Effect of Pore-Water Salinity on the Electrical Resistivity of Partially Saturated Compacted Clay Liners

/e aim of this paper is to investigate the effect of pore-water salinity on the electrical resistivity (ER) of different compacted clay liners (CCLs) in terms of its mineralogical composition. For this purpose, an experimental programme was conducted where ERs of different kaolin-dominant CCL specimens, reconstituted using water having different concentrations of NaCl (0M, 0.5M, and 1.0M), were measured. /e kaolin-dominant CCL specimens tested in this study include pure kaolin, three different kaolinbentonite mixtures, and three different kaolin-sand mixtures. /e experimental results show that the ERs of CCL specimens decrease as the salt concentrations in pore water, moisture content, and dry density increase. At constant density and moisture content, the test results also indicate that increasing the sand content in kaolin-dominant CCL specimens increases its ER regardless of the water salinity level. /is behaviour could be attributed to the lower surface conduction of sand compared to kaolin. However, at constant density and moisture content, increasing the bentonite content in kaolin-dominant CCL specimens decreases its ER in the distilled water environment as surface conduction of bentonite is higher compared to that of kaolin. On the contrary, in saltwater environments, ER increases as the bentonite content increases./is behaviour could be explained in terms of the expected aggregated microstructure of bentonite in the saltwater environment that could reduce the number and area of interparticle contacts, and consequently, increase the ER of CCL specimens.


Introduction
Landfill facilities that usually have a cover and liner system are constructed to store the municipal solid wastes (MSWs) to protect the environment [1].e compacted clay liners (CCLs) in the liner system are constructed as a layer of hydraulic barrier to separate the leachate or toxic pollutants from the MSW into the subsoils or groundwater, to ensure good conditions of the subsoils [2][3][4][5][6].e CCLs are constructed of the natural soils or by combining them together with some other geosynthetics (e.g., geomembrane, geotextiles, and geosynthetic clay liners) based on different hazardous levels of MSW [7].e requirement for CCLs is to have a hydraulic conductivity less than 10 −9 m/s [8].Among these natural soils to form the CCLs, bentonite is considered as an excellent material to have low hydraulic conductivity ranging 10 −11 −10 −12 m/s [9,10].However, due to the unavailability in some regions and strong volume change after swelling/shrinkage [8,11,12], other soils to satisfy the same requirement for the CCLs are also being used [13][14][15][16].One example to replace bentonite is to add a specific percentage of clay (e.g., kaolin) or sand (e.g., silica sand), so the CCLs in the liner system will have less deformation [12].In fact, the suitability and advantages of some man-made clay liners with kaolin-bentonite (K-B) or kaolin-sand (K-S) mixtures with different compositions have been reported in some research [13,17].ey have shown that the additions of kaolin and sand will reduce the swelling capacity and improve the mechanical behaviour of the liner system [14,[18][19][20].
Leachate with different chemical compounds from MSW is a potential issue that will compromise the CCLs, resulting in the subsoils and groundwater significantly contaminated [8].In fact, monitoring the performance of CCLs and leachate is important to the environment and community; thus an observation for CCLs maintained in good condition is critical.ere are several geophysical methods available to be used such as electrical resistivity imaging (ERI), seismic refraction, ground penetrating radar, and multiple-channel analysis of surface waves.Among those techniques, ERI is recommended to evaluate the homogeneity of CCLs as the technique is an e cient method since it diminishes the time and cost constraints [21,22].e ERI technique is used to perceive the continuous changes in the ER of the geomaterials due to the presence of contaminants at a large or small portion.As a result, the relationship between ER of CCLs and other physical properties should be established.Several studies have reported that soil ER can also be controlled by soil mineralogy, pore-water salinity, particlesize distribution, the impact of compaction, the degree of saturation, and porosity [23][24][25][26][27][28].However, investigations on the ER of kaolinite-dominant clay liners a ected by di erent salt concentrations in water have been insu cient.e present paper aims at analysing the in uence of pore-water salinity on the ER of CCLs at di erent compositions.

Characterisation of Materials.
Two types of clay, namely, kaolin (K) and bentonite (B), and one type of sand (S) were used in this study to constitute seven di erent soil mixtures in this study as listed in Table 1.Table 2 provides the geotechnical properties of clays, obtained from [26,27], whereas Figure 1 illustrates the particle-size distribution of the clays (laser di raction method) and sand [29].e maximum and minimum void ratios of the sand were 0.97 and 0.58 [30,31].e liquid and plastic limits of clays are based on [32].

