CEWB4013/CEWB493 Hydropower and Dam Engineering, Lecturer: Dr.

Wong Leong Sing

MODULE 3: Interpretation of maps and results of

geophysical methods and design control techniques of
seepage, pore pressures, internal erosion and piping
3.1 Geophysical exploration
3.1.1 Seismic refraction survey
3.1.2 Cross-hole seismic survey
3.1.3 Resistivity survey
3.2 Seepage for dam engineering
3.2.1 Seepage through earth dams and embankments
3.2.2 Seepage in anisotropic soil
3.2.3 Instability due to seepage
1
3.1 Geophysical exploration

• Several types of geophysical exploration techniques permit a rapid evaluation of

subsoil characteristics.
• These methods also allow rapid coverage of large areas and are less expensive than
conventional exploration by drilling.
• However, in many cases, definitive interpretation of the results is difficult. For that
reason, such techniques should be used for preliminary work only.
• Here, we discuss three types of geophysical exploration technique: the seismic
refraction survey, cross-hole seismic survey, and resistivity survey.

2
3.1.1 Seismic refraction survey

• Seismic refraction surveys are useful in obtaining preliminary information about the
thickness of the layering of various soils and the depth to rock or hard soil at a site.
• Refraction surveys are conducted by impacting the surface, such as at point A in
Figure 3.2a, and observing the first arrival of the disturbance (stress waves) at
several other points (e.g., B,C, D,…).
• The impact can be created by a hammer blow or by a small explosive charge.
• The first arrival of disturbance waves at various points can be recorded by
geophones.

3
Figure 3.1: Method of seismic refraction survey 4
• The impact on the ground surface creates two types of stress wave: P waves (or
plane waves) and S waves (or shear waves).
• P waves travel faster than S waves; hence, the first arrival of disturbance waves will
be related to the velocities of the P waves in various layers.
• The velocity of P waves in a medium is

Eq. (3.1)

where
Es = modulus of elasticity of the medium
γ = unit weight of the medium
g = acceleration due to gravity
μs = Poisson’s ratio

Figure 3.2: Seismic refraction survey 5

• To determine the velocity v of P waves in various layers and the
thicknesses of those layers, we use the following procedure:
Step 1. Obtain the times of first arrival, t1, t2, t3, …, at various distances x1,
x2, x3, … from the point of impact.
Step 2. Plot a graph of time t against distance x. The graph will look like
the one shown in Figure 3.2b.
Step 3. Determine the slopes of the lines ab, bc, cd, … :

Here, v1, v2, v3, … are the P-wave velocities in layers I, II, III, … ,
respectively (Figure 3.2a).
6
Step 4. Determine the thickness of the top layer:

Eq. (3.2)

The value of xc can be obtained from the plot, as shown in Figure 3.2b.
Step 5. Determine the thickness of the second layer:
Eq. (3.3)

Here, Ti2 is the time intercept of the line cd in Figure 3.2b, extended
backwards.
• The velocities of P waves in various layers indicate the types of soil or
rock that are present below the ground surface.
• The range of the P-wave velocity that is generally encountered in
different types of soil and rock at shallow depths is given in Table 3.1.
7
Table 3.1: Range of P-Wave Velocity in Various Soils and Rocks

8
In analyzing the results of a refraction survey, two limitations need to be kept in
mind:
1. The basic equations for the survey - that is, Eqs. (3.2) and (3.3) - are based
on the assumption that the P-wave velocity v1 < v2 < v3 < … .
2. When a soil is saturated below the water table, the P-wave velocity may be
deceptive.
P waves can travel with a velocity of about 1500 m/sec (5000 ft/sec) through
water.
For dry, loose soils, the velocity may be well below 1500 m/sec.
However, in a saturated condition, the waves will travel through water that is
present in the void spaces with a velocity of about 1500 m/sec.
If the presence of groundwater has not been detected, the P-wave velocity may
be erroneously interpreted to indicate a stronger material (e.g., sandstone) than
is actually present in situ.
In general, geophysical interpretations should always be verified by the results
obtained from borings.
9
Worked Example 3.1.
The results of a refraction survey at a site are given in the following table:

