Permeability

7.1 Bernoulli’s Equation

From fluid mechanics, we know that, according to Bernoulli’s equation, the total head at a point in
water under motion can be given by the sum of the pressure, velocity, and elevation heads, or

Note that the elevation head, Z, is the vertical distance of a given point above or below a
datum plane. The pressure head is the water pressure, u, at that point divided by the unit
weight of water, gw.

If Bernoulli’s equation is applied to the flow of water through a porous soil medium,
the term containing the velocity head can be neglected because the seepage velocity is
small, and the total head at any point can be adequately represented by

The loss of head between two points, A and B, can be given by

The head loss, Δh, can be expressed in a nondimensional form as

KETERANGAN

i = hydraulic gradient

L = distance between points A and B—that is, the length of flow over which the loss of head occurred

In general, the variation of the velocity v with the hydraulic gradient i is as shown
in Figure 7.2. This figure is divided into three zones:
1. Laminar flow zone (Zone I)
2. Transition zone (Zone II)
3. Turbulent flow zone (Zone III)
In most soils, the flow of water through the void spaces can be considered laminar;
thus,

7.2 Darcy’s Law

In 1856, Darcy published a simple equation for the discharge velocity of water through
saturated soils, which may be expressed as

v = discharge velocity, which is the quantity of water flowing in unit time through a unit gross cross-
sectional area of soil at right angles to the direction of flow.

k = hydraulic conductivity ( otherwise known as the coefficient of permeability )

This equation was based primarily on Darcy’s observations about the flow of water
through clean sands. both are valid for laminar
flow conditions and applicable for a wide range of soils. v is the discharge velocity ofwater based on
the gross cross-sectional area of the soil. However, the actual velocity of water (that is, the seepage
velocity) through the void spaces is greater than v. A relationship between the discharge velocity and
the seepage velocity can be derived by referring to Figure 7.3, which shows a soil of length L with
a gross cross-sectional area A. If the quantity of water flowing through the soil in unit time
is q, then

Vv = volume of voids in the specimen

Vs = volume of soil solids in the specimen

e = void ratio

n = porosity

Darcy’s law as defined by Eq. (7.6) implies that the discharge velocity v bears a
linear relationship to the hydraulic gradient i and passes through the origin as shown in
Figure 7.4. Hansbo (1960), however, reported the test results for four undisturbed

natural clays. On the basis of his results, a hydraulic gradient i’ (see Figure 7.4) appears
to exist, at which

7.3 Hydraulic Conductivity

The hydraulic conductivity of soils depends on several factors: fluid viscosity, poresize
distribution, grain-size distribution, void ratio, roughness of mineral particles, and degree of soil
saturation. In clayey soils, structure plays an important role in hydrauli conductivity. Other major
factors that affect the permeability of clays are the ionic concentration and the thickness of layers of
water held to the clay particles.
The value of hydraulic conductivity (k) varies widely for different soils. Some typical values for
saturated soils are given in Table 7.1. The hydraulic conductivity of unsaturated soils is lower and
increases rapidly with the degree of saturation.
The hydraulic conductivity of a soil is also related to the properties of the fluid flowing through it by the
equation

ˠw = unit weight of water

η = viscosity of water

K = absolute permeability

It is conventional to express the value of k at a temperature of 20_C.Within the range of test

temperatures, we can assume that
7.4 Laboratory Determination of Hydraulic
Conductivity
Constant-Head Test
A typical arrangement of the constant-head permeability test is shown in Figure 7.5. In
this type of laboratory setup, the water supply at the inlet is adjusted in such a way that
the difference of head between the inlet and the outlet remains constant during the test
period. After a constant flow rate is established, water is collected in a graduated flask for
a known duration.
The total volume of water collected may be expressed as

Q = volume of water collected

A = area of cross section of the soil specimen

t = duration of water collection

Falling-Head Test
A typical arrangement of the falling-head permeability test is shown in Figure 7.6. Water
from a standpipe flows through the soil. The initial head difference h1 at time t = 0 is
recorded, and water is allowed to flow through the soil specimen such that the final head
difference at time t = t2 is h2.
with limits of time from 0 to t and the right side with limits of head difference from h1 to h2 gives

