Advertisement

Environmental Science and Pollution Research

, Volume 26, Issue 18, pp 18316–18328 | Cite as

Fresh underground light non-aqueous liquid (LNAPL) pollution source zone monitoring in an outdoor experiment using cross-hole electrical resistivity tomography

  • Shuai Shao
  • Xiujun GuoEmail author
  • Chang Gao
Research Article
  • 49 Downloads

Abstract

Real-time monitoring of source zone of light non-aqueous liquids (LNAPLs) is important for preventing accidental pollution and taking effective underground pollution remediation measures. As a high-precision monitoring technology, cross-hole electrical resistivity tomography (CHERT) has been widely used to obtain static information regarding underground stratigraphic structures and dynamic information regarding fluid motion. The time-lapse data processing method can be used to obtain pollution zone dynamic change information. Since the leakage rate directly affects the diffusion range of pollution, this study simulated the initial evolution process of LNAPL pollution source regions under two different leakage rates. The time-lapse monitoring of the above process was performed using CHERT. The test results show that CHERT can be used to observe the migration of LNAPLs and the initial evolution of the contaminated area. Differences in leakage rate will result in variation in soil wettability and fluid distribution, which will cause changes in soil resistivity in the corresponding region. The low-saturation LNAPL-contaminated area may exhibit low-resistivity characteristics and is easily overlooked in field investigations. In addition, the degree of contamination in the saturated zone can be quantitatively evaluated by CHERT; however, the pollution range and oil saturation value determined by CHERT are underestimated. The results showed the electrical variation characteristics of the initial evolution process of the fresh pollution source area and provide data that will enable early warnings of LNAPL leakage. This shows that CHERT is a promising tool for monitoring LNAPL pollution source leakage even if further research is needed to fully understand the effect of hydrological processes on electronic signals.

Keywords

LNAPL pollution Real-time monitoring Wettability Cross-hole electrical resistivity tomography Quantitative assessment 

Introduction

The migration of light non-aqueous liquids (LNAPLs) in the soil is complicated, and the process from the beginning of a LNAPL leak to the formation of pools and lenses is difficult to directly observe. The pools and lenses of LNAPLs become long-term sources of pollution and continuously release pollutants into the groundwater (Zhou and Cardiff 2017). The formation of the pollution source zone is directly related to the leakage condition. The larger the leakage aperture and the leakage speed, the wider the diffusion range (Chen et al. 2016). In addition, due to the physical properties (e.g., density and viscosity) of LNAPLs and soil physical parameters (e.g., permeability, water storage coefficient) and other physical factors, it is difficult to predict the distribution and migration of LNAPLs. Description of the migration process after LNAPL leakage and the scope of pollution will play an important role in the remediation of subsequent contaminated sites (Hu et al. 2010). The traditional contaminated site survey relies mainly on sparse drilling and sampling analysis, and can only provide discrete data (Peter et al. 2008). A tracer test can be used depict the distribution of contaminated areas and related saturation information (Illman et al. 2010; Rao et al. 2000; Schubert et al. 2007; Yeh and Zhu 2007). Geophysical methods have been increasingly applied in the investigation of underground LNAPL pollution, as they can provide underground continuous profile information under non-invasive conditions (Cassiani et al. 2014; Halihan et al. 2017).

When the soil is contaminated with LNAPLs, the physical properties will change. The electrical resistivity is the basic physical index of the soil and has been applied to determine whether or not pollution has occurred (Atekwana and Atekwana 2010; Hamzah et al. 2009; Godio et al. 2010). However, the relationship between the contaminated area and resistivity is complex. When the contaminated zone exhibits high resistivity, it reflects the electrical properties of the hydrocarbon as an electrical insulator. The insulating hydrocarbon layer floats on the water table to produce a shadowing or a screening effect on the resistivity measurements (Rosales et al. 2014; Sadowski 1988). The reasons for the decrease in resistivity in contaminated areas, which are more complicated, include the following: (i) When hydrocarbons invade the soil, they wash out the salt on the surface of the particles, causing briny halos around the hydrocarbons (Andres and Canace 1984; Atekwana et al. 2004; Delgado-Rodríguez et al. 2006; Sauck 2000; Shevnin et al. 2003; Waxman and Thomas 1974); (ii) biochemical processes lead to increased organic acids and enhanced mineral weathering (Atekwana and Atekwana 2010); and (iii) changes in oil wettability (Revil et al. 2011).