Sample Preparation and Experimental Setup.
For each CCL mixture, a known mass of dry soil mixture was mixed with a speci c amount of water to obtain the desired moisture content.Four to ve di erent moisture contents were targeted in this study ranging between 5% and 25%.To investigate the e ect of pore-water salinity on ER, three di erent NaCl-water solutions (0.0 M, 0.5 M, and 1.0 M) were used in this study.An automatic mortar mixer was used to guarantee the proper mixture of soil specimen and water.To ensure uniform distribution of moisture, water was sprayed on each specimen to reach the targeted water content.After the mixing procedures, each soil mixture was carefully placed inside air-tight bags at room temperature for 24 hrs for moisture equilibrium, in compliance with some relevant research [33,34].After 24 hrs, the soil mixture was placed into a ring, which has a diameter of 50 mm and a height of 20 mm.en, the top of the test specimen was properly levelled.
To get di erent dry densities of the test specimens, static compaction using a displacement-controlled loading method was applied in this study as shown in Figure 2.Under the static compaction, a negligible discrepancy in density along the sample's height can be achieved if the specimen's diameter is greater than the doubled height [35].
e ring used in this study satis es that condition.e ring, including the test specimen, was placed in a loading frame where the upper surface of the specimen was subjected to a constant displacement rate of 0.2 mm/min via a top loading steel cap as presented in Figure 2. Once the targeted height was achieved, the loading procedure was stopped.e specimen was left under a constant volume   Advances in Materials Science and Engineering condition for around 1.0 hr to ensure the homogeneity and stress relaxation of the specimen and to minimise its elastic rebound during the unloading step.After the unloading, the ring was carefully taken out of the frame.In the next step, the nal height, ER, and the total mass of the specimen were measured.e Wenner four-electrode method was applied to measure the ER of test specimens (Figure 3) [36], which has been discussed in the succeeding section.Once the measurements were done, the ring was repositioned into the loading frame to obtain the next targeted height.is process was repeated four times to obtain ER-dry density relationship of the tested soil mixtures at certain moisture content.Since the moisture content was kept constant, the degree of saturation increased as the height decreased under static compaction process.e experimental programme in this study produced over 400 readings of ER for test specimens having di erent soil mixtures, moisture contents, dry densities, and pore-water salinities.Figures 4 and 5 illustrate the range of dry density, water contents by mass (% W c ), degree of saturation (%Sr) of the di erent test specimens in this study.

Measuring Electrical Resistivity.
e ER of test specimens was measured using the Wenner four-electrode method [36].For this purpose, the steel loading cap used for the static compaction purpose was replaced by a modi ed plastic cap containing four copper electrodes (two outer and two inner electrodes, each 0.8 mm in diameter) as shown in Figure 3.As per the con guration, the electrical current ows through the two outer electrodes and the voltage drop was monitored by the inner electrodes.It has been reported in the literature that four-terminal pair con guration deals with perturbation better than the two-electrode method, as the former is capable of reducing the electrical interference and the electrode polarisation [37].
According to the Wenner four-electrode method, the four electrodes are aligned linearly with equal spacing, a, from each other (approximately 5 mm) and a protrusion of 1 mm into the soil specimen.A voltage was applied between the outer electrodes, and the corresponding current, I, and the voltage drop between the inner electrodes, V, were calculated.e calculated values were computed in the ER measuring equation, which is as follows: e Wenner four-electrode method provides an average ER of a hemisphere of space within the test specimen where the radius of hemisphere space is approximately proportional to the electrode spacing and the term 2πa is a geometric factor de ned based on a semiin nite boundary condition (half space).e impact of geometric factors is decisive to con rm the precise measurement of ER.To calibrate the experimental setup used in this study, the test cell was lled at di erent heights with two reference salt concentrated water solutions of known ER.Similar approach was used before by Suits et al. [37].e calibration results showed that the measured ER is quite identical to that of reference salt solutions, up to 10 mm of height.erefore, an accurate measurement of ER can be made, using the developed cell in this study, provided the thickness of the test specimen is greater than 10 mm.

Results and Discussion
Figure 6 shows the typical test results in the 3D domain of electrical resistivity (ER), water content (W c ), and dry density (ρ d ) for k7s3 and k7b3 where di erent water in terms of the salinity levels (0.0 M and 0.5 M) were used.e results in Figure 6 indicate that a constitutive electrical resistivity surface (CERS) in the domain of ER-W c -ρ d can be determined for each CCL mixture where the range and the geometric con guration of CERS are functions of the CCL composition and the pore-water salinity.In general, the typical results in Figure 6 show that, regardless of CCL composition and porewater salinity, ER decreased nonlinearly as the water content and dry density increased.Furthermore, as the ne content/ plasticity of CCL and/or pore-water salinity level increased, ER decreased as seen in Figure 6.
To understand the e ect of pore-water salinity on ER of di erent soils in terms of its composition, moisture content, and dry density; a parametric sensitivity analysis was conducted using the extensive experimental results of this study.For this purpose, a statistical approach was used to model the di erent generated CERS in this study where the minimum correlation coe cient (R 2 ) of the models was >0.98 as presented in Figure 6.In the following sections, a conceptual understanding for the possible e ect of changing pore-water salinity on ER of CCL will be discussed.
en, the  experimental evidence on the e ect of pore-water salinity on ER results, from this study, will be presented and discussed taking into consideration the possible role of mineralogical composition of CCL.

