Engineering Geology 290 (2021) 106193

Contents lists available at ScienceDirect

Engineering Geology
journal homepage: www.elsevier.com/locate/enggeo

A modified series-parallel electrical resistivity model of saturated sand/

clay mixture
Md Farhad Hasan a, *, Hossam Abuel-Naga a, E.-C. Leong b
a
Department of Civil Engineering, La Trobe University, Melbourne, Australia
b
School of Civil and Environmental Engineering, Nanyang Technological University, Singapore

A R T I C L E I N F O A B S T R A C T

Keywords: Electrical resistivity tomography, ERT, is a geophysical method that is used to image subsurface geomaterials for
Electrical conductivity various geological, environmental and geotechnical engineering problems. The success of ERT depends on the
Saturated soils availability of a robust electrical resistivity model of soils that is able to characterise the different type of soils
Surface conduction
using its electrical resistivity measurements. The aim of this paper is to present an electrical resistivity/con­
Series-parallel
Clay
ductivity model for isotropic saturated sand/clay mixture that considers the effects of sand content, and surface
Sand conduction of clay content. The proposed model is formulated using a modified series-parallel mixing model
theory which uses phase-volumetric and surface conduction model parameters. The phase-volumetric parmeters
describe the configuration of the different soil/water phases in a representative soil unit cell which includes solid
soil particles, free water, diffuse double layer (DDL) water. The surface conduction model parameters are used to
describe the volume fraction of DDL water in the unit cell, and its electrical conductivity compared to the
electrical conductivity of free water. Simple test methods are used in this study to determine the proposed model
parameters. Laboratory experimental programme was also conducted in this study to validate the proposed
model. Two types of clays, namely, kaolin and bentonite, and one type of clean sand were used to constitute five
different sand/clay and clay/clay mixtures. Dynamic compaction method using Standard Proctor tools was used
to constitute the testing samples. The electrical conductivities of different reconstituted soil samples were
measured and compared with the proposed model predictions, and a good agreement was achieved. The small
differences between the model predictions and the experimental measurements could be attributed to the slight
deviation of the laboratory test samples from the model assumptions, which require the soil sample to fulfil fully
saturated and electrical isotropy conditions.

1. Introduction engineering, mineral exploration, mining applications and archaeolog­

ical fields (Giao et al., 2003; Sudha et al., 2009; Kibria and Hossain,
Electrical resistivity tomography (ERT) is a non-destructive 2012; Kneisel et al., 2014; Yeh et al., 2015; Barron et al., 2016).
geophysical imaging technique that is mainly used to represent varia­ ERT dates back to 1912 when Schlumberger introduced the idea of
tions in the electrical resistivity of the subsurface soils. As the electrical using electrical resistivity measurements to study subsurface rock bodies
resistivity of soils is controlled by its geotechnical properties, ERT could (Barker, 1979; Dahlin, 2001). In principle, ERT involves using a large
be considered as an effective imaging technique to determine subsurface number of electrodes in the area of interest. Electrical signals are
geotechnical parameters such as soil texture (Triantafilis and Lesch, transmitted through selected electrode locations, while electrical po­
2005; Lesch et al., 2005; Abdu et al., 2008), coarse content in soils tential measurements are conducted at other electrode locations. This
(Basso et al., 2010; Tetegan et al., 2012; Rossi et al., 2013), soil structure process is repeated systematically, and the resulting data set enables the
and compaction (Seladji et al., 2010; Al Rashid et al., 2018; Hasan et al., reconstruction of an electrical resistivity cross-section of the survey area.
2018; Lu et al., 2019), and soil hydraulic conductivity (Doussan and The contrasts in the electrical properties enable the mapping of different
Ruy, 2009; Neyshabouri et al., 2013; Niu et al., 2015). Moreover, ERT geological materials in the survey area.
finds applications in environmental engineering, hydrology, civil Based on the above working principle, the application of ERT as a

* Corresponding author.
E-mail addresses: [email protected], [email protected] (M.F. Hasan), [email protected] (H. Abuel-Naga).

https://doi.org/10.1016/j.enggeo.2021.106193
Received 28 March 2020; Received in revised form 5 May 2021; Accepted 11 May 2021
Available online 20 May 2021
0013-7952/© 2021 Elsevier B.V. All rights reserved.
M.F. Hasan et al. Engineering Geology 290 (2021) 106193

Sand particle

Pore (filled
with free
water

Idealized unit cell mixing model

Parallel model Isotropic Series-Parallel model Series model

(1-x)/2 Water
1-n Water

1.0 Solid Water = √1 − Solid

n Solid
(1-x)/2

1-n n (1-x)/2 = √1 − (1-x)/2 1.0

Modified Series-Parallel model

Layer 1 Water Water

Equivalent layer by
Layer 2 Solid
Parallel model
Layer 3 Water

Slicing the unit cell horizontally Equivalent series layer system

Fig. 1. Series-Parallel mixing model of fully saturated sands, where n is the porosity, and x is introduced to show mathematical representation of the specified part of
the unit cell.

