PERMEABILITY

The flow of gravitational water through soil is called percolation, and the property of soil which permits
such percolation is called permeability. The percolation of water through soil was first studied by Darcy
(1856) who demonstrate experimentally that for laminar flow conditions in a saturated soil the discharge
velocity, v of water is proportional to the hydraulic gradient i.

v = k i --- --- --- (1)

Where, k = Darcy coefficient of permeability, m/s

The hydraulic gradient i is the ratio of head loss h over a distance l.

The discharge velocity v is defined as the quantity of water, q percolating through a cross sectional area
A in unit time.

In equation 1, v is the discharge velocity of water based on the gross cross sectional area of the soil.
However, the actual velocity of water (that is seepage velocity) through the void spaces is greater than v.
If quantity of water flowing through soil in unit time is q, then

q = v A = Av vs --- --- --- (2)

where, vs = seepage velocity, Av = area of void in the cross section of the specimen.

However, A = Av + As --- --- --- (3)

Where, As = area of soil solids in the cross section of the specimen.

Combining equation 2 & 3 gives,

q = v (Av + As) = Av vs

1
or, = = --- --- --- (4)

Where, vv = volume of voids in the specimen

vs = volume of soil solids in the specimen

Equation 4 can be written as

= = --- --- --- (5)

Where, e = void ration

n = porosity
The constant of proportionality between the seepage velocity and hydraulic gradient is called coefficient
of percolation kp.

vs = kp i ---- ----- ----- ----- (6)

Approximate values of coefficient of permeability for different types of soils are given in Table – 1.

Table – 1. Coefficient of permeability of various soils

Coefficient of
Type of soil permeability
(cm/sec)
Clean gravel 1.0-100
Coarse sand 1.0-0.01
Fine sand 0.01-0.001
Silty clay 0.001-0.00001
Clay <0.000001
Validity of Darcy’s Law

Darcy’s law is valid only for laminar flow. It has been borne out by experiments that the limits of
validity of Darcy’s law may be fixed with respect to particle size, velocity of flow and hydraulic
gradient. Francher, Lewis and Barnes demonstrated experimentally that flow through sands remains
laminar and Darcy’s law valid so long as the Reynolds number expressed in the form of equation 7
given below, is equal to or less than unity:

2
 ---- ----- ----- ----- (7)

Where, v = velocity of flow (cm/sec)

Da = diameter of average particle, assumed spherical (cm)
According to Allen Hazen’s experiments, the rate of flow remains proportional to the hydraulic gradient
if the effective size of the oil does not exceed 3 mm. The validity limits of Darcy’s law as plotted by
Prinz, based on Piefke’s experiments on sand are presented in Figure 2.

It is observed from Fig. 2 that the law of proportionality between velocity and hydraulic gradient
deviates faster in very fine sands than in coarser soils. Coarser the soil, the lower is the hydraulic
gradient required for laminar flow.

In short, the Darcy’s law is valid for a wide range of materials and flow conditions, both for coarse
grained and fine grained soils. But it is rather difficult to precisely define the validity limits for a specific
set of conditions, without performing extensive permeability tests. For the ground water flow occurring
in nature and normally encountered in soil engineering, the law is generally within its validity limits.

Factors Affecting Permeability

The hydraulic conductivity or permeability of soil depends on several factors: fluid viscosity, pore size
distribution, grain size distribution, void ratio, roughness of mineral particles and degree of soil
saturation. In clayey soils, structure plays an important role in hydraulic conductivity. Other major

3
factors that affect the permeability of clays are the ionic concentration and thickness of layers of water
held to the clay particles.

a) Effect of Size and Shape of Particles: Flow of gravitational water through minute pore
passages of soil is analogous to flow through capillary tubes. The flow through capillary tubes as
given by Poiseuille is directly proportional to the square of the tube diameter, and hence in soils,
flow should be proportional to the square of the average pore dimensions. Since it is not possible
to measure the dimensions of intricate pore passages in soil, attempts are made to correlate rate
of flow and permeability to the particle size which has the direct greatest effect on the pore
dimensions. It can be expected that permeability should be proportional to the square of some
average particle size in the soil. For fairly uniform sand, Hazen (1930) proposed an empirical

relationship for hydraulic conductivity in the form , where c = a constant

that varies from 1.0 to 1.5, and D10 = the effective size in mm.

