NPTEL- Advanced Geotechnical Engineering

Module 2 Lecture 7 Permeability and Seepage -3 Topics

1.1.6 Determination of Coefficient of Permeability in the Field
Pumping from wells Packer test

1.1.7 Theoretical Solution for Coefficient of Permeability 1.1.8 Variation of Permeability with Void Ratio in Sand

1.1.6 Determination of Coefficient of Permeability in the Field

It is sometimes difficult to obtain undisturbed soil specimens from the field. For large construction projects, it is advisable to conduct permeability tests in situ and compare the results with those obtained in the laboratory. Several techniques are presently available for determination of the coefficient of permeability in the field, such as pumping from wells, borehole tests, etc., and some of these methods will be treated briefly in this section.

Pumping from wells: Gravity wells. Figure 2.12 shows a permeable layer underlain by an

impermeable stratum. The coefficient of permeability of the top permeable layer can be determined by pumping from a well at a constant rate and observing the steady state water table in nearby observation wells. The steady state is established when the water level in the test well and the observation wells become constant. At the steady state, the rate of discharge due to pumping can be expressed as

From Figure 2.12 is approximately equal to (this is referred to as Dupuits assumption), and . Substituting these in the above equation for rate of discharge, we get

Dept. of Civil Engg. Indian Institute of Technology, Kanpur,

NPTEL- Advanced Geotechnical Engineering

Figure 2.12 Determination of coefficient of permeability by pumping from wells-gravity well

So,

(1.33)

If the values of are known from field measurements, the coefficient of permeability can be calculated from the simple relation given in equation. (1.23) According to Kozeny (1933), the maximum radius of influences, R (Figure 2.12), for drawdown due to pumping can be given by

(1.34) Where

Also note that if we substitute

then (1.35)

Where H is the depth of the original groundwater table from the impermeable layer. The depth h at any distance r from the well thus

Dept. of Civil Engg. Indian Institute of Technology, Kanpur,

NPTEL- Advanced Geotechnical Engineering

Or

(1.36)

It must be pointed out that Dupuits assumption (i.e., that does introduces large errors regarding the actual phreatic line near the wells during steady state pumping. This is shown in Figure 2.12. For greater than to , the phreatic line predicted by equation (1.36) will coincide with the actual phreatic line. The relation for the coefficient of permeability given is equation (1.33) has been developed on the assumption that the well fully penetrates the permeable layer. If the well partially penetrates the permeable layer as shown in Figure 2.13, the coefficient of permeability can be better represented by the following relation (Mansur and Kaufman 1962): (1.37) The notations used in the right-hand side of equation (1.37) are shown in Figure 2.13.

Figure 2.13 Pumping from partially penetrating gravity wells

Artesian wells. The coefficient of permeability for a confined aquifier can also be determined from well pumping tests. Figure 2.14 shows an artesian well penetrating the full depth of an aquifier from which water is pumped out at a constant rate. Pumping is continued until a steady state is reached. The rate of water pumped out at steady is given be (1.38) Where T is the thickness of the confined aquifier, or (1.39) Solution of equation (1.39) gives (1.40) Hence, the coefficient of permeability k can be determined by observing the drawdown in two observation wells, as shown in Figure 2.14.
Dept. of Civil Engg. Indian Institute of Technology, Kanpur, 3

NPTEL- Advanced Geotechnical Engineering

Figure 2.14 Determination of coefficient of permeability by pumping from wells-confined aquifier case

If we substitute

we get (1.41)

Bore hole test recommended by USBR. Well pumping tests are costly, and so for economic reasons bore holes are used in many cases to estimate the coefficient of permeability of soils and rocks (U. S. Bureau of Reclamation, 1961). Open-end test. Figure 2.15 shows the schematic diagram for determination of the permeability of soils by means of the open-end test. Casings are inserted in the bore holes, and they extend to the soil layers whose permeability needs to be determined. The groundwater table can be above or below the bottom of the casings.

Figure 2.15 Open-end test for soil permeability in the field. (Redrawn after U. S. Bureau of Reclamation, 1961, by permission of the United States Department of the Interior, Water and Power Resources Service.)
Dept. of Civil Engg. Indian Institute of Technology, Kanpur, 4

NPTEL- Advanced Geotechnical Engineering

Before conducting the permeability test, the hole is cleaned out. The test begins by adding clear water to maintain gravity flow at a constant head. The coefficient of permeability can be determined by the equation (1.42) Where

To ensure that equation (1.42) is used correctly, the following points should be noted: In Figure 2.15a and b, the constant water level maintained inside the casing coincides with the top of the casing. In Figure 2.15a, h is measured up to the bottom of the casing. In soils that have low permeability, it may be desirable to apply some pressure to the water. This is the case in Figure 2.15c and d. for these cases, the value of h used in equation (1.42) is given by (1.43) Any consistent set of units may be used in equation (1.42). If q is expressed in gal/min, then (1.44) The values of for various sizes of standard casings used for field exploration are given in table 1.3. The inside and outside diameters of these casings are given in table 1.4. Table 1.3 Values of for various sizes of casings Size of casings EX AX BX NX After U. S. Bureau of Reclamation (1961)

204,000 160,000 129,000 102,000

Packer test : The arrangement for carrying out permeability tests in a bore hole below the casing is

shown in Figure 2.16. These tests can be made above or below the ground water table. However, the soils should be such that the bore hole will stay open without any casing. The test procedure is illustrated in Figure 2.16. The equations for calculation of the coefficient of permeability are as follows:

Dept. of Civil Engg. Indian Institute of Technology, Kanpur,

NPTEL- Advanced Geotechnical Engineering

Figure 2.16 The packer test for soil permeability. (Redrawn after U. S. Bureau of Reclamation, 1961, by permission of the United States Department of the Interior, Water and Power Resources Service.)