geotechnical site investigation method mainly depends on the avail­ To overcome this limitation, Waxman and Smits (1968) added
ability of a strong correlation between soil properties, and its electrical another pathway to Archie's model. Pore water and solid surfaces were
resistivity. In fact, several studies have shown that the electrical re­ assumed as two parallel conduction paths for electrical current through
sistivity of soils is a function of several soil properties, including soil the soil medium. In line with the model of Waxman and Smits (1968),
mineralogy (Kalinski and Kelly, 1994), particle size distribution (Hasan Rhoades et al. (1976) incorporated the volumetric water content in their
et al., 2018), void ratio (Hasan et al., 2018; Al Rashid et al., 2018), pore model. The Rhoades et al. (1976) model was able to predict the lower
water connectivity (Al Rashid et al., 2018; Hasan et al., 2018; Lu et al., range of electrical resistivity, but it predicts negative values of solid
2019; Hasan, 2021), degree of water saturation (Al Rashid et al., 2018; particle's conductivity for some specific clays with high cation exchange
Hasan et al., 2018; Lu et al., 2019; Chung et al., 2019; Alsharari et al., capacity (CEC) (Shah and Singh, 2005). Therefore, Rhoades et al. (1989)
2020; Hasan, 2021), pore water salinity (Abu-Hassanein et al., 1996; added a series pathway for electrical current flow through the mixture of
McCarter and Desmazes, 1997; Samouëlian et al., 2005; Hasan, 2021), soil solids and pore water in an upgraded model that provides a
the electrical resistivity of the fluid and temperature (Abu-Hassanein reasonable prediction over the full range of electrical resistivity data
et al., 1996; McCarter and Desmazes, 1997; Samouëlian et al., 2005; (Mitchell and Soga, 2005). However, determining parameters in
Long et al., 2012; Oh et al., 2014; Son et al., 2020; Hasan, 2021). Rhoades et al. (1989) model remains a challenging task to implement.
Archie (1942) proposed a model for predicting the electrical re­ Therefore, there is a need to develop an electrical resistivity model that
sistivity of sandy soils by assuming that the electrical current passes uses parameters that have physical meaning and can be measured by
through the pore water only. However, since the effect of surface charge simple test methods. In this study, a modified series-parallel model that
of clay particles on electrical resistivity was neglected, Archie's model satisfies these requirements is introduced and validated experimentally.
failed to predict the electrical resistivity of fine-grained soils (Klein and
Santamarina, 2003; Oh et al., 2014; Glover, 2010). Nevertheless, few 2. Modified series-parallel mixing (MS-PM) model for saturated
researchers had attempted to upgrade Archie's model to fit the electrical sands
behaviour of fine-grained soils and proposed a modified Archie's model
(Atkins Jr and Smith, 1961; Salem and Chilingarian, 1999; Glover, Saturated soil is a two-phase composite material where the solid soil
2010). However, those modified models were not established based on a particles can be considered as the dispersed phase, and the soil water is
robust theoretical background. Most of the modified Archie's models are the continuous phase. The geometric configuration of the mixing model
case-dependent, and therefore, require a large volume of experimental of the saturated soils is very complicated. However, some simplifications
tests for determining its parameters. have been used to facilitate the mathematical modelling of the con

2
M.F. Hasan et al. Engineering Geology 290 (2021) 106193

7.0 composite material, which could be zero if one of the phases is a non-
Modified Series Parallel conductive material which is the case of most solid soil particles.
6.0 Parallel However, these parallel or series mixing models do not represent the
Conventional Series Parallel saturated soil in its natural state. Series-parallel mixing model could, to
5.0 Lewis-Nielsen some extent, represent the saturated soil. However, there is an unlimited
number of possible combinations of series and parallel mixing models.
4.0 Fig. 1 shows a special case of series-parallel mixing model where
isotropic conduction behaviour is assumed. Fig. 2 shows the electrical
conductivity at different porosities, n, of parallel and conventional
R

3.0
isotropic series-parallel mixing models shown in Fig. 1. The electrical
2.0 conductivity results in Fig. 2 are calculated using the conventional
approach and presented in terms of the relative electrical resistivity
1.0 ratio, R, as follows:
ρ σFW
0.0 Rconv = = (1)
ρFW σ
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
n where ρ and σ are the overall electrical resistivity and conductivity of
saturated soil unit cell, respectively, and ρFW and σFW, are the electrical
Fig. 2. Electrical conductivity of saturated sand at different porosities n using resistivity and conductivity of soil-free water. For the conventional
different models. series-parallel model,
1
duction behaviour of the two-phase composite material. The parallel Rsp = (2)
1− ϕ
mixing model considers the upper boundary of the conduction through
two-phase composite materials, as shown in Fig. 1, where the saturated ϕ = 1− n (3)
soil is presented using a unit cell. On the other side, the series mixing
model represents the lower conductivity boundary of two-phase where n is the porosity. Several theoretical, empirical and semi-

Clay particles
with DDL water

Aggregate
Inter-aggregate
pore (filled
with free water

Sand particle

Modified Series-Parallel model

Free DDL
L1 Layer 1 (V
VL1)
water water
L2 Layer 2 (VL2)
L3 Layer 3(VL3)
1.0
Solid L4 Layer 4 (VL4)
L3 Layer 3 (VL3)
L2 Layer 2 (VL2)
L1 Layer 1(VL1)

L1 L2 L3 L4 L3 L2 L1 1.0
1.0

Slicing the unit cell horizontally Equivalent series layer system

Fig. 3. Series-Parallel mixing model of fully saturated sand/clay mixture.