The equation proposed by Kozeny – Carman relates permeability to particle size, void ratio (e),
angularity of particles, specific surface (S), and viscosity of water (ηw). the general equation is:

--- --- --- --- 8

For a collection of spherical particles uniformly distributed in size between diameters d 1 and d2,
the specific surface (i.e. the surface area per unit volume of grains), S is obtained from the

equation . If d1 and d2 are expressed in millimeters, S is expressed in mm 2/mm3, or

mm-1. The term is the geometric mean particle diameter. Values of f stated by Loudon
(1952) are as follows:

Rounded grains: f = 1.1

Sub-rounded grains: f = 1.25
Angular grains: f = 1.4
Equation 8 can be expressed in the customary SI units given below:

k = coefficient of permeability = m/s

g = acceleration due to gravity = 9.81 m/s2
γw = dynamic viscosity of water = 1 mPas
e = void ratio of soil
S = specific surface of grains (mm2/mm3) = mm-1
Substituting these values in equation 8 with the appropriate multiplying prefixes:

4
 m/s

m/s

b) Effect of Temperature: Permeability, k is not constant for a given soil but is related to the
dynamic viscosity of the fluid (water). Viscosity varies with temperature, and increases by about
30% from 200 C to 100 C. Therefore the water temperature should always be taken into account
when performing permeability tests, whether in the laboratory or in the field. Permeability varies
inversely with the viscosity of water (or any other fluid). It is usual to convert the permeability

results to a standard temperature (270 C) for comparison purpose. Thus . Where, kT

and ηT are the values of permeability and viscosity at test temperature respectively.

c) Effect of Void Ratio: Two approximate relationships which are valid only for coarse grained

soils are: , and . It has also been found that a semi logarithmic plot of void ratio

versus permeability is approximately a straight line both for coarse and fine grained soils. All
soils contain voids or pores and can be described as porous. However, to allow water to flow, at
least some of the voids must be continuous and then the soil can be described as permeable.
Permeability is not dependent on the amount of void present. The porosity of clay is usually
greater than that of sand but the clay could be a million times less permeable than the sand. Some
typical values are given below.

Typical - k
Soil Porosity
(m/s)
Clay 0.3 – 0.5 10-9
sand 0.2 – 0.4 10-3
This difference is due to the nature and size of the voids. In sand, particularly clean sands
without fines, most of the voids are continuous, more direct and relatively large, while in clays
the voids are smaller, more tortuous and with several stagnant voids.

1. On the basis of laboratory experiments, the U S Department of Navy (1971) provided an

empirical correlation between k and D10(mm) for granular soils with the uniformity
coefficient varying between 2 and 12 and D10 / D5 < 1.4. According to their experimental

5
 observations, Samarasinghe et. el., suggested that the hydraulic conductivity of normally
consolidated clays can be given by the following equation:

, where C3 anad n are constants to be determined experimentally.

d) Effect of Structural Arrangement of Particles and Stratification: Stratified soil masses have
marked variations in their permeability in directions parallel and perpendicular to stratification.
The permeability parallel to stratification being always greater. The horizontal permeability
parallel to the stratification is largely governed by the most pervious strata, whereas, the vertical
permeability perpendicular to stratification greatly depends upon the least pervious strata. At a
given void ratio the shape and arrangement of voids will vary with the structural arrangement of
soil particles and hence permeability will also vary with structural arrangement of soil particle.
The effect of structural disturbance on permeability is more pronounced in fine grained soils then
coarse grained soils.

e) Effect of Degree of Saturation and Other Foreign Matter: Darcy’s law and other relations for
determining the permeability is valid only if sample is fully saturated. Permeability is greatly
affected if air remains in pores of soil. Similar to air bubble, foreign mater, if present in water,
may plug the soil pores and reduces the permeability.

f) Effect of Adsorbed Water: The adsorbed water surrounding the fine soil particles is not free to
move and hence, it causes an obstruction to the flow of free water by reducing the effective pore
space available for the passage of water. According to Casagrande, 0.1 may be taken the void
ratio occupied by adsorbed water. Adsorbed water has a marked influence on the permeability of
clays.

LABORATORY DETERMINATION OF PERMEABILITY

Permeability of soil can be determined in the laboratory with the help of permeameters. Constant head
and variable (falling) head permeameter are the two general categories used for determination of
permeability. As the name imply, the seepage takes place in the constant head permeameter, through the
soil sample at a constant head during the experiment; and in the variable head permeameter, the head is
allowed to fall.

Constant Head Permeability Test

6
The constant hydraulic head causing the flow is the difference in elevations of water levels in the bottom
and overhead tanks. When steady state flow is reached, quantity of water flow for a desired time interval
is measured. From Darcy’s law:

Q/t = k i A

where Q = total flow quantity (ml), t = time interval (sec), l = length of the specimen (cm), A = cross
sectional area of specimen (cm2), and K = coefficient of permeability (cm/sec).

The constant head test is run on coarse grained soils which have a reasonable discharge to be measured.

Falling Head Permeability Test

The falling head permeability test is applicable for fine grained soils where the flow through specimen is
very limited. Figure shows the arrangement for falling head permeability test.

7
A stand pipe having cross sectional area a is attached to the top of the permeameter mould and water is
allowed to pass through the specimen. The water level in the stand pipe constantly falls with time and
observations are recorded after steady state flow condition. The required observations are change in
head of water with elapse of time.

The head at any time instant t is equal to the difference in the water level in the stand pipe and the
bottom tank. Let h1 and h2 are the water heads at time interval t1 and t2 respectively. Let h be the water
head at any intermediate time t, and –dh (minus sign has been used since h decreases with increase in t)
be the change in the water head in a small time interval dt. Hence, from Darcy’s law the rate of flow q is
given by:

, where i = hydraulic gradient at time t = h/l.