(1.45) And Where Constant rate of flow into bore hole Length of test hole Radius of bore hole Differential head of water For convenience, (1.47) The values of are given in table 1.5. (1.46)

Table 1.4 diameters of standard casings Casing EX AX BX

Inside diameter, in

Outside diameter, in

Dept. of Civil Engg. Indian Institute of Technology, Kanpur,

NPTEL- Advanced Geotechnical Engineering

NX Note: 1 in = 25.4 mm.

Variable-head tests by means of piezometer observation wells. The U. S. Department of the Navy (1971) has adopted some standard variable-head tests for determination of the coefficient of permeability by means of piezometer observation wells. These methods are described in table 1.6. Careful attention should be paid to the notations. Figure 2.17 gives the shape factor coefficient S used for condition (A). Figure 2.18 gives the shape factor coefficient used for condition (F-1). Figure 2.19 shows the analysis of permeability by variable-head tests with reference to table 1.6. Table 1.6 computation of permeability from variable head tests (for observation well of constant cross section

Example 1.1 Refer to Figure 2.20 for the pumping test from a gravity well. If and the pumping rate from the well is determine the coefficient of permeability .

Dept. of Civil Engg. Indian Institute of Technology, Kanpur,

NPTEL- Advanced Geotechnical Engineering

Example 1.2 refer to Figure 2.21. For the steady state-condition, . The coefficient of permeability of the layer is 0.03 mm/s. for the steady-state pumping condition, estimate the rate of discharge . Solution from equation (1.37),

So,

Figure 2.22 Shape factor coefficient used for condition (A) of table 1.6. (after U. S. Navy 1971, based on Figure 11-11 from Soil Engineering, 3d ed, by Merlin G. Spangler and Richard L. Hardy. Copyright 1951 1963, 1973 by Harper and Row, Publishers, Inc. Reprinted by permission of the publisher.)

Dept. of Civil Engg. Indian Institute of Technology, Kanpur,

NPTEL- Advanced Geotechnical Engineering

Figure 2.23 Shape factor coefficient,

- condition (F-1) of table 1.6 (After U. S. Navy, 1971)

Observation well in isotropic soil: Obtain shape factor from table 1.6

Piezometer in isotropic soil: Radius of intake point, R, differs from radius of stand-pipe, r

Test in anisotropic soil: Estimate ratio of horizontal to vertical permeability and divide horizontal dimensions of the intake by: to compute mean permeability . For condition (C), table 1.6:

For condition (C)

Dept. of Civil Engg. Indian Institute of Technology, Kanpur,

NPTEL- Advanced Geotechnical Engineering

For this equation for q, we can construct the following table:

25 30 40 50 100 From the above table, the rate of discharge is approximately Table 1.5 values of 1 2 3 4 5 6 7 8 9 10 15 20 [equation (1.47)] EX 31,000 19,400 14,00 11,600 9,800 8,500 7,500 6,800 6,200 5,700 4,100 3,200

0.5 0.48 0.45 0.43 0.37 .

AX 28,500 18,100 13,600 11,000 9,300 8,100 7,200 6,500 5,900 5,400 3,900 3,100

BX 25,800 16,800 12,700 10,300 8,800 7,600 6,800 6,100 5,600 5,200 3,700 3,000

NX 23,300 15,500 11,800 9,700 8,200 7,200 6,400 5,800 5,300 4,900 3,600 2,800

After U. S. Bureau of Reclamation (1961) noted: 1 ft = 0.3048 m.

Dept. of Civil Engg. Indian Institute of Technology, Kanpur,

10

NPTEL- Advanced Geotechnical Engineering

1.1.7 Theoretical Solution for Coefficient of Permeability

It was pointed out earlier in this chapter that the flow through soils finer than coarse gravel is laminar. The interconnected voids in a given soil mass can be visualized as a number of capillary tubes through which water can flow (Figure 2.24).