3
M.F. Hasan et al. Engineering Geology 290 (2021) 106193

empirical conductivity models of two-phase composite materials are as follows (Fig. 3):
available in the literature (Maxwell, 1881; Fricke, 1931; Archie, 1942; √̅̅̅̅̅̅̅̅̅̅̅
Lewis and Nielsen, 1970; Pal, 2007). According to Pal (2008), the Lewis- 1− 1 − n − 2L2
L1 = (7)
Nielsen Model appears superior to the other models as it is already 2
validated by a vast amount of experimental data. Lewis-Nielsen Model of L2 = L4 (8)
saturated soil where the electrical conductivity of the solid soil particles
is zero can be expressed as follow: √̅̅̅̅̅̅̅̅̅̅̅
1 − n − L4
L3 = (9)
1 + 0.5ψ ϕ 2
RLNP = (4)
1− ϕ The equivalent electrical conductivity, σLn, of each sliced layer, can
( ) be determined by the parallel mixing model as follows:
1 − ϕm
ψ = 1+ ϕ (5) σL1= (1 − L4 )σFW + L4 σDDL (10)
ϕ2m
( √̅̅̅̅̅̅̅̅̅̅̅ ) ( √̅̅̅̅̅̅̅̅̅̅̅ )
where ϕm is the maximum packing volume fraction of solid particles (the σL2= 1 − 2L4 − 1 − n σFW + 2L4 + 1 − n σDDL (11)
minimum possible value of porosity, n). The electrical conductivity re­
sults of Lewis-Nielsen Model for a random close packing of sphere par­ ( √̅̅̅̅̅̅̅̅̅̅̅ )
σL3 = 1 − 2L4 − 1 − n σFW + 2L4 σDDL (12)
ticles at ϕm = 0.65 (practical value for soils) are also plotted in Fig. 2.
The results of Lewis-Nielsen Model fall between the conventional par­ ( √̅̅̅̅̅̅̅̅̅̅̅ )
allel and isotropic series-parallel mixing models. In other words, the σL4 = 1 − 1 − n σDDL (13)
conventional isotropic series-parallel mixing model underestimates the
electrical conductivity of saturated soil compared to Lewis-Nielsen where σFW, and σDDL are the electrical conductivity of free water and
Model. This behaviour could be attributed to the neglect of the elec­ DDL water, respectively. The overall electrical conductivity of the unit
trical conductivity of the water in the series part of series-parallel mixing cell, σ, can be determined using the series mixing model as follows:
model. To overcome this limitation, a Modified Series-Parallel Mixing
(MS-PM) model for isotropic saturated soils is proposed in this study. 1 2L1 2L2 2L3 L4
= + + + (14)
The proposed approach involves slicing the unit cell horizontally and σ σL1 σL2 σL3 σL4
calculating the equivalent conductivity of the horizontally sliced layers Based on the above equations, the volume fractions and the electrical
using the parallel mixing model. This will change series-parallel unit cell conductivities of the three phases (free water, DDL water, solid particles)
to an equivalent series unit cell, as shown in Fig. 1. Then, The relative should be determined to calculate the overall electrical conductivity of
resistivity, R, of the equivalent unit cell (MS-PM model) can be unit cell.
expressed as follows:
√̅̅̅
√̅̅̅ ϕ 3.1. Determination of unit cell volume fractions
RMS− PM = 1 − ϕ+ √̅̅̅ (6)
1− ϕ
For sand/clay mixture unit cell, the total volume of solid soil parti­
The results of the MS-PM model are also plotted in Fig. 2, and it cles, Vs, comprises the volume of fine solids, Vfs, (passing sieve no. 200),
almost coincides with the results of Lewis-Nielsen Model. Therefore, the and coarse solids, Vcs , (retaining on sieve no. 200) where the volumetric
proposed MS-PM model is equivalent to Lewis-Nielsen Model and can be fine content, Δ, is expressed as follows:
used, with a reasonable level of confidence, in predicting the conduction
behaviour of saturated sand soils. However, it should be mentioned that Vfs
∆= (15)
both Lewis-Nielsen Model and MS-PM model cannot be used for satu­ Vs
rated clay soils as their particles have electrical surface conduction that
where,
is considered in both models. In the next section, the electrical surface
conduction will be incorporated in the MS-PM model. Vs = Vcs + Vfs = 1 − n (16)

3. Modified series-parallel mixing (MS-PM) model for saturated The total volumetric water content (θ), the volumetric water content
sand/clay mixtures of free water (θFW) and volumetric water content of DDL water (θDDL) are
defined as follows:
The unit cell of MS-PM model for the isotropic saturated soil (solid Vw
particles are insulator) is shown in Fig. 1. where the solid soil particles θ= (17)
V
are represented as a square concentrated at the centre of the unit cell of
soil as a non-conductive medium with an area of (1-n), where n is the soil VDDL
θDDL = w
(18)
porosity. For the clay soils, the electrical surface conduction is included V
in the unit cell by representing its diffuse double layer (DDL) water as a
uniform layer covering the perimeter of the central unit of solid parti­ and
cles, and it also extends to the four sides of the soil element where the θ = θFW + θDDL (19)
electrical potential is applied, as shown in Fig. 3. These extensions
represent the possible electrical current flow at the particles contact where, VDDL
w , and V are the volumes of DDL water, and total soil,
through their DDL water layer. A similar approach was used by Haigh respectively. Using the soil phase relation of fully saturated soils, the
(2012) in his soil unit cell to represent the capillary water at the sand volumetric water content, θ, is equal to the soil porosity, n, and hence
particles contacts in unsaturated soils. Therefore, the clayey soil unit cell Eq. 17 can be expressed as follows:
includes three phases, namely, solid particles, DDL water, and free
Vv wc Gs
water. θsat = n = = (20)
V 1 + wc Gs
Following the proposed slicing approach for the soil unit cell, the
symmetric soil unit cell could be divided into seven parallel layers and where, Vv, wc, and Gs are volume of the void, gravitational water con­
the thickness of each layer, denoted as Ln, n = 1 to 4, can be determined tent, and specific gravity, respectively. According to Hasan et al. (2018),

4
M.F. Hasan et al. Engineering Geology 290 (2021) 106193

R2=0.98
-0.01800 Vef f DDL
s = Vs + Vw (23)
-0.003500
0.01100 here, Vfs and Vefs are the volumes of fines solid and effective fines solid,
0.02550
respectively; χ is the soil parameter that expresses the ratio between Vfs
and Vef
s , and it can be determined using a simple test (Hasan et al., 2018).
0.04000
0.05450

The term χt is also used to represent the ratio of , and it can be

2 VDDL
w +Vs
0.1
0.06900
Vs
related to χ as follows:
0.08350
0
0.1 0.09800
8
0.0 0.1125
χt = Δ(χ − 1) + 1 (24)
6 0.1270
0.0
L2 04
The term L2 can be linked with χt and n as follows:
0.
L2 values (√̅̅̅̅̅̅̅̅̅̅̅ )2 [ (√̅̅̅̅̅̅̅̅̅̅̅ )]
02
0.
(χ t − 1)(1 − n) = 1 − n + 2L2 − (1 − n) + 2L2 1 − 1 − n + 2L2
00
0.