Therefore,

Or,

Integrating between two time limits, we get

8
or

(t2-t1) is time lapsed between two observations and denoted as t.

--- ---- -----

The laboratory observations are recording of water head h1 and h2 with reference to time t1 and t2.

Equivalent Hydraulic Conductivity in Stratified Soil

In a stratified deposit where the hydraulic conductivity for flow in a given direction changes from layer
to layer, an equivalent hydraulic conductivity can be computed to simplify calculations.

Figure shows n layers of soil with flow in the horizontal direction. Let us consider a cross section of unit
length passing through the n layer and perpendicular to the direction of flow. The flow through the cross
section in unit time can be written as:

… …1

Where, v = average discharge velocity

v1, v2, v3… vn = discharge velocities of flow in layers.

If kH1, kH2, kH3… kHn are the hydraulic conductivities of the individual layers in the horizontal direction,
and kH(eq) is the equivalent hydraulic conductivity in the horizontal direction, then from Darcy’s law,

v = kH(eq) ieq; v1 = kH1 i1; v2 = kH2 i2; v3 = kH3 i3; … vn = kHn in

Substitute the preceding relations for velocities in equation – 1.

kH(eq) ieq H = kH1 i1 H1 + kH2 i2 H2 + kH3 i3 H3 + ….+ kHn in Hn

But ieq = i1 = i2 = i3 = … = in

kH(eq) =

Figure shows n layers of soil with flow in the vertical direction. In this case, the velocity of flow through
all the layers is the same. However, the total head loss, h, is equal to the sum of the head losses in all
layers. Thus,

9
v1 = v2 = v3 = …= vn, and h = h1 + h2 + h3 +… + hn

But, h1 = i1 H1; h2 = i2 H2; h3 = i3 H3……; hn =in Hn

Therefore, h = i1 H1 + i2 H2 + i3 H3+…+ in Hn … … 2

Now, if kv(eq) = average permeability perpendicular to the bedding plane, we have

v = kv(eq) i = kv(eq) or h =

Also, i1 = ; i2 = ; i3 = ; ….. ; in =

Substituting these values in equation 2, we get

or --- --- ---3

CONSIDARATIONS FOR HYDRAULIC CONDUCTIVITY OF CLAYEY SOILS IN FIELD

COMPACTION

In some compaction work in clayey soils, the compaction must be done in a manner so that a certain
specified upper level of hydraulic conductivity of the soil is achieved. Examples of such works are
compaction of the core of an earth dam and installation of clay liners in solid waste disposal sites.

To prevent groundwater pollution from leachates generated from solid-waste disposal sites, the U. S.
Environmental Protection Agency (EPA) requires that clay liners have a hydraulic conductive of 10 -7
cm/sec or less. To achieve this value, the contractor must ensure that the soil meets the following
criteria:

1. The soil should have at least 20% fines (fine silt and clay-size particles).

2. The plasticity index (PI) should be greater than 10. Soils that have a PI greater than about 30 are
difficult to work with in the field.

3. The soil should not include more than 10% gravel-size particles.

4. The soil should not contain any particles or chunks of rock that are larger than 25 to 50 mm.

10
In many instances, the soil found at the construction site may be somewhat nonplastic. Such soil may be
blended with imported clay minerals like sodium bentonite (about 4 to 6%) to achieve the desired range
of hydraulic conductivity.

Moisture Content and Dry Unit Weight of Compaction

Figure shows the variation of the dry unit weight of compaction and hydraulic conductivity with the
molding moisture content of a clayey soil. From the figure, we can see that for similar dry unit weight, a
soil will have a lower hydraulic conductivity when it is compacted on the wet side of the optimum. In
addition, a heavy sheepsfoot roller can introduce larger shear strains during compaction that create a
more dispersed structure in the soil. This type of compacted soil will have an even lower hydraulic
conductivity. During compaction, the clods must be broken down mechanically to as small as possible.
A very heavy roller used for compaction helps to break them down.

Bonding between lifts: Bonding between successive lifts is also an important factor; otherwise,
permeant can move through a vertical crack in the compacted clay and then travel along the interface
between two lifts until it finds another crack, as is schematically shown in Figure below:

11
Bonding can substantially reduce the overall hydraulic conductivity of compacted clay.

Field Tests

Field tests to determine coefficient of permeability can be broadly grouped into three categories as
below:

1. Pumping test (Pump – out test)

2. Borehole test (Pump – in test)
3. Variable head test
1. Pumping Test: In this test water is pumped out of wells at a constant rate. When steady state
conditions are reached the data collected from field are used to determine k. Three cases of such
test are discussed below.

(a) Fully penetrating gravity well: Figure shows the arrangement for a fully penetrating gravity
well. Water is pumped from pumping well at a steady rate. The natural ground water level gets lowered
and the change in water level is observed in the observation wells.