7.5 Relationships for Hydraulic Conductivity—

Granular Soil
For fairly uniform sand (that is, sand with a small uniformity coefficient), Hazen (1930) proposed
an empirical relationship for hydraulic conductivity in the form

c = a constant that varies from 1.0 to 1.5

D10 = the effective size, in mm

Over the last several years, experimental observations have shown that the magnitude
of c for various types of granular soils may vary by three orders of magnitude (Carrier,
2003) and, hence, is not very reliable.
Another form of equation that gives fairly good results in estimating the hydraulic
conductivity of sandy soils is based on the Kozeny-Carman equation (Kozeny, 1927;
Carman, 1938, 1956). The derivation of this equation is not presented here. Interested
readers are referred to any advanced soil mechanics book. According to the Kozeny-
Carman equation

The magnitude of SF may vary between 6 to 8, depending on the angularity of the soil
particles. Carrier (2003) further suggested a slight modification to Eq. (7.29), which can be
written as

assume that laminar flow condition does exist. More recently, Chapuis (2004) proposed an empirical
relationship for k in conjunction

D10 = effective size (mm).

The preceding equation is valid for natural, uniform sand and gravel to predict k that is in the range of
10-1 to 10-3 cm/s. This can be extended to natural, silty sands without plasticity. It is not valid for
crushed materials or silty soils with some plasticity. Based on laboratory experimental results, Amer
and Awad (1974) proposed the following relationship for k in granular soil:
k is in cm/sec

Cu = uniformity coefficient

D10 = effective size (mm)

ρw = density of water (g/cm 3)
η = viscosity (g s/cm2)

Mention was made at the end of Section 7.1 that turbulent flow conditions may exist
in very coarse sands and gravels and that Darcy’s law may not be valid for these materials.
However, under a low hydraulic gradient, laminar flow conditions usually exist. Kenney, Lau,
and Ofoegbu (1984) conducted laboratory tests on granular soils in which the particle sizes
in various specimens ranged from 0.074 to 25.4 mm. The uniformity coefficients, Cu, of
these specimens ranged from 1.04 to 12. All permeability tests were conducted at a relative
density of 80% or more. These tests showed that for laminar flow conditions,

D5 = diameter (mm) through which 5% of soil passes.

On the basis of laboratory experiments, the U.S. Department of Navy (1971) provided an
empirical correlation between k (ft/min) and D10 (mm) for granular soils with the uniformity
coefficient varying between 2 and 12 andD10 /D5 < 1.4.

7.6 Relationships for Hydraulic Conductivity— Cohesive Soils

Taylor (1948) proposed a linear relationship between the logarithm of k and the void
ratio as

ko = in situ hydraulic conductivity at a void ratio eo

k = hydraulic conductivity at a void ratio e
Ck = hydraulic conductivity change index

For a wide range of void ratio, Mesri and Olson (1971) suggested the use of a linear relationship
between log k and log e in the form

Samarasinghe, et al. (1982) conducted laboratory tests on New Liskeard clay and proposed that, for
normally consolidated clays,

where C and n are constants to be determined experimentally.

 Tavenas, et al. (1983) also gave a correlation between the void ratio and the hydraulic
conductivity of clayey soil. This correlation is shown in Figure 7.13. An important point to note,
however, is that in Figure 7.13, PI, the plasticity index, and CF, the clay-size fraction in the soil, are in
fraction (decimal) form. One should keep in mind, however, that any empirical relationship of this type
is for estimation only, because the magnitude of k is a highly variable parameter and depends on
several factors.

7.7 Directional Variation of Permeability

Most soils are not isotropic with respect to permeability. In a given soil deposit, the magnitude
of k changes with respect to the direction of flow. Figure 7.14 shows a soil layer
through which water flows in a direction inclined at an angle α with the vertical. Let the
hydraulic conductivity in the vertical (α = 0) and horizontal (α = 90˚) directions be kV
and kH, respectively. The magnitudes of kV and kH in a given soil depend on several
factors, including the method of deposition in the field.
Figure 7.15 shows the laboratory test results obtained by Fukushima and Ishii
(1986) related to kV and kH for compacted Masa-do soil (weathered granite). The soil

specimens were initially compacted at a certain moisture content, and the hydraulic conductivity was
determined at 100% saturation. Note that, for any given molding moisture content and confining
pressure, kH is larger than kV.
There are several published results for fine-grained soils that show the ratio of kH/kV
may vary over a wide range. Table 7.3 provided a summary of some of those studies.