Electrical resistivity tomography (ERT) has been recognized as an effective technique to characterize the contaminated areas of LNAPLs (Barker et al. 2001; Thabit and Khalid 2016). However, due to geological heterogeneities, the application of surface ERT is limited by static detection, which may result in a greater change in resistivity than the change in resistivity associated with LNAPLs (Cardarelli and Filippo 2009; Perri et al. 2012). Cross-hole electrical resistivity tomography (CHERT) has been increasingly applied to studies of fluid migration (Slater et al. 2000; Franco et al. 2009) and saturation distribution (Slater et al. 2010) in hydrological testing because of its higher resolution. With the development of time-lapse inversion algorithms (Karaoulis et al. 2014; Orlando and Renzi 2015), CHERT has great potential for monitoring changes in underground pollution areas.

Compared with groundwater, non-aqueous liquids have high-resistivity characteristics. Good results have been achieved in past studies through the application of CHERT to the monitoring of contaminated areas of dense non-aqueous phase liquids (DNAPLs) (Power et al. 2015). The invasion of DNAPLs can displace water, causing changes in soil resistivity (Tomlinson et al. 2017). The LNAPLs often occupy the original pore space during the migration in the vadose zone and can discharge a certain amount of water under sufficient oil pressure. The electrical resistivity was not much different from that of pre-contaminated soil (Atekwana et al. 2000; Grumman Jr and Daniels 1995), so the monitoring process is more difficult. In a survey of sites contaminated with crude oil in Italy, CHERT was used to establish a connection between a contaminated area and a specific electronic signal at a small scale. The pollution area could not be directly identified from the CHERT profile without time-lapse monitoring (Cassiani et al. 2014). In the northwestern part of Italy, combined with geophysical, geochemical monitoring, and biological analysis, a small-scale description of an aged LNAPL-contaminated area was conducted and it was found that the application of CHERT could identify the contaminated area at the top of the saturated area (Arato et al. 2014). CHERT also has practical limitations, for example, in order to achieve sufficient resolution, the spacing between vertical holes is limited by the depth of the hole (Power et al. 2015).

The water wettability of the soil changes following the intrusion of a large number of LNAPLs (Quyum et al. 2002). The change of wettability affects the distribution of oil–water two-phase materials in the soil (Wu et al. 2008), resulting in corresponding changes in electrical resistivity (Revil et al. 2011). This process is difficult to detect during the static detection of ERT in oil-contaminated areas. For potential LNAPL leakage areas, time-lapse monitoring can be used to monitor LNAPL leaks and the evolution of the pollution source zone by deploying multiple CHERT devices. The overall goal of this research was to apply CHERT to study the process of diesel from leakage to the formation of contaminated lenses on the surface of the water table, and to determine whether the CHERT and time-lapse inversion algorithms have sufficient sensitivity to describe the contaminant hydrological processes. The characteristics of the migration process of LNAPL were visualized, and the law of resistivity change during the migration process was summarized. At the same time, of particular interest is the change in resistivity caused by changes in soil wettability during this process, a feature not seen in the literature on past ERT detection of fresh LNAPL contamination. Combined with the Archie formula, CHERT was here used to assess the degree of soil pollution.

Materials and methods

Experiment setup and materials

The experimental sand tank (1.5 m long, 1 m deep, and 1.2 m wide) was placed outdoors and filled with clean river sand (Fig. 1). The particle size composition and physical properties of the sand are shown in Table 1. Layered filling and compaction took place during backfilling of the sand tank. A membrane of high-density polyethylene material was laid on the bottom and 50 cm up the sidewalls of the sand tank. After the sand layer was laid, water was injected into the sand tank, and the water level in the sand tank was monitored using the monitoring tube. Injection was ceased when the water level was 0.7 m below the ground. In order to determine soil moisture content at different depths within the background soil layer, soil samples at different depths were obtained by drilling, and the moisture content of the soil samples was measured using the drying method. Soil water saturation at different depths were calculated (Table 2). The LNAPL used in the experiments was diesel, a typical material for organically contaminated sites (Table 3). The injection point for simulated point source diesel leakage was placed 0.2 m below the surface. A peristaltic pump was used to simulate the intrusion process of infiltration rate of 30 ml/min and 150 ml/min. The different injection speeds will affect the fluid migration state and the distribution of water in the soil.
Fig. 1