.0
2 1.8 (25)
-0 . 2
0
1.6
L2 from Eq. 25 was subject to 3D surface analysis, as shown in Fig. 4,
4
0. 1.4 and the equation to determine L2 was found as the following:
F
t

1.2
(26)
6
0.

L2 = L4 = 0.01 + 0.035n + 0.08χ t

n

1.0
8
0.

Fig. 4. Surface analysis to predict L2 as a function of n and χt. 3.2. Determination of electrical conductivities of free and DDL water

θDDL for a unit volume of soil can be expressed as follows: The electrical resistivity of the free water can be determined by
θDDL = ∆[(1 − n)(χ − 1) ] (21) squeezing the soil sample and measuring the electrical resistivity of the
squeezed water. The electrical behaviour of the DDL layer is mainly
where, controlled by the influence of ion mobility which depends on the pore
water salinity level and the clay surface density of electrical charges. The
χ=
Vef
s
(22) amount of surface density of negative electrical charges is proportional
Vfs to the amount of isomorphous substitutions and the type of clay. In this
case, kaolin and montmorillonite clays have the highest and lowest
and order of surface density of charges, respectively, among various types of
clays. In addition, the pore water solution provides the system with

Fig. 5. Proposed simple approach to determine electrical conductivity of DDL water, σDDL.

5
M.F. Hasan et al. Engineering Geology 290 (2021) 106193

Fig. 6. Effect of clay properties on R-n relationship at (a) α = 2, (b) α = 10, and (c) α = 100.

positive charges. More positive charges in pore solution indicate stron­ The parameter σs and χ could be determined experimentally as
ger attraction between the surface of soil particles and the pore water proposed by Hasan et al. (2018).
and, as a result, lower ion mobility. The ions in the free water are
completely mobile. Ions are attracted to the surface negative charges, 3.3. Effect of clay properties on R
when those are closer to the surface, and within the DDL layer; thus, the
mobility of the ions is reduced in this region. The closest cations to the The proposed model in this study uses the parameters χt and α to
surface are held firmly by the negative charges and are immobile (Bolt, consider the effect of clay mineralogy and its interaction with the soil
1935; Mitchell and Soga, 2005; Mojid et al., 2007). Bolt (1935) postu­ water on the electrical properties of sand/clay mixtures. The term χt
lated that by increasing the concentration of ions in the pore solution, describes the volume fraction of DDL water in the unit cell, whereas the
the electrical resistivity of DDL becomes proportional to the free water term α presents the relationship between the electrical conductivity of
electrical resistivity, ρFW, with the coefficient of proportionality defined DDL and free water. Therefore, both parameters control the electrical
as α, and therefore, ρFW = αρDDL, where α ≥ 1.0. surface conduction of clay. It should also be mentioned that χt is a
Tabbagh and Cosenza (2007) reported that for concentric spherical function of the volumetric content of the clay, Δ, clay mineralogy, and
volumes with radii x and y that correspond to the non-conducting solid pore water salinity whereas α is a function of clay mineralogy, and pore
core and the surrounding conductive DDL, the relationship between the water salinity. For most of the natural soils, χt and α can be in the range
overall conductivity of the sphere, σs, and σDDL can be written as the of 1 to 2.5 and 1 to 200, respectively.
following: Fig. 6 shows the prediction of the proposed model in this study for R
of sand/clay mixture as n, χt, and α change. For α =2 (Fig. 6a), the model
x3 + 2y3
σDDL = σs (27) predictions show that R decreases as n and χt increase. This behaviour is
2(y3 − x3 )
expected as increasing n increases the volume of free water, which is
more conductive than the solid soil particles. At constant n and α,
where for the fines solid in the unit cell, x and y can be expressed as
increasing χt means replacing free water by DDL water. As DDL water
follows:
has higher electrical conductivity than free water, R decreases. How­
3 ever, it can be observed that in this case, the values of R is greater than
x3 = (1 − n)∆ (28)
4π 1.0, and it should be 1.0 when n = 1.0. As α increases, this behaviour
changes. At α = 10 (Fig. 6b), the behaviour of R as n increases is a
3
y3 = [(1 − n)∆χ ] (29) function of χt. At χt = 1.1, R decreases as n increases, whereas an
4π opposite behaviour can be observed at χt = 1.4. However, at χt = 1.1, R is
Fig. 5 shows the idealisation of the clay particle and its surrounded almost equal to 1.0 regardless of the n value.
DDL water into an effective clay particle where the volume of solid clay This behaviour could be explained in terms of the ratio between the
particle and its DDL water are merged together and assigned an equiv­ electrical conductivity of free water, σFW, and the equivalent electrical
alent electrical conductivity, σs, then, converting the rectangular effec­ conductivity of the combined volume of DDL water and solid particles,
tive clay particle to an equivalent sphere particle and use Tabbagh and σs, as follows:
Cosenza (2007) equation to get σDDL as follows:
• Case A: σFW = σs, R = 1.0 regardless of n value.
σs (1 + 2χ)
σDDL = (30) • Case B: σFW > σs, R > 1.0, and it decreases as n increases. At n = 1.0,
2(χ − 1)
R becomes 1.0
Therefore, α, can be expressed as follows: • Case C σFW < σs, R < 1.0, and it increases as n increases. At n = 1.0, R
becomes 1.0
ρFW σs (1 + 2χ)
α= = (31)
ρDDL 2σFW (χ − 1)

6
 Passing %

Gs
10B

SiO2
10 K

Al2O3

0
10
20
30
40
50
60
70
80
90
100
5 K-5B

Table 2
Table 1

100 g)
3 K-4B-3S
3 K-6B-1S

Properties

0.001
Mixture name
M.F. Hasan et al.

Liquid limit (%)

Plastic limit (%)

Properties of sand

Minimum Void Ratio

Maximum Void Ratio

Sand

Kaolin
Total surface area (m2/g)
0

Bentonite
30
30
50
100

pH in water (28%-40% solid)

Surface charge density (μC/m2)

Cation exchange capacity (meq/
Properties of clays used in tests

4. Experimental validation
0.01
Kaolin (K) %

20
32
74

0.58
0.97
38.8
45.2
0.36
2.58

4.1. Test materials and specifications

0.075
Kaolin

(weight %)
0

40
60
50

100

Diameter (mm)
Composition of soil mixtures (by weight) used in tests

0.1
Bentonite (B) %

Main chemical composition

Fig. 7. Particle size distributions of tested soils.

follow Case C where R < 1.0 regardless of n and χt values.