Determine the P-wave velocities and the thickness of the material

encountered. 10
Solution
Velocity in Figure 3.3, the times of first arrival of the P waves are plotted
against the distance of the geophone from the source of disturbance. The
plot has three straight-line segments. The velocity of the top three layers
can now be calculated as:

11
Figure 3.3: Plot of first arrival time of P wave versus distance of geophone from
source of disturbance 12
Comparing the velocities obtained here with those given in Table 3.1
indicates that the third layer is a rock layer.
Thickness of Layers
From Figure 3.3, xc = 10.5 m, so
13
14
Again, from Eq. (3.3)

15
3.1.2 Cross-hole seismic survey

• The velocity of shear waves created as the result of an impact to a given layer of soil
can be effectively determined by the cross-hole seismic survey.
• The principle of this technique is illustrated in Figure 3.4, which shows two holes
drilled into the ground a distance L apart.
• A vertical impulse is created at the bottom of one borehole by means of an impulse
rod. The shear waves thus generated are recorded by a vertically sensitive
transducer. The velocity of shear waves can be calculated as
Eq. (3.4)
where t = travel time of the waves.
16
Figure 3.4: Cross-hole method
of seismic survey

17
• The shear modulus Gs of the soil at the depth at which the test is taken
can be determined from the relation

Eq. (3.5)

where
vs = velocity of shear waves
γ = unit weight of soil
g = acceleration due to gravity
18
3.1.3 Resistivity survey

• Another geophysical method for subsoil exploration is the electrical resistivity survey.
• The electrical resistivity of any conducting material having a length L and an area of
cross section A can be defined as
Eq. (3.6)
• where R = electrical resistance.
• The unit of resistivity is ohm-centimeter or ohm-meter.
• The resistivity of various soils depend primarily on their moisture content and also on
the concentration of dissolved ions in them.

19
• Saturated clays have a very low resistivity; dry soils and rocks have a
high resistivity.
• The range of resistivity generally encountered in various soils and rocks
is given in Table 3.2.
• The most common procedure for measuring the electrical resistivity of a
soil profile makes use of four electrodes driven into the ground and
spaced equally along a straight line.
• The procedure is generally referred to as the Wenner method (Figure
3.5a).
• The two outside electrodes are used to send an electrical current I
(usually a dc current with nonpolarizing potential electrodes) into the
ground.
• The current is typically in the range of 50 to 100 milliamperes.

20
• The voltage drop, V, is measured between the two inside electrodes.
• If the soil profile is homogeneous, its electrical resistivity is

Eq. (3.7)

• In most cases, the soil profile may consist of various layers with different
resistivities, and Eq. (3.7) will yield the apparent resistivity.
• To obtain the actual resistivity of various layers and their thicknesses, one may
use an empirical method that involves conducting tests at various electrode
spacings (i.e., d is changed). The sum of the apparent resistivities, or, is
plotted against the spacing d, as shown in Figure 3.5b.
• The plot thus obtained has relatively straight segments, the slopes of which
give the resistivity of individual layers.
• The thicknesses of various layers can be estimated as shown in Figure 3.5b.
• The resistivity survey is particularly useful in locating gravel deposits within a
fine-grained soil.
21
Table 3.2: Representative Values of Resistivity

22
Figure 3.5: Electrical resistivity
survey: (a) Wenner method; (b)
empirical method for determining
resistivity and thickness of each
layer

23
3.2 Seepage for dam engineering

• Seepage involves movement of water in soils, and is often a critical problem in dam
engineering.
• Seepage depends on several factors, including permeability of the soil and the
pressure gradient, essentially the combination of forces acting on water through
gravity and other factors.
• Permeability can vary over a wide range, depending on soil structure and
composition, making possible the safe design of earth dams with negligible leakage
loss.