Conceptual Understanding for the E ect of Pore-Water
Salinity on ER of CCL.ER of clays is a function of water content and salinity, pore-water connectivity, surface conduction of clay particle, number and area of interparticle contacts (microstructure con guration and dry density), and temperature [26,27].It is known that as the salinity of pore water increases, its ER decreases and consequently ER of clay should also decrease.However, it is believed that the e ect of pore-water salinity on ER goes beyond this direct e ect.Hasan et al. [27] reported that the surface conduction of clay particle is a function of clay plasticity and salinity level of pore water.
e surface conduction increases as the clay plasticity and pore-water salinity increase.
erefore, the contribution of surface conduction into ER of clay increases and brings ER further down as the salinity level of pore water increases.
Several studies [5,[38][39][40] reported that the salinity of the water could a ect the microstructure con guration of compacted clay liner.In fact, the microstructure con guration of CCL is controlled by the magnitude of developed interparticle repulsive forces during the hydration process of clay particles [11].
e hydration-induced interparticle

4
Advances in Materials Science and Engineering forces depend on the available water content, salinity of water, and mineralogy of clay particles [41,42].For partially saturated clays, as the salinity of water decreases and water content and clay plasticity increase, the interparticle repulsive forces increase [43].Several researchers showed that reconstituting a clay specimen with distilled water could produce a monomodal pore size microstructure (uniform pore size) whereas aggregated microstructure (bimodal pore size) is expected as the salinity of pore water increases [5,38,44,45].erefore, for clay specimens reconstituted at constant water content and dry density but using di erent water in terms of its salinity level, a di erent number and area of interparticle contacts are expected.Consequently, ER of these specimens will also be a ected.

E ect of Pore-Water Salinity on ER of Pure Kaolin
Specimens.e results in Figures 7 and 8 depict the evolution of ER of K10 CCL specimens as pore-water salinity, moisture content, and dry density change.e results indicated that, as the pore-water salinity, dry density, and moisture content increased, ER decreased nonlinearly.ese results can be interpreted in terms of the increase in electrical conductivity of pore water as its salinity increased, the expected increase in the number and area of contact between the soil particles as the dry density increased which enhanced the ow of the electrical current through the moist soils, and the increase in pore-water connectivity as the water content increased which o ered more conductive channels through the soils.Advances in Materials Science and Engineering erefore, at constant moisture content and dry density, the di erences observed in ER of the specimens could be explained in terms of the change in electrical conductivity of pore water and the microstructure con guration of clay specimen, as water salinity increases.

E ect of Pore-Water Salinity on ER of K-S Specimens.
Under constant water content and dry density conditions, replacing the kaolin in the pure kaolin specimen (K10) with coarse soil (sand) and/or changing the salinity of pore water should change ER of CCL as shown in Figure 9.In fact, two mechanisms can be proposed to explain the observed ER change in Figure 9. ese mechanisms consider the expected change in surface conduction and the availability of water for clay particles as the mineralogical composition of CCL and pore-water salinity change.From the mineralogy perspective, as kaolin has higher surface conduction than sand, increasing the content of sand in CCL will reduce the overall surface conduction contribution of CCL and consequently 6 Advances in Materials Science and Engineering increase its ER.On the contrary, as the speci c surface area of sand and its electrochemical surface activity is much smaller than kaolin, replacing the kaolin with sand will involve releasing of the adsorbed water of the replaced kaolin in the pore space. is released water will increase the water availability for the remaining kaolin clay particles to enhance its hydration level which should improve its surface conduction level.Furthermore, this released water can also play a role in improving the pore-water connectivity.Consequently, this behaviour could decrease ER as sand content increases.
In conclusion, the overall ER change behaviour as sand content increases is controlled by the relative contribution of the two mechanisms.e result in Figure 9(a) reveals that for low water content (15%) and distilled water (0 M NaCl) cases, as the sand content increases, the contribution of surface conduction change and water availability mechanisms were almost equal.erefore, ER did not change as the sand content increased.However, for the distilled water case, as the water content increased (from 20% to 25%), ER increased as sand content increased, as shown in Figures 9(b) and 9(c). is behaviour elucidates the existence of a critical water content which is enough to fully hydrate the clay particles and establish primary water channels through the pores.So, if the soil specimen holds water content greater than this critical value, the released water due to the replacement process will have an insigni cant role in enhancing the surface conduction level of the clay particle and will also play a secondary role in improving pore-water connectivity.Consequently, the contribution of the water availability mechanism will be less than the contribution of surface conduction mechanism due to the replacement process and ER increased as the sand content increased.
It should also be mentioned that, for partially saturated clay specimen in saline water environment, the released water due to the sand replacement process could enhance the chemically induced microstructure change that depends Advances in Materials Science and Engineering on the hydration level of the clay particle.As mentioned before, the clay particles tend to have an aggregated microstructure when it is hydrated by saline water.Consequently, as the hydration level increases, the aggregation level increases.Since the aggregated microstructure can be characterised as the bimodal pore size system, the existence of large pores could reduce pore-water connectivity and consequently leads to an increase in e results in Figure 9 for saline water cases show that the aggregation and surface conduction mechanism due to the replacement process dominated the ER change behaviour as sand content increased.erefore, for these cases, ER increased as sand content increased.
Figures 10(a e results indicated that increasing the dry density would decrease ER.However, the overall ER change behaviour as the sand content and pore-water salinity change stayed the same.