0
0
0

30
10

80
53

9.5

750
504

13.6
63.8
2.68

10.24
Bentonite
Sand (S) %

Based on the above explanation, the results of α = 100 as shown in

whereas Fig. 7 represents the particle size distribution of the clays, and
the geotechnical properties of clays, as provided by the suppliers,

sand, obtained by the laser diffraction method (ASTM D422-63e2,

type of clean sand (S) were used in this study to constitute five different

2007). Two types of water in terms of its electrical conductivity

sand/clay and clay/clay mixtures, as listed in Table 1. Table 2 provides
Two types of clays, namely, kaolin (K) and bentonite (B), and one
Fig. 6c, indicate that as α increases beyond a certain limit, the soil will

7
Table 3
Comprehensive research data comprising experimental and numerical results with physical parametric values
General Information Electrical Conductivity Parameters Volume Fraction Parameters Physical Parameters Experimental Results Predicted Results

Soils W⋅C (%)* Blows σw (S/m) σs (S/m) α Δ ϴ χ χt e S (%) σvl (S/m) σhl (S/m) λ σavg (S/m) R (exp) R (predicted)

10 K 10.2 5 0.0014 0.00376 124.76 1.0 0.55 1.04 1.04 1.22 94.15 0.002231 0.002471 0.90 0.002351 0.60 0.59
10 0.51 1.04 94.44 0.002267 0.0025 0.91 0.00238 0.59 0.59
15 0.48 0.92 94.47 0.002379 0.002612 0.91 0.0025 0.56 0.56
25 0.43 0.75 95.23 0.002571 0.002758 0.91 0.00263 0.53 0.54
5 K-5B 10.3 5 0.0014 0.0335 91.65 1.0 0.64 1.53 1.53 1.78 95.27 0.011 0.0116 0.92 0.0111 0.126 0.12
10 0.60 1.5 95.38 0.012 0.0131 0.91 0.0125 0.112 0.11
15 0.54 1.33 95.63 0.0131 0.0147 0.90 0.0139 0.101 0.10
25 0.52 1.00 96.00 0.0151 0.0167 0.91 0.0159 0.088 0.086
10B 10.3 5 0.0014 0.082 167.04 1.0 0.69 1.81 1.81 2.23 97.29 0.0261 0.0269 0.97 0.0265 0.053 0.04
10 0.65 1.86 97.41 0.0287 0.0309 0.93 0.0298 0.047 0.034
15 0.62 1.63 97.57 0.0324 0.0359 0.91 0.0342 0.031 0.031
25 0.56 1.27 97.94 0.0409 0.0446 0.92 0 0.0428 0.025 0.032
3 K-4B-3S 11.44 10 0.011 0.083 153.87 0.703 0.55 1.11 1.077 1.22 95.67 0.0221 0.0234 0.94 0.0228 0.48 0.40
15 0.53 1.13 97.28 0.026 0.0276 0.946 0.0268 0.41 0.39
20 0.51 1.04 96.49 0.0279 0.0288 0.969 0.0284 0.38 0.37
25 0.47 0.89 97.55 0.0286 0.0296 0.966 0.0291 0.37 0.36
3 K-6B-1S 10.06 10 0.011 0.133 124.33 0.901 0.52 1.16 1.08 1.08 97.48 0.0396 0.0431 0.919 0.0414 0.26 0.22
15 0.50 1.00 97.49 0.0424 0.0443 0.957 0.0434 0.25 0.21
20 0.47 0.89 98.28 0.0465 0.0488 0.952 0.077 0.23 0.20
25 0.45 0.82 98.65 0.0493 0.0541 0.911 0.0517 0.21 0.19
*
At the commencement of compaction.
Engineering Geology 290 (2021) 106193
M.F. Hasan et al. Engineering Geology 290 (2021) 106193

1
3K-6B-1S

Predicons
0.9 3K-4B-3S

Model
10K
0.8
5K-5B

0.7 10B
3K-6B-1S

Experimental
0.6 3K-4B-3S

results
10K
0.5

R
5K-5B
0.4 10B

0.3

0.2

0.1

0
0.4 0.45 0.5 0.55 0.6 0.65 0.7 0.75
n

Fig. 9. Comparison between the experimental results and the proposed model
predictions