24
3.2.1 Seepage through earth dams and embankments

• Whereas the flow under impermeable structures, such as concrete or masonry dams,
is confined, the seepage taking place through a permeable structure, such as earth
dam, is unconfined.
• In such problems, the upper boundary of the seepage zone is the phreatic surface,
this being the top flow line and along which the pressure is that of the atmosphere.
• The first step in drawing a flow net for seepage through a dam or embankment is
therefore to locate and draw the phreatic surface or top flow line.

25
Entry and exit conditions
• The upstream surface of the dam is the entry surface to the seepage
zone and since it is the equipotential Φmax. Representing the maximum
pressure head, the flow line must intersect it at right angles (Figure 3.6).

Figure 3.6: Phreatic surface detail at entry to seepage zone

26
• At the downstream or exit surface, the theoretical parabola may have to
be modified depending on the conditions at the toe (Figure 3.7).

Figure 3.7: Phreatic surface detail at exit to seepage zone

27
Figure 3.8: Seepage in earth dam
28
3.2.2 Seepage in anisotropic soils

• Where a soil consists of a number of layers of different soil types each with different
permeability, the horizontal and vertical permeabilities will be different.
• The presence of thin sandy layers (high permeability) in thick layers of fine soil (low
permeability) can very often produce relatively high horizontal seepage rates.
• Considering the soil mass shown in Figure 3.9 which consists of three layers, each of which
has a different coefficient of permeability.
• Horizontal flow (i.e. tangential to strata). The head lost between the entry and exit faces will
be the same for each later:
• h1 = h2 = h3 = h
• Hence the hydraulic gradients are the same:
• i1 = i2 = i3 = i

29
Figure 3.9: Flow in
stratified soil
(a) Horizontal flow
(b) Vertical flow

30
Giving Eq. (3.8)

31
 • In both Equations 3.8 and 3.9,
the number of components in both
the numerator and denominator is
Giving equal to the number of layers.
Eq. (3.9)
32
Worked Example 3.2.
A stratified soil consists approximately of alternating layers of sand and silt.
The sand layers are generally 150 mm in thickness and have a
permeability of k = 6.5 × 10-1 mm/s, the silt layers are 1.80 m thick and
have a k = 2.5 × 10-4 mm/s. Assuming that within each layer flow
conditions are isotropic, determine the ratio of the horizontal permeability
to that of the vertical.
Solution:
One strata cycle will consist of a layer of sand (1) and a layer of silt (2).

33
34
Flow net in anisotropic soil
In an anisotropic soil kx ≠ kz and, and in the average direction of flow, kf
has a value between kx and kz. The Laplace equation for two-dimensional
flow therefore becomes:
Eq. (3.10)

Eq. (3.11)

Eq. (3.12)
Equation [3.12]

where H = total head loss Eq. (3.13)

Nf and Nd = number of flow
channels and potential drops,
respectively (from
flow net drawn) Eq. (3.14) 35
To construct the flow net, use the following procedure:

• Note that when flow nets are drawn in transformed sections (in anisotropic soils), the
flow lines and the equipotential lines are orthogonal.
• However, when they are redrawn in a true section, these lines are not at right angles
to each other.
• This fact is shown in Figure 3.10. In this figure, it is assumed that kx = 6kz.
• Figure 3.10a shows a flow element in a transformed section.
• The flow element has been redrawn in a true section in Figure 3.10b.
36
Figure 3.10: A flow
element in anisotropic
soil: (a) in transformed
section; (b) in true
section
37
Worked Example 3.3.
A dam section is shown in Figure 3.11a. The hydraulic conductivity of the
permeable layer in the vertical and horizontal directions are 2 × 10-2 mm/s
and 4 × 10-2 mm/s, respectively. Based on the drawn flow net, calculate
the seepage loss of the dam in m3/day/m.