7.8 Equivalent Hydraulic Conductivity in Stratified Soil

In a stratified soil 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. The following derivations relate to the equivalent hydraulic conductivities for flow
in vertical and horizontal directions through multilayered soils with horizontal stratification.
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 total flow
through the cross section in unit time can be written as
An excellent example of naturally deposited layered soil is varved soil, which is a rhythmically layered
sediment of coarse and fine minerals. Varved soils result from annual seasonal fluctuation of
sediment conditions in glacial lakes. Figure 7.18 shows the variation of moisture content and grain-
size distribution in New Liskeard, Canada, varved soil. Each varve is about 41 to 51 mm (1.6 to 2.0
in.) thick and consists of two homogeneous layers of soil—one coarse and onefine—with a transition
layer between.
7.9 Permeability Test in the Field
by Pumping from Wells
In the field, the average hydraulic conductivity of a soil deposit in the direction of flow can be
determined by performing pumping tests from wells. Figure 7.21 shows a case where the top
permeable layer, whose hydraulic conductivity has to be determined, is unconfined and underlain by
an impermeable layer. During the test, water is pumped out at a constant rate from a test well that has
a perforated casing. Several observation wells at various radial distances are made around the test
well. Continuous observations of the water level in the test well and in the observation wells are made
after the start of pumping, until a steady state is reached. The steady state is established when the
water level in the test and observation wells becomes constant. The expression for the rate of flow of
groundwater into the well, which is equal to the rate of discharge from pumping, can be written as

The average hydraulic conductivity for a confined aquifer can also be determined by
conducting a pumping test from a well with a perforated casing that penetrates the full
depth of the aquifer and by observing the piezometric level in a number of observation wells
at various radial distances (Figure 7.22). Pumping is continued at a uniform rate q until a
steady state is reached.
This gives the hydraulic conductivity in the direction of flow as
7.10 In Situ Hydraulic Conductivity
of Compacted Clay Soils
Daniel (1989) provided an excellent review of nine methods to estimate the in situ hydraulic
conductivity of compacted clay layers. Three of these methods are described.

Boutwell Permeameter
A schematic diagram of the Boutwell permeameter is shown in Figure 7.23. A hole is first
drilled and a casing is placed in it (Figure 7.23a). The casing is filled with water and a
falling-head permeability test is conducted. Based on the test results, the hydraulic conductivity
k1 is calculated as

where d = diameter of the standpipe

D = diameter of the casing

h1 = head at time t1

h2 = head at time t2
Constant-Head Borehole Permeameter

In this arrangement a constant head h is maintained by supplying water, and the rate of flow q is
measured. The hydraulic conductivity can be calculated as

Typical values of α range from 0.002 to 0.01 cm -1 for fine-grained soil.

Porous Probes
Porous probes (Figure 7.26) are pushed or driven into the soil. Constant- or falling-head permeability
tests are performed. The hydraulic conductivity is calculated as follows :
The constant head is given by

The falling head is given by

For probes with permeable bases (Figure a),

For probes with impermeable bases (Figure b),

7.11 Summary and General Comments

In this chapter, we discussed Darcy’s law, the definition of hydraulic conductivity, laboratory
determinations of hydraulic conductivity and the empirical relations for it, and field
determinations of hydraulic conductivity of various types of soil. Hydraulic conductivity
of various soil layers is highly variable. The empirical relations for hydraulic conductivity
should be used as a general guide for all practical considerations. The accuracy of the values
of k determined in the laboratory depends on several factors:
1. Temperature of the fluid
2. Viscosity of the fluid
3. Trapped air bubbles present in the soil specimen
4. Degree of saturation of the soil specimen
5. Migration of fines during testing
6. Duplication of field conditions in the laboratory
The hydraulic conductivity of saturated cohesive soils also can be determined by laboratory
consolidation tests. The actual value of the hydraulic conductivity in the field also may be somewhat
different than that obtained in the laboratory because of the nonhomogeneity of the soil. Hence,
proper care should be taken in assessing the order of the magnitude of k for all design considerations.