12
 When steady state conditions are reached, data are collected from the pumping well and observation
wells. From the observations recorded, k can be computed as

k = , where q = constant rate of pumping, r 1 & r2 = radial distance of

observation wells 1 and 2 respectively from pumping well, h 1 & h2 = depth of water
above impervious stratum in observation wells 1 and 2 respectively.

k can be alternatively obtained from,

k= Equ. 1

where R = radius of influence, H = natural depth of ground water above impervious

stratum, rw = radius of well & hw = depth of water in pumping well above impervious
stratum after draw down.

According to Kozeny (1933), R can be expressed as, R = Equ. 2 where n

= soil porosity, t = time required to establish steady conditions.

Recommendations for values of R made by U. S. Department of Agriculture – Bureau of

Soils are shown in Table below:

Table: Values of Radius of Influence, R

Soil description Particle size R
(mm) (m)
Coarse gravel >10 >1500
Medium gravel 2 – 10 500 – 1500
Fine gravel 1–2 400 – 500
Coarse sand 0.5 – 1 200 – 400
Medium sand 0.25 – 0.5 100 – 200
Fine sand 0.1 – 0.25 50 – 100
Very fine sand 0.05 – 0.1 10 – 50
Silty sand 0.025 – 0.05 5 – 10

13
#4 A pump out test is carried out with a fully penetrating gravity well. Steady state conditions are
reached after 27 hours of pumping at a constant rate of 1320 l/min. The depth of ground water is 7.8 m.
The draw down water depth in the well of 30 cm diameter is 6.15 m. Void ratio of in-place soil is 0.45.
Compute the coefficient of permeability of the soil.

rw = 15 cm, n = = 0.31.

Since no observation well data have been provided k will be calculated using Equ. 1 & 2. Atrial and
error procedure is adopted as explain below.

Trial – 1. Assuming k = 10-3 cm/s, from Equ. 2

R= = 3155 cm

Substituting in eqn 1, k = = 1.63 X 10-1 cm/s

Due to large difference between assumed and computed values another trial is necessary.

Trial – 2 Assume k = 10-2cm/s, then

R= = 5610 cm, and

k= = 1.8 X 10-1 cm/s

Trial – 3 Assume k = 10-2 cm/s, then from calculations,

R = 11870 cm, and k = 2.03 X 10-1 cm/s O. K.

The coefficient of permeability of soil deposit is 0.2 cm/s. It can be noted that k is not very
sensitive to value of R. Hence, a reasonable value of R can be used in the calculations.

(b)Partially penetrating gravity well: A pump-out test from partially penetrating gravity well is shown
in Figure. k is expressed by the relationship (Mansur & Kaufman, 1962).

14
 , where m = , and s = depth below

bottom of well to impervious stratum.

(c)Fully penetrating artesian well: Artesian conditions can exist in a confined aquifer sandwiched
between two impervious layers. A well fully penetrating the pervious aquifer is shown in Figure.

When steady state conditions are reached at a constant pumping rate of q, coefficient of permeability
can be determined from the equation:

k= , where T = thickness of pervious aquifer and r2>r1

2. Borehole Tests:

15
Pump-out tests are expensive and time consuming. Borehole tests which are relatively economical and
easy to conduct have been recommended by USBR. There are two different types of borehole tests.
These are: (a) Open-end test and (b) Packer test. Borehole tests are carried out using standard casing
pipes. The dimensions of USBR standard casings are given in Table below:

Casing Inside diameter Outside diameter

designation inch (mm) inch (mm)
EX
(38.1) (46)
AX
(48.4) (57.2)
BX
(60.3) (73)
NX 3 (76.2)
(88.9)

(a) Open-end test:

Figure shows the different situations in open-end tests. For all the situations shown in the figure k can

be determined from the equation k = , where r = inside radius of casing, h = head of water = h p +

hgr, hp = pressure head over gravity head, hgr = gravity head and, q = rate of supply of water to maintain
head h.

(b) Packer test: In packer test (Figure) the rate of flow of water required to maintain a constant head over
the desired length of soil stratum is observed. The test can be carried out either while drilling of
borehole or after completion of borehole. Permeability k is determined from the following formulae,

k= , for L>10r

k= , for 10r > L > r

16
where q = constant rate of flow, h = head of water = hp + hgr, L = length of test section, r = radius of
borehole.

SEEPAGE AND FLOW NETS

Seepage Force

As water flows through the voids of a soil it transfers some of its energy to the soil particles, and a force
a force is applied by the flowing water which, in certain circumstances, can be detrimental to stability of
the soil and any structure on the soil. This seepage force (and seepage pressure) can be derived by
considering a block in a flow net bounded by two flow lines and two equipotential lines:

Water force on LHS = γw h1 A

Water force on RHS = γw h2 A

Area per unit width of section = Δl 1

Volume affected by seepage force = V = Δl2 1

Hydraulic gradient i = Δh / Δl

Force applied to sand particles = force on LHS – force on RHS

= γw h1 Δl 1 - γw h2 Δl 1
= γw Δh Δl 1
= γw (Δh/Δl) Δl2 1 = γw i V

seepage force J (units of kN) is then,

17
J = γw i V, and seepage pressure is γ w i, which is the seepage force per unit volume (kN/m 3). The term
seepage pressure will be misleading since it is not a pressure.