Solution
From the given data,

kz = 2 × 10-2 mm/s = 1.728 m/day

kx = 4 × 10-2 mm/s = 3.456 m/day

and H = 6.1 m. For drawing the flow net,

38
Figure 3.11

39
 Figure 3.11b.
Figure 3.11b,

40
3.2.3 Instability due to seepage (‘piping’)

• The term piping is used to describe the unstable condition which can occur when the
vertical component of seepage pressure acting in an upward direction exceeds the
downward weight of the soil.
• When the upward seepage force becomes equal to the submerged weight of the soil,
no frictional resistance can be developed between the particles. The soil/water
mixture therefore has no shear strength and act as a liquid.
• If the upward seepage forces exceed the submerged weight, the particles may be
carried upwards to be deposited at the ground surface. Thus a ‘pipe’ is formed in the
soil near the surface.

41
• Piping failure can lead to the collapse of a supporting structure, such as
the toe of a dam, or part of a cofferdam.
• It is necessary, therefore, to check the potential instability condition
during the design of water-retaining structures.
• Where a flow net has been drawn to represent the seepage conditions, a
simple rule-of-thumb method may be used to determine a factor of safety
against the occurrence of piping.
• This is done by considering the downstream face of the structure.
• Consider the case of a coffer dam as shown in Figure 3.121.
• The effective weight of the prism of soil ABCD will be

42
Figure 3.12: Factor of safety against piping for cofferdam
43
Eq. (3.15)

44
• The method of calculating us is based on the pressure head multiply by
the gravity.
• The factor of safety against piping maybe increased in a number of
ways.
• For example, in the case of a cofferdam the depth of penetration of the
piles could be increased, or a layer of coarse filter material could be laid
on the downstream side before pumping down to the final level.
• In the case of dams, both an increase in the factor of safety against
piping and a reduction in the quantity of seepage can be obtained by
increasing the length of the flow path.
• This may be done by driving a row of sheet piles, preferably at or near
the upstream face, or by laying an apron of impermeable paving in
front of the upstream face.
• A layer of coarse filter material laid on the downstream side is another
possibility (See Figure 3.13). 45
Figure 3.13: Methods of improving seepage conditions (a) Cofferdam (b) Concrete or masonry dam
46
 • Coarse filter material – Coarse gravels.
• Sheet piles – Steel sheet piles.
• Impermeable apron – Concrete layer.

Figure 3.14: Cofferdam sheet piles

47
Worked Example 3.4.
Figure 3.12 shows the cross section of a cofferdam at 15 m height in an
alluvial silty clay, which is underlain by an impermeable stratum of bedrock.
The average pressure head (hs) at the base BC was determined to be 6.85
m. The lower and upper water levels were found to be 12 and 20 m
respectively from the impermeable stratum. The soil level at the upper
water level was measured to be 16 m from the impermeable stratum. The
upper water level was noticed to be 1 m from the surface of the cofferdam.
The saturated unit weight of the alluvial silty clay is 18 kN m-3. Evaluate
the factor of safety against piping for the cofferdam. If the cofferdam is
unsafe, suggest a remediation measure and based on the measure;
design a new factor of safety which is deemed to be safe for the coffer
dam.

48
Solution
• Average pore water pressure at the base BC, us = hs × g = 6.85 × 9.81 =
67.20 kPa.
• Depth of the prism, D = 15 - (20 - 12 + 1) = 6 m.
• Saturated unit weight of alluvial silty clay, γsat = 18 kN m-3.
• Unit weight of water, γw = 9.81 kN m-3.

F( piping) 
 sat   w D 18  9.816
  0.73 < 4
us 67.20
• The cofferdam is unsafe.
• Coarse filter material (Coarse basaltic gravels) is suggested as a
remedial measure.
• Unit weight of basaltic gravel, γgravel = 32 kN m-3.
49
• Design a new factor of safety against piping, F(piping) = 4.10.

 sat   w D  ( gravel  H gravel )

F pipingnew 
us
18  9.816  (32  H gravel )
4.10 
67.20
• Hgravel = 7.07 m ≈ 7.10 m.

50