Survey.

Price, M. (2009). Barometric water-level fluctuations and their measurement using vented and non-
vented pressure transducers. Quarterly Journal of Engineering Geology and Hydrogeology, 42:245–
250.
Prime Minister’s Science, Engineering and Innovation Council (1999). Dryland salinity and its impacts
on rural industries and the landscape. World Wide Web. Version 2, as at 7 January 1999.
Shugg, A. and Knight, M. J. (1994). Hydrogeology of the Mineral Springs of Central Victoria. In Wa-
ter Down Under 94: Groundwater Papers, pages 215–221, Barton, ACT. Institution of Engineers,
Australia.
Standards Australia (1990). Test pumping of water wells. Australian Standard AS 2368-1990, Standards
Australia.
Theis, C. V. (1935). The relationship between the lowering of the piezometric surface and the rate and
duration of discharge of a well using groundwater storage. Transactions American Geophysical Union,
2:519–524.

Townley, L. R. and Trefry, M. G. (2000). Surface water-groundwater interaction near shallow circular
lakes: Flow geometry in three dimensions. Water Resources Research, 36:935–948.
Tribble, G. (2008). Ground water on tropical pacific islands - understanding a vital resource. Circular
1312, US Gelogical Survey.
Turner, I. L., Coates, P., and Acworth, R. I. (1996). The effects of tides and waves on water-table
elevations in coastal zones. Hydrogeology Journal, 4(2):51–69.
Wang, H. F. and Anderson, M. P. (1982). Introduction to Groundwater Modelling - Finite Difference and
Finite Element Methods. Academic Press, Inc.
Weaver, T. R., Cartwright, I., Tweed, S. O., Ahearne, D., Cooper, M., Czapnik, K., and Tranter, J. (2006).
Controls on chemistry during fracture-hosted flow of cold CO2-bearing mineral waters, Daylesford,
Victoria, Australia: Implications for resource protection. Applied Geochemistry, 21(2):289–304.
Winter, T. C., Harvey, J. W., Franke, O. L., and Alley, W. M. (1998). Ground Water and Surface Water -
A Single Resource. Geological Survey Circular 1139, U.S. Geological Survey. Available from USGS
Denver, Colorado.

Water Resources Engineering CVEN3501 96