Critical Hydraulic Gradient

When the velocity of flow exceeds a certain critical value, turbulence occurs and Darcy’s law no longer
applies. If the flow is in downward direction, frictional drag of the water on the surfaces of the soil
particles tends to bring the particles closer together and increases the contact pressure between them. If
the water flows upwards, frictional drag tends to lift the particles and force them apart. If the seepage
pressure becomes equal to the pressure due to submerged weight of the soil, the effective pressure is
reduced to zero. In such cases, a cohesionless soil loses all its shear strength, and the soil particles have a
tendency to move up in the direction of flow. This phenomenon of lifting of soil particles is called quick
condition, boiling condition or quick sand. The hydraulic gradient under which the particles begin to
lose contact is known as the ‘critical hydraulic gradient’, ic, and is equal to the ratio of the submerged
density of the soil to the density of water.

Consider a block in a flow net at the soil surface on the exit side, such as inside a sheet pile cofferdam or
down stream of a concrete or earth dam. To the right of the sheet piling the seepage force J, acts in the
same direction as the gravity force, R, in an element of soil, so effective stresses and hence shear
strength are increased. However, to the left of the sheet piling where the seepage exits from the soil, the
upward seepage force, J, is acting against the downward gravity force S. This downward force acts
within and between the soil grains and represents the effective stresses in the element. Hence, S = γsub V

However the upward seepage force J, also acts on the soil grains thereby reducing the effective stresses
and shear strength.

The ‘quick’ condition occurs when S = J and the effective stresses are zero. Then the soil has no strength
at all, interlock between the grains is removed and the soil is in a quick or active fluid state. Small

18
upward seepage will be seen as localized boiling, carrying sand particles upwards which flow outwards
to form small volcano – like mounds.

There is a critical hydraulic gradient, ic, at which this condition occurs, when: S = J

γsub V = γw ic V

This critical hydraulic gradient will depend, therefore, on the particle density and particle packing. For
light weight particles in a loose conditions ic could be as low as 0.6 but for densely packed quartz grains
a value over 1.0 is likely.

Seepage Uplift of Cohesive Stratum

In case of cohesive soils, interparticles attractive forces produce a condition in which a mass of soil
rather than individual grains may be lifted due to upward head difference. Consider the example of
excavation into a clay deposit underlain by coarse sand within a cofferdam, as shown in Figure:

Water level inside the cofferdam is kept lowered at the excavation line by pumping out. The
permeability of clay is so low that there is hardly any seepage and there is no seepage forces consider.

Neglecting any side friction, the critical thickness Hc at which the clay will be on the verge of uplift
(safety factor F = 1) is obtained by equating the downward and upward pressures as the bottom plane of
the clay.

Hc γsat = hpγw = (h + hc)γw, where γsat = saturated unit weight of clay (may not always be fully saturated),
hp = piezometric head at the bottom of clay, h = head difference.

19
Rearranging above equation Hc = or Hc =

Laboratory Experiment for Quick Condition

When flow takes place in an upward direction, the seepage pressure also acts in the upward direction
and the effective pressure is reduced. If the seepage pressure becomes equal to the pressure due to
submerged weight of the soil, the effective pressure is reduced to zero. In such a case, a cohesionless
soil loses all its shear strength, and the soil particles have a tendency to move up in the direction of flow.
Thus, during the quick condition,

σ ‘ z γ’ – ps =0
Or, ps = z γ’ or i z γw = z γ’
From which i = ic =

Figure shows a set up to demonstrate the phenomenon of quick sand. Water flows in an upward
direction through a saturated soil sample of thickness z under a hydraulic head h. This head can be
increased or decreased by moving the supply tank in the upward or downward direction. When the soil
particles are in the state of critical equilibrium, the total upward force at the bottom of the soil becomes
equal to the total weight of all the materials above the surface considered.

Equating the upward and downward forces at the level a-a, we have, (h+z) γw = z γsat A

Therefore, h γw = z(γsat – γw) = zγ’

Therefore,

20
Flow Nets

A flow net consists of two sets of lines: flow line and equipotential line. A flow net is usually used to
represent the steady state condition. For example, on impounding a reservoir behind an earth dam, the
soil voids must first become saturated before a steady state flow through the dam can develop.

Flow lines: These are paths along which water can flow through a cross section. There are infinite
numbers of flow lines available but only a few needs to be selected for an adequate flow net. The space
between adjacent flow lines is called flow channels. Each flow channel has an equal flow quantity, Δq,
so the total seepage flow is given by Δq multiplied by the number of flow channels. The number of flow
channels need not be a whole number. This can be useful when plotting a flow net up to an impermeable
boundary when only a part flow channel is left. The seepage quantity passing through this remaining
part channel is proportional to its width relative to a full channel.

Equipotential lines: These are lines of equal energy level or equal total head. As the water flows
through the pore spaces its energy is dissipated by friction and the equipotential lines act like contours to

21
show how the energy is lost. The interval between adjacent equipotential represent a constant difference
in total head loss, Δh, and the total head, h, lost around the structure is shared equally among
equipotential drops.