Figure 2.24 Flow of water through tortuous channels in soil

According to the Hagen-Poiseuilles equation, the quantity of flow of water in unit time, q, through a capillary tube of radius R can be given by (1.48) Where Unit weight of water Absolute coefficient of viscosity Area cross section of tube Hydraulic gradient The hydraulic radius of the capillary tube can be given by (1.49) From equation (1.48) and (1.49), (1.50)
Dept. of Civil Engg. Indian Institute of Technology, Kanpur, 11

NPTEL- Advanced Geotechnical Engineering

For flow through two parallel plates, we can also derive (1.51) So, for laminar flow conditions, the flow through any cross section can be given by a general equation: (1.52) Where is the shape factor. Also, the average velocity of flow is given by (1.53) For an actual soil, the interconnected void spaces can be assumed to be a number of tortuous channels (Figure 2.24), and for these the term S in equation (1.53) is equal to now,

(1.54) If the total volume of soil is , the volume of voids is where n is porosity. Let surface area per unit volume of soil (bulk). From equation (1.54), be equal to the

(1.55) Substituting equation (1.55) into equation (1.53) and taking velocity through soil), we get is the actual seepage

(1.56) It must be pointed out that the hydraulic gradient used for soils is the macroscopic gradient. The factor S in equation (1.56) is the microscopic gradient for flow through soils. Referring to Figure 2.24, . So, (1.57) Or (1.58) Where T is tortuosity, .

Again, the seepage velocity in soils is (1.59) Where is the discharge velocity. Substitution of equation (1.59) and (1.58) into equation (1.56) yields

Dept. of Civil Engg. Indian Institute of Technology, Kanpur,

12

NPTEL- Advanced Geotechnical Engineering

Or (1.60) In equation (1.60), is the surface area per unit volume of soil. If we define volume of soil solids, then as the surface area per unit (1.61) Where is the volume of soil solids in a bulk volume that is,

So, Combining equation (1.60) and (1.62), we obtain

(1.62)

(1.63) Where e is the void ratio. This relation is the Kozeny-Carman (Kozeny, 1927; Carman, 1956). Comparing equation (1.4 and 1.63), we find that the coefficient of permeability is

(1.64) This absolute permeability was defined by equation (1.6) as

Comparing equation (1.6 and 1.64), (1.65) The Kozeny-Carman equation works well for describing coarse-grained soils such as sand and some silts. For these cases, the coefficient of permeability bears a linear relation to . However, serious discrepancies are observed when the Kozeny-Carman equation is applied to clayey soils. For granular soils, the shape factor is approximately 2.5 and the tortuosity factor T is about .

1.1.8 Variation of Permeability with Void Ratio in Sand

Based on equation (1.64), the coefficient of permeability can be written as (1.66) Or
Dept. of Civil Engg. Indian Institute of Technology, Kanpur,

(1.67)
13

NPTEL- Advanced Geotechnical Engineering

Where

are the coefficients of permeability of a given soil at void ratios of

respectively.

Several other relations for the coefficient of permeability and void ratio have been suggested. They are of the form (1.68) (1.69) A.Hazen (1911) gave an empirical relation for permeability of filter sands as (1.70) Where is in and is the effective size of the soil in cm.

Equation (1.70) was obtained from the test results of Hazen where the effective size of soils varied from 0.1 to 3 mm and the uniformity coefficient for all soils was less than 5. The coefficient 100 is an average value. The individual test results showed a variation of the coefficient from 41 to 146. Although Hazens relation is approximate, it shows a similarity to equation (1.69). A. Casagrande has also given an empirical relation for k for fine or medium clean sands with bulky grain as (1.71) Where is the coefficient of permeability at a void ratio of 0.85.

Variation of Permeability with Void Ratio in Clay The Kozeny-Carman equation does not successfully explain the variation of the coefficient of permeability with void ratio for clayey soils. The discrepancies between the theoretical and experimental values are shown in Figures 2.25 and 2.26. These results are based on consolidation permeability tests (Olsen, 1961, 1962). The marked degree of variation between the theoretical and experimental values arise from several factors, including deviations from Darcys law high viscosity of the pore water, and unequal pore sizes. Olsen has developed a model to account for the variation of permeability due to unequal pore sizes.

Dept. of Civil Engg. Indian Institute of Technology, Kanpur,

14

NPTEL- Advanced Geotechnical Engineering

Figure 2.25 Coefficient of permeability for sodium illite. (After H. W. Olsen. Hydraulic Flow through Saturated Clays, SC. D. Thesis, Massachusetts Institute of Technology, 1961) Example 1.3 results of a permeability test are drawn in Figure 2.26(a) Calculate the composite shape factor, , of the Kozeny-Carmen equation, given poise. (b) If determine . Compare this value with the theoretical value for a sphere of diameter Solution Part (a): from equation (1.64)

The value of 2.26). So

is the slope of the straight line for the plot of

against k (Figure

Part (b): (Note the units carefully.)

For

Dept. of Civil Engg. Indian Institute of Technology, Kanpur,

15

NPTEL- Advanced Geotechnical Engineering

This value of

agrees closely with the estimated value of

Figure 2.26 Ratio of the measured flow rate to that predicted by the Kozeny-Carman equation for several clays. Curve 1: Sodium illite, . Curve 3: Natural kaolinite, Distilled water . Curve 4: Sodium Boston blue clay, . Curve 5: Sodium Kaolinite, 1% (by Wt.) sodium tetraphosphate. Curve 6: Calcium Boston blue clay, . (After H. W. Olsen, Hydraulic Flow through Saturated Clays, Sc. Thesis Massachusetts Institute of Technology 1961)

Dept. of Civil Engg. Indian Institute of Technology, Kanpur,

16