(a) Sketch of the experimental setup. (b) Photograph of the experimental apparatus. The triangle in (a) is the positions of the sampling points, and the diesel content and water content in the soil sample are measured

Table 1

The particle size composition and physical properties of the sand

Density (g cm−3)

Specific gravity

Mean grain size (%)

< 0.075

0.075-0.25

0.25-0.5

0.5-1

1–2

> 2

1.88

2.61

2

10.4

9.6

19.4

27.3

31.3

Table 2

Measured water saturation at different depths of the background soil layer

Depth (m)

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

Saturation

0.31

0.33

0.33

0.37

0.52

0.63

0.76

0.81

Table 3

Fluid properties of diesel and water used in the experiments

Properties

Diesel

Water

Density (g cm−3)

0.842

1.000

Viscosity (cP)

3.94

1.003

Interfacial liquid tension (N m−1)

0.0172

First, experiment 1 with a diesel injection rate of 30 ml/min was implemented (Injection tube 1). The injection time was 2 h, and the continuous observations lasted for 16 h. After stopping the measurements, samples were taken at the sampling points shown by triangles in Fig. 1(a). Subsequently, experiment 2 was conducted using an injection rate of 150 ml/min (Injection tube 2). The injection time was also 2 h, and the continuous observations lasted for 16 h (Fig. 1(a)). Two samples at each position were recorded as A and B. Volatilization of diesel during drying would cause errors in measurement of the diesel and water content; therefore, the diesel drying loss coefficient was introduced to correct the traditional drying method to determine the diesel content and water content of the diesel-contaminated soil (Liang et al. 2011; Pan et al. 2018). The samples were dried at 105 °C for 24 h and then cooled to a constant weight in a dry box. Based on the experimental sand and diesel, the drying loss coefficient r was measured. The drying loss coefficient is the ratio of the amount of volatilized diesel to the total diesel in the sand before drying. The formula for determining the diesel content and water content is as follows:
$$ n=\frac{m_0}{\left(1-r\right)\left({m}_r-{m}_0\right)} $$
(1)
$$ w=\frac{m_t-{m}_r-{m}_0\left(\frac{r}{1-r}\right)}{m_r-{m}_0} $$
(2)
where n is the diesel content of contaminated soil samples, w is the water content of contaminated soil samples, mt is the mass of the contaminated sand sample, mr is the mass of the contaminated sand sample after drying, and m0 is the mass of the diesel in the dried sample, which was measured by extraction and UV spectrophotometry (UV2550, Shimadzu Corporation). The diesel content and water content of sample A and B indicated the values of the same sampling point. To reduce the measurement error, the diesel content and water content measured by sample A and sample B were averaged.

Acquisition and processing of CHERT data

For cross-hole data acquisition, three insulated probes were installed in the sand tank. Each probe was equipped with 20 electrodes, the electrode spacing was 0.04 m, the probe spacing was 0.65 m, and the first electrode of the probe was 0.1 m from the surface (Fig. 1). Unlike ground ERT, there is no air-soil interface in CHERT. The subsurface ERT measurement is an array of electrodes on the surface of a half-space. For this measurement, the geometric factor is given by k = 2πa, where a is the spacing between the electrodes. The electrode arrays of the CHERT are located in an infinite medium underground, and the geometric factor is given by k = 4πa (Loke 2004). When the surface ERT is measured, the accuracy will decrease with the depth of detection. To obtain more accurate results, the sensor can be placed underground near the target area (Loke 2004; Zhao et al. 1986).

CHERT has a variety of array optimization methods, including pole–pole, pole–bipole, bipole–pole, and bipole–bipole (Daily and Owen 1991; Demirel and Candansayar 2017). The earliest and most widely used is the pole–pole array (Daily and Owen 1991; Shima 1992). However, since there are two additional remote electrodes in the pole–pole array, they must be placed away from the study area, and systematic errors are introduced during the inversion (Zhou and Greenhalgh 2000). In contrast, pole–bipole, bipole–pole, and bipole–bipole arrays have higher resolution and can eliminate the effects of remote electrodes (Sasaki 1992; Wilkinson et al. 2010; Zhou and Greenhalgh 2000). However, if a current pole and a potential pole coexist in the borehole, a shadow effect and a symmetric effect will be produced as a result of a highly symmetric sensitivity pattern (Tsourlos et al. 2011). When monitoring the LNAPL rapid migration process, acquisition time becomes critical. The acquisition time is directly related to the amount of data (Demirel and Candansayar 2017). The pole–bipole, bipole–pole, and bipole–bipole arrays collect more data than the pole–pole array. Although more data volume will increase resolution, the acquisition time must be adjusted to the length of the LNAPL dynamic process (Bellmunt et al. 2016). Considering the advantages and disadvantages of each array and the suitability of monitoring the LNAPL migration process, the electrode configuration used in this experiment was pole–pole (Loke 1999). The remote electrodes were placed at a distance of 30 m from the study area.