Flow Net Construction

There are numbers of procedure available by which an engineer may draw a flow net for the solution of
a seepage problem. Among these are:

1. Mathematical solution.

2. Electrical analogy.

3. Trial sketching.

4. Models, including finite element and finite difference solution.

Mathematical Solution - (Laplace equation)

The basis for flow net theory is the principle of continuity – that the volume of water flowing into a
given volume of saturated soil must equal the volume flowing out.

Figure …. Shows a soil element of size Δx, Δy and Δz. Considering two dimensional flow, water is
assumed to flow only in the plane of the x and z axis. If vx and vz are the discharge velocity of flow at the

22
entry to the element, the exit velocities will be , and respectively. It is assumed

that the water is incompressible fluid, soil is fully saturated and the seepage force does not alter any
physical state of the soil element. The quantity of flow that enters the soil element per unit time is equal
to the water exit the soil element:

vx Δz Δy + vz Δx Δy =

or

the above equation is called continuity equation for two dimensional flow in xz plane. The hydraulic

gradient components for the element , and can be inserted in Darcy’s law to give:

; and , where the total head, h, is decreasing in the direction of vx and

vz.

Differentiating above equation, we get

, and

Substituting the value in equation….

or

Above equation is known as Laplace equation for two dimensional flows.

Mathematically the equation states that the algebric sum of the gradient changes in the x and z directions
is zero. Graphically it represents two sets of curves intersecting at right angles.

Electrical analogy

The Darcy’s law and the Ohm’s law governing the flow of electricity are comparable to each other. The
comparison between two is as given in Table hereunder:

23
 Flow of water Flow of electricity

Darcy’s law: Ohm’s law:

q = rate of flow of water I = rate of flow of current
h = drop in head E = potential drop
L = length of path of percolation l = length of path of current
A = cross sectional area through a = cross sectional area of
which flow occurs. conductor
k = coefficient of permeability K = coefficient of conductivity.
Thus, an electrical model can be prepared whose boundary conditions are similar to those which govern
the flow of water through soil. The equipotential lines are drawn by joining the equal voltage points.

The electrical analogy is a shallow a shallow tray with flat bottom constructed of an insulating material
and water. A small quantity of salt or hydrochloric acid solution is added to make water a conducting
media. The hydraulic boundaries are simulated on the tray. The boundary equipotential lines are made of
copper strips, and boundary flow lines are simulated by non-conducting strips such as perspex, ebonite
or other such material. The non-conducting strip is fixed in the tray by means of some non-conducting
adhesive, such as plasticene, bee wax, etc.

An alternating current in the range of 20V is applied across the boundary equipotential strips. A
potentiometer is also connected in parallel with the current source. To plot the potential distribution
within the tray, the particular potential point on the potentiometer is connected through a null indicator
system, such as galvanometer, CRO etc., to a probing point which is moved within the conducting water

24
of the tray. When the balance is obtained as indicated by the null indicator, the probing point in the tray
and the contact point at the potentiometer have the same potential. The coordinates of the grid points can
be read on a graph paper kept under the transparent bottom sheet of the model tray. By moving the probe
at various grids points and obtain the null points at different positions, an equipotential line can be
obtained. Thus, corresponding to different potential points on the potentiometer, various equipotential
lines of the flow net are traced out. Orthogonal flow lines conforming to the boundary conditions are
then drawn.

Trial sketching

In the method of trial sketching, the first step is to draw to a convenient scale a cross section of the soil
structure at a point along the axis normal to the direction of seepage flow. Next, it is necessary to
establish the two boundary flow lines and two boundary equipotential lines. Then, by trial, sketch the
intermediate flow lines and equipotential lines, adhering rigidly to right angle intersections and
essentially square figures. Locate the unknown lines as well as possible, and continue sketching until an
inconsistency develops. Each consistency will indicate the direction and magnitude of change for a
second trial. Successive trials must be made until the net is reasonably consistent throughout. From three
to five flow lines will generally suffice. Keep in mind that every transition in a homogeneous soil is
smooth and either elliptical or parabolic in shape. The size of the squares in each flow nets constructed
by experts and should try to develop a sense of fitness and intuition to assist in trial sketching.

25
A certain skill is required in drawing a flow net but adequate results can quickly be obtained providing
the following rules are observed:

 Flow lines and equipotential lines must cross at right angles.

 The areas formed by intersecting flow lines and equipotential lines must be as near square as
possible.

 Impermeable boundaries shall be identified. These are flow lines such as the surface of a clay
layer, the vertical surface of sheet piling or the underside of a concrete dam.

 Permeable boundaries shall be identified. Where a permeable soil boundary is in contact with
open water as on the upstream face of an earth dam, this boundary will be an equipotential, i.e.
the total head is constant on this boundary. This also applies to a horizontal water table within a
permeable deposit when flow is occurring vertically down wards, as alongside a sheet piled
excavation.