Fig. 8. Arrangement for measuring electrical conductivity of saturated clays

(σ vl: principal axis orientated parallel to the compaction direction; σ hl: principal Table 4
axis orientated perpendicular to the compaction direction) (Hasan et al., 2018) Electrical conductivity readings of diluted kaolin solutions using different
electrical conductivity probes and their impact on model parameters
(0.0014 S/m, 0.011 S/m) were used in this study to constitute the testing EC Accuracy (S/ σ1* (S/ σ2* (S/ σs (S/m) χ α
sand/clay mixture specimens as listed in Table 3. The saline water was Probe m) m) m)
prepared by adding different salt concentrations in the de-ionised water. A ±0.6% 0.0074 0.0096 0.00376 1.033 124.76
The maximum and minimum void ratios of the sand were 0.97 and 0.58, B ±2-2.5% 0.0071 0.0095 0.00374 1.04 102.85
respectively (ASTM D4253-16, 2016; ASTM D4254-16, 2016). The *
Electrical conductivity reading of the diluted soil solutions following the
liquid limit and plastic limit of clays were based on ASTM D4318-10e1, approach proposed by Hasan et al. (2018)
2010.
To determine σs and χ of different clay mixtures, two diluted solu­ as shown in Fig. 8. The trimmed specimens were used to determine the
tions with different clay contents (5 g/L and 10 g/L) were prepared, and vertical and horizontal electrical conductivity values, σvl and σhl, of each
their electrical conductivity values were measured. Back calculation saturated compacted clay specimen using the four electrodes test setup
technique using series-parallel mixing model of the diluted solution was (Abu-Hassanein et al., 1996). In this test setup, a one-dimensional
used to determine σs and χ (Hasan et al., 2018). The values of σs and χ, of electrical field was applied across the cylindrical specimen via circular
each clay mixture used in this study are listed in Table 3. plate electrodes pressed against the ends of the specimen and two inner
electrodes inserted at equal spaces, L, along the height of the testing
specimen to measure the electrical potential difference, v, as shown in
4.2. Preparation of specimens and experimental design
Fig. 8. The electrical conductivity, σ, can be expressed as follows:

The experimental programme was designed to understand the effect σ=

IL
(32)
of soil composition (different sand/clay mixture), clay mineralogy vA
(kaolin, and bentonite), and porosity on the electrical conductivity of
where, I and A are the electrical current and the cross-sectional area of
saturated soil. Dynamic compaction following the standard Proctor
the test specimen normal to the current flow direction, respectively.
method (ASTM D698, 2012) was used in this study to constitute the
testing specimens. As the proposed model in this study is only valid for
isotropic soils, the dynamically constituted specimens were compacted 4.3. Test results and discussion
at the dry side of the compaction curve as the compacted soil specimens
are expected to hold an isotropic fabric under this condition (Abu-Has­ Table 3 lists the experimental results of the test conducted in this
sanein et al., 1996; Hasan et al., 2018). For this purpose, low water study to determine the electrical conductivities of the tested specimens
contents in the range between 10% to 16% were used as listed in Table 3. and its electrical and physical parameters that can be used to predict its
After mixing soil with the targeted water content, the soil mixture electrical conductivity according to the proposed model in this study.
was kept in air-tight bags for moisture equilibration at room tempera­ The comparison between the experimentally measured R of the tested
ture for 24 h, as described in Hasan et al. (2018) and Leong and Abuel- soil specimens and the proposed model prediction is shown in Table 3
Naga (2018). Then, the specimens were constituted at different dry and Fig. 9 where good agreement can be observed. The slight difference
densities (porosities) by changing the compaction efforts (no. of blows between the experimental and the prediction results can be attributed to
per layer). Four different numbers of blows per layer (10, 15, 20, 25) the deviation of the test specimen properties from the proposed model
were used in this study. The compacted specimens were placed inside a assumptions in terms of fully saturated and electrical isotropy condi­
CBR soaking tank to fully saturate them. The soaking water has a similar tions. The degree of saturation of the test specimens in this study varies
electrical conductivity as the water used in constituting the soil between 94.15% and 98.65%, whereas the isotropy degree λ (σvl/σhl) is
specimens. in the range of 0.90 to 0.97. It should also be mentioned that the random
To check the level of electrical resistivity isotropy of each compacted distribution of the different types of soil (sand, kaolin, bentonite) used to
testing sample, each fully saturated compacted sample was trimmed into constitute the test specimens could create different spatial variations of
two cylinders (3 cm in diameter and 6 cm in height) having their prin­ soil composition within the test specimen, which could affect the elec­
cipal axes orientated vertical and horizontal to the compaction direction trical conductivity measurements.

8
M.F. Hasan et al. Engineering Geology 290 (2021) 106193

0.68 • The deviation of the soil sample from the fully saturated and
0.66
isotropic fabric condition would negatively affect the accuracy of the
model predictions.
0.64
0.62 Declaration of Competing Interest
0.6
The authors declare that they have no known competing financial
R