The electrode was sequentially supplied as the power supply electrode A, the remaining electrode was observed as the measurement electrode M, and the observed potential was converted into the apparent resistivity. There was no boundary in the upper part of the sand tank, and the influence of the lower half of the sand tank was neglected in the experiment. Time-lapse ERT measurements describe dynamic changes in subsurface properties and determine the extent of subsurface regions affected by a particular process (Baker and Moore 1998; Cassiani et al. 2006; Oldenborger et al. 2007). Time-lapse CHERT datasets were collected every few hours prior to diesel injection to identify anomalies that may be associated with the acquisition procedure, equipment, or bad electrode contacts. The schedule of time-lapse CHERT measurements is presented in Table 4.
Table 4

Schedule of time-lapse CHERT measurements

Data collection moment

Pollution source area condition

Before diesel filling

Clean soil

40 min after the beginning of diesel injection

Pollution source area state 1 under diesel fluid pressure

80 min after the beginning of diesel injection

Pollution source area state 2 under diesel fluid pressure

120 min after the beginning of diesel injection

Pollution source area state after diesel injection

2 h after completion of injection of diesel

Natural evolution state 1 of pollution source area

6 h after completion of injection of diesel

Natural evolution state 2 of pollution source area

16 h after completion of injection of diesel

Pollution source area state at the end of monitoring

The resistivity profiles acquired were separately inverted, and the acquisition profile of the cleaned background layer before diesel injection was used as a reference model. It is important to note that applying this method requires that all time-lapse data be collected using the same inversion method. The inversion process performed by RES2DIN is based on a least squares algorithm (Loke and Barker 1996), allowing us to obtain a two-dimensional distribution of actual resistivity data. The root-mean-square (RMS) obtained in each individual inversion ranged from 0.6 to 1.7%. In order to show the subtle underground changes in the diesel injection process, we perform data ratio inversion or normalization (Daily et al. 1992):
$$ R=\frac{R_t}{R_0}{R}_h $$
(3)
where Rt is the resistivity measured at time t, R0 is the background resistivity and Rh is the resistivity for a homogenous (reference) resistivity distribution model, Rh is 10 Ω·m in this calculation process. The resistivity ratio is multiplied by Rh to obtain the normalized datum. The inversion of the normalized data set produces an image of the resistivity changes relative to the homogenous model. The changes in resistivity are expressed as the resistivity ratio with respect to the background. A resistivity ratio greater than 1 indicates that the model becomes more resistive than the reference model, and a resistivity ratio less than 1 indicates a decrease in resistivity.

Results

When the diesel was injected at a rate of 30 ml/min, the diesel moving front maintained a regular arc-shaped movement, and under the action of capillarity, the lateral direction also expanded. CHERT captured the real-time motion of the diesel plume. The resistivity ratio profile reflects the diesel intrusion area and the direction of migration (Fig. 2). The more regular arc-shaped low-resistivity areas are shown in Fig. 2(a), (b). As the amount of diesel continues to increase, the arcuate edge moves downward (Fig. 2(c)). After the diesel injection was stopped, the diesel continued to diffuse downward and horizontally under gravity and capillary action within 2 h, and the diffusion area was larger than before (Fig. 2(d)), a wider range of low-resistivity areas appeared. The resistivity of the contaminated area gradually increased with time (Fig. 2(e)). All leaked diesel was fixed in a certain space, and after 16 h, the resistivity of the contaminated area increased (Fig. 2(f)). As shown in Fig. 2, the resistivity decreases in the beginning and then increases.
Fig. 2