When water flows across a boundary between soils of different permeabilities the flow lines bend, the
flow channel width alters and the distance between equipotentials changes so that the blocks become
rectangular. The quantity of flow in both deposits must be the same. When water flows from a soil of
high permeability to one of low permeability A and i must increase so the flow channel width increases
and the distance, l, between the equipotentials decreases.

Seepage Quantity

These can be determined from flow nets. Flow net is assumed to be two dimensional so a unit width of
the cross section is considered. The total flow around a structure will then depend on its overall length.
Since the blocks on a flow net are square the width of a flow channel will be equal to its length Δl.

Therefore, area A = 1 Δl

Total flow q = Δq in each channel X number of channels

= Δq X nf

Head loss between equipotentials = Δh, and Δh

26
Hydraulic gradient for any block = i =

The flow in each channel = Δq = A k i

= =

The total flow is then given by q =

Although drawing square elements for a flow net is convenient, it is not always necessary. Alternatively,
one can draw a rectangular mesh for a flow channel, as shown in Figure below, provided that the width
to length ratios for all the rectangular elements in the flow net are the same. In this case equation for rate
of flow through the channel can be modified to:

If b1/l1 = b2/l2 = b3/l3 = …. = n (i.e., the elements are not square), and can be modified:

or

27
Above figure shows a flow net for seepage around a single row of sheet piles. Note that flow channels 1
and 2 have square elements. Hence, the rate of flow through these two channels can be obtained from
equation:

However, flow channel 3 has rectangular elements. These elements have a width to length ratio of about

0.38; hence:

So, the total rate of seepage can be given as:

The rate of seepage per unit length in anisotropic soil can be calculated by modifying equation as:

, where H is total head loss, and Nf and Nd = number of flow channels and potential

drops, respectively.

Seepage pressure

The hydraulic potential h at any point located after n potential drops, each of value Δh is given by:

h = H – nΔh.

Thr seepage pressure at any point equals the hydraulic potential multiplied by the unit weight of water.

ps = h γw = (H – nΔh) γw

Exit Gradient

The exit gradient is the hydraulic gradient at the down stream end of the flow line where the percolating
water leaves the soil mass and emerges into the free water at the down stream. The exit gradient can be
calculated from the following expression, in which Δh represents the potential drop and l the average

length of the last field in the flow net at exit end: ie =

GRADED SOIL FILTERS

As per Sherard, et el

28
 Soil
Group Soil Description D15 Design Criteria
No
Fine silts and clays with >85% passing
1 D15 ≤ 9d85, but not smaller than 0.2 mm
#200 sieve
Silty and clayey sands and sandy silts and
2 clays with 40 – 85% passing the #200 D15 ≤ 0.7 mm
sieve
Silty and clayey sands and gravels with
3 D15 ≤ (4 d85 – 0.7 mm) + 0.7 mm
15 – 39% passing the #200 sieve
Silty and clayey sands and gravelly sands
4 D15 ≤ 4 d85
with ≤ 15% passing the #200 sieve
# The design of a proposed earth dam is to include a gravel drain to control the ground water inside
the dam. This drain is to be protected with a graded soil filter zone. The grain size distribution curves of
the soils used to build the core of the dam and the gravel used to build the drain are shown in Figure .
Determine the range of acceptable filter material.

Solution: Migration of core soils into filter: 91% passing #200 sieve, so this is soil type 1 as per above
table.

D85 = 0.056 mm

D15 ≤ 9 d85 → D15 ≤ (9) (0.056 mm) → D15 ≤ 0.50 mm

Therefore, the filter soils must have D15 ≤ 0.50 mm, as shown by the mark on Figure.

Migration of filter soils in to drain:

29
D15 = 4.0 mm

D15 ≤ 4 d85 → 4.00 mm ≤ (4) (d85) → 1.0 ≤ d85

Therefore, the filter soils must have D85 ≥ 1.0 mm, as shown by the mark on figure.

Based on these two marks, the range of acceptable filter material is as shown in figure.

OTHER CRITERIA

1. The filter material should be coarse enough so that the percolating water moves easily with out
any build up of water pressure in the filter. D15 (f) > 5 D15(b)

2. The filter material should be fine enough that the soil particles of the base material are not

washed through the filter. D15(filter material) < D85(base material). In a graded filter, each

layer is designed considering it as a filter and the preceding layer as a base material. The particle
size of the layers increases in the direction of flow.

3. For circular holes in the pipe: (D85 of filter material) / Diameter of the hole > 1.2.

4. For slotted openings: (D85 of filter material) / (Width of slot) > 1.4

# Determine the approximate limits of the filter material required for the soil which has D 15 = 0.01
mm and D85 = 0.1 mm, and grading curve as shown in figure.

From equation, D15(f) > 5 D15(b)

D15(f) > 5 X 0.01 > 0.05 mm

From equation, D15(f) < 5 D85(b)

D15(f) < 5 X 0.10 < 0.5 mm

Therefore, D15 of filter should lie between 0.05 mm and 0.5 mm.

As the gradation curve of the filter material should be roughly parallel to that of the base material, the
hatched portion indicates the limits of the material suitable as filter.