E ect of Pore-Water Salinity on ER of K-B Specimens.
e results in Figure 11 highlight the e ect of di erent porewater salinity levels on the evolution of ER of kaolin clay specimens as the bentonite content increased at di erent water content levels but constant dry density.e results show that, for the case of distilled water, ER decreased as the bentonite content increased regardless of the water content level.However, for the cases of saline pore water, increasing the bentonite content increased ER regardless of the water content level.is behaviour can also be explained in the light of the change in the mineralogy and the availability of water for clay particles as the bentonite content increased.
Compared to kaolin, bentonite clay has a higher speci c surface area and electrochemical surface activity.erefore, compared to kaolin, the bentonite holds a higher surface conduction and requires more water for its hydration process.At constant water content and density, the bentonite's higher surface conduction should lead to a reduction in ER as the bentonite content increases.However, the  Advances in Materials Science and Engineering bentonite needs higher water content for its hydration process, which should lead to an increase in ER due to the shortage in the water available for hydration process as the bentonite content increases.For a clay specimen with high water content, free water can be available in its pores.Nevertheless, when the kaolin is replaced with bentonite, at constant water content, this free water be used to supply the demand of bentonite hydration process which requires more water content than the kaolin.e use of the free water, in this case, will reduce the pore-water connectivity and should also increase ER.
e results in Figure 11 for distilled water also suggest that the overall increase in the surface conduction as the increase in bentonite content dominates the observed behaviour.Consequently, ER decreased as bentonite content increased.However, the increase in ER as the bentonite content increased for cases of saline pore water can be attributed to the shortage of water available for hydration process of bentonite and the possible aggregation microstructure which decreased connectivity of pore water.

Conclusions
e in uence of pore-water salinity on the ER of three types of kaolin-dominant CCL specimens (kaolin, kaolin-sand, and kaolin-bentonite) has been experimentally investigated.
e study was conducted by constituting CCL specimens at di erent NaCl concentrations of water (0 M, 0.5 M, and 1 M) corresponding to seven types of soil mixtures, di erent dry densities (ρ d ), water contents (W c ), sand/bentonite fractions, and degree of saturation (%Sr).A typical constitutive electrical resistivity surface (CERS) based on a statistical approach was established in order to understand the correlation between the ER, W c , and ρ d at di erent pore-water salinities and soil plasticity.e test results, in general, demonstrate that the ERs of the CCL specimens decrease nonlinearly with increasing dry densities, moisture contents, or pore-water salinities.However, at constant density and water content, the increased salt concentrations had an   Advances in Materials Science and Engineering insigni cant impact on conductivity as the ERs increased if the sand fraction increased in the tested specimens.is behaviour can be attributed to sand's lower surface conduction compared to that of kaolin.Consequently, bentonite has higher surface conduction than kaolin, and hence, the ERs of kaolin-bentonite declined abruptly if bentonite fraction increased concurrently in the CCL specimens, constituted using distilled water.Nevertheless, it was also found that ERs of the kaolin-bentonite increased surprisingly despite increasing bentonite fraction at 0.5 M and 1 M of NaCl concentrations.A conceivable explanation for this result can be interpreted in terms of the aggregated microstructure of bentonite (bimodal pore size) and shortage of water available for the hydration process of bentonite, which weakens the connectivity of pore water, leading to a reduction in the number and area of the interparticle contacts.

Data Availability
No data were used to support this study.e research is mostly experimental, which was validated by numerical presentation and theoretical explanation.Proper references have been cited throughout the research article.

Figure 2 :
Figure 2: Controlling dry density using the displacement-controlled loading system.
) and10(b)  exhibit the e ect of dry density change on ER change behaviour of K-S specimens as the sand content and pore-water salinity change.

8
) and10(d)  show the e ect of dry density change on ER change behaviour of K-B specimens as the