CHERT monitoring data of diesel injected into the sand at a rate of 30 ml/min. Data presented in (a)–(c) were acquired after the beginning of diesel injection: 40 min, 80 min, and 120 min for (a)–(c), respectively. Data presented in (d)–(f) were acquired after completion of injection of diesel: 2 h, 6 h, and 16 h for (d)–(f), respectively. The values are expressed in resistivity ratios with respect to background

When diesel was injected into the soil at a rate of 150 ml/min, the main transporting force of the diesel came from the fluid pressure due to the large initial velocity of the leak. Diesel was blocked at the point of leakage, and the fluid pressure rose sharply, causing the diesel to spread. The migration rate in the vertical direction was greater than that in the horizontal direction. After 40 min, the pollution plume of the diesel reached the water level, and the resistivity of the diffusion area increased (Fig. 3(a)). Different forms of diesel masses are formed at different stages of migration and migrate as a whole in the form of oil masses (Newell et al. 2011). This migration process is apparent in the Fig. 3(a) section as a region of three trapped high-resistivity anomalies. Subsequently, a large trapped high-resistivity anomaly region was formed and moved downward (Fig. 3(b)–(d)). Diesel invades the soil–water–air three-phase system and migrates. Water and diesel are immiscible, and multi-phase seepage occurs. The infiltration of diesel causes displacement of water, resulting in redistribution of water in the soil (Schroth et al. 1998). The displaced water is collected in the direction of diesel migration, and the stripe-like anomaly decreases in resistivity. Below the water table, the resistivity decreased due to the increase of water content (Fig. 3(a)–(d)). Diesel continued to migrate to the capillary fringe under fluid pressure and gravity. The capillary water was displaced by diesel, the capillary fringe disappeared, and the diesel floated on the water table to form a contaminated lens.
Fig. 3

CHERT monitoring data of diesel injected into the sand at a rate of 150 ml/min. Data presented in (a)–(c) were acquired after the beginning of diesel injection: 40 min, 80 min, and 120 min for (a)–(c), respectively. Data presented in (d)–(f) were acquired after completion of injection of diesel: 2 h, 6 h, and 16 h for (d)–(f), respectively. The values are expressed in resistivity ratios with respect to background. The longer two black lines in (f) indicate the position of the borehole BH5 and BH7 for measurement of moisture content and diesel content. The red dotted line area in (f) is the study area associated with saturation exponent n.

After the diesel injection was stopped, the diesel in the upper part of the soil gradually moved downward under the action of gravity, and the diesel content of the soil was continuously reduced for the remainder of the study period. In the soil below the leak point, which was where the maximum amount of diesel infiltration occurred, the residual diesel saturation of the soil after 16 h of redistribution was larger than at any other point. The high residual diesel saturation region exhibited high-resistivity characteristics (Fig. 3(e), (f)). Finally, a stable contaminated lens of diesel formed at the position of the water table, and the resistivity increased by up to 300%. As shown in Fig. 3(e), (f), the diesel-contaminated area in the middle and right side of the section (red dotted line area in Fig. 3(f)) changed from a high-resistivity area to a low-resistivity area compared with the background value.

Discussion

Fluid distribution state and resistivity characteristics

Wettability refers to the comprehensive characteristics of the soil–fluid system. It describes the tendency of two immiscible fluids to adhere and adsorb on the solid surface. The surface wettability of aqueous media has an essential effect on factors such as capillary pressure, residual oil or water saturation, and relative permeability, and is an important factor influencing the microscopic distribution of fluids in the pores. The experimental coarse sand was clean river sand and belongs to the natural hydrophilic medium. The water saturation of the soil below 0.2 m underground was greater than 0.3 (Table 2), which belongs to the water-wet system. The water is mainly filled with small pores and adsorbed on the surface of the soil particles to form a wetting film, which plays a leading role in the conductivity of the soil (Mungan and Moore 1968; Suman 1997). There was almost no difference between the measured moisture content of the contaminated soil layer and the moisture content of the background soil layer after experiment 1. For diesel intrusion at a rate of 30 ml/min, the diesel fluid pressure was relatively small and was unable to displace the moisture attached to the soil particles; thus, the diesel mainly occupied the air phase portion of the soil pores. The water originally distributed in the large pores did not participate in conduction, but this part of the water was then pushed by diesel into a part of the wetting film, resulting in enhanced conductivity in the invaded area. Therefore, the area invaded by diesel shows a low resistivity ratio in the profile (Fig. 2(a)–(d)). In the current study, after the diesel invaded the large pores, it did not stop the migration. Since the soil did not reach the condition of complete wettability, the soil particles were not completely wrapped by water, and there was a position for the diesel to adsorb. Over time, the diesel gradually contacts, infiltrates, adheres to the soil particles, and migrates at the water–gas interface (Weber and Huang 1996), which changes the state of the continuous wetting film, and the regional resistivity increases. Figure 4 is a contour plot of measured values of soil diesel saturation after completion of monitoring in Experiment 1 (16 h after completion of injection of diesel). As shown in Fig. 4, the actual distribution of diesel is larger than the area identified by CHERT.
Fig. 4