30
GEOSYNTHETICS FILTERS

EOS = Equivalent opening size OR AOS = Apparent opening size, which is expressed either as O 95 or as
equivalent sieve size. For example, a geotextile with EOS = AOS #30 (O 95 = 0.60 mm) has openings
similar to those in a #30 sieve. Such a fabric would retain 95% of soil particles that have a diameter of
0.60 mm.

Carroll (1983) recommends selecting geotextile for filtration based the following criterion: O95 < (2 or 3)
(D85)soil.

The geotextile filter fabric also must have sufficient permittivity to pass the required groundwater flow

rate. This parameter is measure of the fabric’s ability to pass water, and is defined as: ψ = ,

combining with Darcy’s Law ψ = ,

Where: ψ = permittivity of the geotextile, kn = hydraulic conductivity for flows normal to the fabric
face, t = thickness of the geotextile, Q = flow rate normal to fabric, and Δh = head loss through fabric.

The minimum required permittivity may be determined by assigning an allowable head loss and using
above equation.

# A sandy silt with D85 = 0.10 mm is to be drained by a perforated drainage pipe surrounded by a
¾ inch gravel and a filter fabric as shown in figure. The estimated flow rate to this drain will be 5
gal/min per lineal foot (half of which is from each side), and the head loss through the fabric must not
exceed 0.1 ft. Select an appropriate filter fabric.

Solution:

Check maximum allowable apparent opening size (AOS).

31
O95 < (2 or 3) (D85)soil

< (2 or 3) (0.10 mm)

< 0.2 – 0.3 mm

Check minimum permittivity requirement.

Q = (5 gal/min) = 0.01 ft3/s

A = (0.5 ft + 1.0 ft + 0.5 ft) (1.0 ft) = 2.0 ft2

Q/A = (0.01 ft3/s) / (2.0 ft2) = 0.005 ft/s

Minimum required ψ = (Q/A) / Δh = (0.005 ft/s) / 0.1 ft = 0.05 s-1

Occasionally, geotextile drains may be subject to soil clogging, which blocks the pores and prevents or
severely restricts the flow of water. Although this is usually not a problem, the following conditions
have been found to cause clogging (Koerner, 1998):

1. cohesionless sands and silts

2. gap – graded soils, and

3. high hydraulic gradients

When all three of these conditions are present, it is better to use graded soil filters.

Clogging also can occur in soils where chemical precipitates tend to build up on the fabric, or when
the ground water contains a large concentration of biological matter (i.e., municipal landfill
leachate).

32
REFERENCES

Barnes G. E. (1995), “Soil Mechanics – Principles and practice”, Macmillan Press Ltd. London.
Carman, P. S. (1939) “Permeability of saturated sands, soils and clays”, Jr. Agric. Sci. Vol. XXIX, No.
11.
Darcy, H. (1856), “Les fontaines publique de la ville de Dijon”. Dalmont, Paris.
Das, B. M (1997), “Principles of Geotechnical Engineering”, PWS publishing company, Boston.
Fancher, G. H., Lewis J. A. and Barnes K. B. (1933) “Some physical characteristics of oil sands”,
Mineral Industries Experiment Station, Bulletin 12, Pennsylvania State Collage, USA.
Hazen A (1892, 1895), “Some physical properties of sands and gravels with special reference to their
use in filtration”. 24th Annual Report, Massachusetts State Board of Health, Also, The filtration of Public
Water Supplies.
Head, K. H. (1982), “Manual of Soil Laboratory testing”, Pentech Press, London.
Kozeny, J. (1927), “Uber kapillare leitung des wassers in boden”, Ber Wien Akad., 136a-271.
Lambe, T. W. and Whitman, R. V. (1979), “Soil Mechanics”, Wiley, New York.
Loudon, A. G. (1952), “The computation of permeability from simple soil tests”, Geotechnique, Vol. 3,
No. 4.
Peck, R. B., Hanson, W. E. and Thornburn, T. H. (1973), “Foundation Engineering”, Second edition.J.
Wiley, New York.
Prinz, E (1923) “Handbuch der hydrologie”, Berlin
Reynolds, O. (1883) “An experimental investigation of the circumstances which determine whether the
motion of water shall be direct or sinuous and the law of resistance in parallel channels”, Trans. Royal
Soc., Vol. 174, London.
Sherard, J. L., Dunnigan, L. P. and Decker, R. S. (1976), “Identification and nature of dispersive soils”,
J. Geo. Eng. Div., ASCE, Paper 12052, April.
Sherard, J. L., Dunnigan, L. P. and Decker, R. S. (1976), “Pinhole test for identifying dispersive soils”,
J. Geo. Eng. Div., ASCE, Paper 11846, January.
Singh, A. and Chowdhary, G. R. (1994), “Soil engineering in theory and practice”, CBS Publishers and
Distributors, New Delhi.
Taylor, D. W. (1948), “Fundamentals of soil mechanics”, Chapman & Hall, London.
Terzag γsat hi, K and Peck, R. B. (1967), “Soil mechanics in engineering practice”, Wiley, New York.

33