0.58
interests or personal relationships that could have appeared to influence
0.56
Experimental the work reported in this paper.
0.54 Model Predicon (EC Probe A)
0.52 Model Predicon (EC Probe B) References
0.5
Abdu, H., Robinson, D.A., Seyfried, M., Jones, S.B., 2008. Geophysical imaging of
0.4 0.45 0.5 0.55 0.6
watershed subsurface patterns and prediction of soil texture and water holding
n capacity. Water Resour. Res. 44.
Abu-Hassanein, Z.S., Benson, C.H., Blotz, L.R., 1996. Electrical resistivity of compacted
Fig. 10. Effect of the accuracy of electrical conductivity probe on the model clays. J. Geotech. Eng. 122, 397–406.
Alsharari, B., Olenko, A., Abuel-Naga, H., 2020. Modeling of electrical resistivity of soil
predictions (10K).
based on geotechnical properties. Expert Syst. Appl. 141, 112966.
Archie, G.E., 1942. The electrical resistivity log as an aid in determining some reservoir
Finally, the accuracy of the electrical conductivity probe that is used characteristics. Trans. AIME 146, 54–62.
ASTM D422-63e2, 2007. Standard Test Methods for Particle Size Analysis of Soils.
to measure the electrical conductivity of the soil diluted suspensions to Annual book of ASTM Standards.
determine vales of σs and χ, of each clay mixture can also affect the ASTM D4253-16, 2016. Standard Test Methods for Maximum Index Density and Unit
proposed model predictions. Two different EC probes in terms of their Weight of Soils Using a Vibratory Table. ASTM International, West Conshohocken,
PA.
accuracy (A and B), as listed in Table 4, were used to determine the ASTM D4254-16, 2016. Standard Test Methods for Minimum Index Density and Unit
electrical conductivity of diluted kaolin solutions to determine its σs and Weight of Soils and Calculation of Relative Density. ASTM International, West
χ, according to Hasan et al. (2018). EC probe B (low accuracy) slightly Conshohocken, PA.
ASTM D4318-10e1, 2010. Standard Test Methods for Liquid Limit, Plastic Limit and
underestimates the electrical conductivity of diluted kaolin solutions Plasticity Index of Soils, Annual Book of ASTM Standards. ASTM International, West
(σ1, σ2), which leads to an overestimation of χ and underestimation of α Conshohocken, PA.
and σs compared to the results of EC probe A, as shown in Table 4. The ASTM D698, 2012. Test Methods for Moisture-Density Relations of Soils and Soil.
Aggregate Mixtures. Method A (Standard Proctor). www.astm.org.
impact of EC probe accuracy on the model prediction results of 10 K
Atkins Jr., E.R., Smith, G.H., 1961. The significance of particle shape in formation
specimens is shown in Fig. 10 where it can be concluded that using high resistivity factor-porosity relationships. J. Pet. Technol. 13, 285–291.
accuracy electrical conductivity probe improves the model predictions. Barker, R.D., 1979. Signal contribution sections and their use in resistivity studies.
It should be mentioned that the results presented in Fig. 9 were obtained Geophys. J. R. Astron. Soc. 59, 123–129.
Barron, A.J.M., Uhlemann, S., Pook, G.G., Oxby, L., 2016. Investigation of suspected
using EC probe A (high accuracy). gulls in the Jurassic limestone strata of the Cotswold Hills, Gloucestershire, England
using electrical resistivity tomography. Geomorph. 268, 1–13.
5. Conclusion Basso, B., Amato, M., Bitella, G., Rossi, R., Kravchenko, A., Sartori, L., Carvahlo, L.M.,
Gomes, J., 2010. Two-dimensional spatial and temporal variation of soil physical
properties in tillage systems using electrical resistivity tomography. Agron. J. 102,
In this study, A new modified series-parallel model (MS-PM) to 440–449.
predict electrical resistivity (ER) of fully saturated isotropic soil has been Bolt, G.H., 1935. 1979. Soil chemistry, B. Physico-chemical models. Bruggeman DAG.
Berechnung verschiedener physikalischer Konsl-anlen von heterogenen substanzen.
proposed and experimentally validated. The proposed model has the Ann Phys (Leipzig) 24, 636–679.
following features; Chung, C.C., Lin, C.P., Yang, S.H., Lin, J.Y., Lin, C.H., 2019. Investigation of non-unique
relationship between soil electrical conductivity and water content due to drying-
wetting rate using TDR. Eng. Geol. 252, 54–64.
• The solid soil particles (non-conductive medium) are represented as Dahlin, T., 2001. The development of DC resistivity imaging techniques. Comput. Geosci-
a square concentrated at the centre of the unit cell of soil and is UK 27, 1019–1029.
surrounded by soil water (conductive medium). Doussan, C., Ruy, S., 2009. Prediction of unsaturated soil hydraulic conductivity with
electrical conductivity. Water Resour. Res. 45 (Eucken A Forsch. Gebiete
• The proposed MS-PM approach involves slicing the unit cell hori­
Ingenieurw. B3 Forschungsheft 1932;353: 16).
zontally and calculating the equivalent conductivity of each slice Fricke, H., 1931. The electric conductivity and capacity of disperse systems. Physics 1,
using the parallel mixing model. Then, applying the series mixing 106–115.
Giao, P.H., Chung, S.G., Kim, D.Y., Tanaka, H., 2003. Electric imaging and laboratory
model on the equivalent horizontally sliced layers to get the overall
resistivity testing for geotechnical investigation of Pusan clay deposits. J. Appl.
electrical conductivity of the unit cell. Geophys. 52, 157–175.
• Surface conduction of the fine solid soil content (passing sieve no. Glover, P.W., 2010. A generalised Archie’s law for n phases. Geophys. 75 (6),
200) is represented in the model through its volume of the associated E247–E265.
Haigh, S.K., 2012. Thermal conductivity of sands. Geotechnique 62, 617.
DDL water and electrical conductivity that is higher than free soil Hasan, M.D., 2021. Electrical Conductivity of Saturated Fine-Grained Soils: Modelling
water. and Soil Classification Applications. Doctoral dissertation. La Trobe.
• The volume of DDL water and its electrical conductivity are function Hasan, M.F., Abuel-Naga, H., Broadbridge, P., Leong, E.-C., 2018. Series-parallel
structure-oriented electrical conductivity model of saturated clays. Appl. Clay Sci.
of the fine volumetric content of the soil, fine soil mineralogy, and 162, 239–251.
pore water salinity. Kalinski, R.J., Kelly, W.E., 1994. Electrical-resistivity measurements for evaluating
• Hasan et al. (2018) approach can be used to determine the volume of compacted-soil liners. J. Geotech. Eng. 120, 451–457.
Kibria, G., Hossain, M.S., 2012. Investigation of geotechnical parameters affecting
DDL water and its electrical conductivity. The results in this study electrical resistivity of compacted clays. J. Geotech. Geoenviron. Eng. 138,
recommend using EC probe with an accuracy of ±0.6% S/m or better 1520–1529.
to determine DDL properties. Klein, K.A., Santamarina, J.C., 2003. Electrical conductivity in soils: underlying
phenomena. J. Environ. Eng. Geoph. 8 (4), 263–273.
• The proposed ER model in this study shows that the evolution of R as
Kneisel, C., Emmert, A., Kästl, J., 2014. Application of 3D electrical resistivity imaging
n increases depends on the ratio of σFW / σs where R decreases as n for mapping frozen ground conditions exemplified by three case studies. Geomorph.
increases where σFW / σs > 1.0 and it shows opposite behaviour if σFW 210, 71–82.
Leong, E.-C., Abuel-Naga, H., 2018. Contribution of osmotic suction to shear strength of
/ σs < 1.0
unsaturated high plasticity silty soil. Geomech. Energy Env. 15, 65–73.

9
M.F. Hasan et al. Engineering Geology 290 (2021) 106193

Lesch, S.M., Corwin, D.L., Robinson, D.A., 2005. Apparent soil electrical conductivity Rhoades, J.D., Raats, P.A.C., Prather, R.J., 1976. Effects of Liquid-phase Electrical
mapping as an agricultural management tool in arid zone soils. Comput. Electron. Conductivity, Water Content, and Surface Conductivity on Bulk Soil Electrical
Agr. 46, 351–378. Conductivity 1. Soil Sci. Soc. Am. J. 40, 651–655.
Lewis, T., Nielsen, L., 1970. Dynamic mechanical properties of particulate-filled Rhoades, J.D., Manteghi, N.A., Shouse, P.J., Alves, W.J., 1989. Soil electrical
composites. J Appl Poly Sci 14, 1449–1471. conductivity and soil salinity: New formulations and calibrations. Soil Sci. Soc. Am.
Long, M., Donohue, S., L’Heureux, J.-S., Solberg, I.-L., Rønning, J.S., Limacher, R., J. 53, 433–439.
O’Connor, P., Sauvin, G., Rømoen, M., Lecomte, I., 2012. Relationship between Rossi, R., Amato, M., Pollice, A., Bitella, G., Gomes, J.J., Bochicchio, R., Baronti, S.,
electrical resistivity and basic geotechnical parameters for marine clays. Can. 2013. Electrical resistivity tomography to detect the effects of tillage in a soil with a
Geotech. J. 49, 1158–1168. variable rock fragment content. Eur. J. Soil Sci. 64, 239–248.
Lu, Y., Abuel-Naga, H., Rashid, Q.A., Hasan, M.F., 2019. Effect of Pore-Water Salinity on Salem, H.S., Chilingarian, G.V., 1999. The cementation factor of Archie’s equation for
the Electrical Resistivity of Partially Saturated Compacted Clay Liners. Adv. Mat. Sci. shaly sandstone reservoirs. J. Pet. Sci. Eng. 23, 83–93.
Eng. 2019. Samouëlian, A., Cousin, I., Tabbagh, A., Bruand, A., Richard, G., 2005. Electrical
Maxwell, J.C., 1881. A Treatise on Electricity and Magnetism, vol. 1. Oxford (435 pp). resistivity survey in soil science: a review. Soil Tillage Res. 83, 173–193.
McCarter, W.J., Desmazes, P., 1997. Soil characterisation using electrical measurements. Seladji, S., Cosenza, P., Tabbagh, A., Ranger, J., Richard, G., 2010. The effect of
Geotechnique 47. compaction on soil electrical resistivity: a laboratory investigation. Eur. J. Soil Sci.
Mitchell, K.J., Soga, K., 2005. Fundamentals of Soil Behaviour. John Wiley & Sons, 61, 1043–1055.
Hoboken, New Jersey. Shah, P.H., Singh, D.N., 2005. Generalised Archie’s law for estimation of soil electrical
Mojid, M.A., Rose, D.A., Wyseure, G.C.L., 2007. A model incorporating the diffuse double conductivity. J. ASTM Int. 2, 1–20.
layer to predict the electrical conductivity of bulk soil. Eur. J. Soil Sci. 58, 560–572. Son, J.S., Song, S.Y., Nam, M.J., 2020. Complex resistivity survey for the evaluation of
Neyshabouri, M.R., Rahmati, M., Doussan, C., Behroozinezhad, B., 2013. Simplified ground reinforcement in a karst area. Eng. Geol. 105555.
estimation of unsaturated soil hydraulic conductivity using bulk electrical Sudha, K., Israil, M., Mittal, S., Rai, J., 2009. Soil characterisation using electrical
conductivity and particle size distribution. Soil Res. 51, 23–33. resistivity tomography and geotechnical investigations. J. Appl. Geophys. 67, 74–79.
Niu, Q., Fratta, D., Wang, Y.-H., 2015. The use of electrical conductivity measurements in Tabbagh, A., Cosenza, P., 2007. Effect of microstructure on the electrical conductivity of
the prediction of hydraulic conductivity of unsaturated soils. J. Hydrol. 522, clay-rich systems. Phys. Chem. Earth 32, 154–160.
475–487. Tetegan, M., Pasquier, C., Besson, A., Nicoullaud, B., Bouthier, A., Bourennane, H.,
Oh, T.-M., Cho, G.-C., Lee, C., 2014. Effect of soil mineralogy and pore-water chemistry Desbourdes, C., King, D., Cousin, I., 2012. Field-scale estimation of the volume
on the electrical resistivity of saturated soils. J. Geotech. Geoenviro. Eng. 140, percentage of rock fragments in stony soils by electrical resistivity. Catena 92,
06014012. 67–74.
Pal, R., 2007. On the electrical conductivity of particulate composites. J. Compos. Mater. Triantafilis, J., Lesch, S.M., 2005. Mapping clay content variation using electromagnetic
41, 2499–2511. induction techniques. Comput. Electron. Agr. 46, 203–237.
Pal, R., 2008. On the Lewis-Nielsen model for thermal/electrical conductivity of Waxman, M.H., Smits, L.J.M., 1968. Electrical conductivities in oil-bearing shaly sands.
composites. Composites: Parl A 39 (2008), 718–726. Soc. Petrol. Eng. J. 8, 107–122.
Rashid, Q.A., Abuel-Naga, H.M., Leong, E.-C., Lu, Y., Al Abadi, H., 2018. Experimental- Yeh, H.-F., Lin, H.-I., Wu, C.-S., Hsu, K.-C., Lee, J.-W., Lee, C.-H., 2015. Electrical
artificial intelligence approach for characterising electrical resistivity of partially resistivity tomography applied to groundwater aquifer at downstream of Chih-Ben
saturated clay liners. Appl. Clay Sci. 156, 1–10. Creek basin. Taiwan. Environ. Earth Sci. 73, 4681–4687.

10