Soil diesel saturation measured by drilling sampling 16 h after completion of injection of diesel (experiment 1). The triangles in the figure are the measured data point with a value other than zero

For diesel intrusion at a rate of 150 ml/min, the main driving force of diesel migration is oil fluid pressure and gravity. During the migration process, part of the water can be displaced, the water content is reduced, and the electrical resistivity is increased. As the diesel content continues to increase, the medium is embodied as an oil–wet system (Lowe et al. 1999; Quyum et al. 2002). The diesel adsorbs on the surface of the soil particles and fills the small pores. The original continuous conductive wetting film becomes broken, and the undisplaced water is distributed in the large pores in an isolated state (Suman 1997), no longer participating in conduction, and the regional resistivity continues to increase (Fig. 3(a)–(d)).

Variation of saturation exponent n and corresponding resistivity characteristics during diesel pollution

In situ subsurface fluid saturation is usually obtained by the Archie formula (Archie 1942) and the default exponent of resistivity measurements. Studies have shown that saturation exponent n varies according to factors such as water saturation and wettability (Toumelin and Torres-Verdín 2005). In the experiment 2, 16 h after completion of injection of diesel, the soil samples were taken from the borehole to measure the water saturation and the diesel saturation. The water saturation and resistivity ratio data of BH5 and BH7 in Fig. 3(f) were extracted and compared with the background clean soil layer saturation (Table 2). Figure 5 presents the measured water saturation and resistivity ratio curves in the BH5 and BH7 positions. The water saturation of both boreholes below 0.2-m depth was reduced. The position of BH5 was closer to the leakage point, and thus the fluid pressure was larger and more water was displaced. Compared with the background, the water saturation was greatly reduced in the range of 0.2–0.6 m underground. According to the resistivity ratio data, the peak value was reached at the depth of 0.4 m, and the resistivity ratio had a good correlation with the diesel saturation at the BH5 position. The water saturation at the BH7 position was higher than that at BH5, but lower than the water saturation of the background soil. The intrusion of diesel also caused the water to be displaced at BH7. However, the resistivity ratio data was less than 1 in the depth range of 0.2 to 0.6 m in the ground, that is, lower than the background clean soil layer. From the measured data of diesel saturation in Fig. 6, it can be seen that the diesel saturation at this depth was about 0.1–0.2. A large amount of diesel intrusion may lead to changes in soil wettability. Analysis was based on the resistivity formula developed by Keller and Frischknecht (1966) for unsaturated soils:
$$ \rho ={\rho}_wa{\phi}^{-m}{S}_w^{-n} $$
(4)
where ρw is the pore water resistivity; ϕ and Sw are the soil effective porosity and water saturation, respectively; a is the formation factor; and m and n are the pore cementation exponent and the saturation exponent, respectively. After diesel intrusion, the resistivity in the red dashed area is reduced compared to the background value (Fig. 3(f)). The resistivity characteristic analysis was based on Eq. (4); a, φ, and m are fixed values (Aguilera 2004); ρw does not change; and the water saturation is reduced compared to the background value (Fig. 5(b)). Under such conditions, only the saturation exponent n in Eq. (4) increased, and the red dot area resistivity decreased. The increase in n also means that the soil wettability has changed. In the oil–wet system, the saturation coefficient n is larger than that of the water-wet system (Donaldson and Siddiqui 1989; Mungan and Moore 1968). The zone (the red dashed area of Fig. 3(f)) exhibited a decrease of resistivity ratio compared with the background with a diesel saturation of 0.1 to 0.2 and a water saturation of 0.2. Although the residual diesel saturation was larger at the position of BH5, the saturation exponent n increased, but the water saturation was too low due to the displacement of high fluid pressure, which was characterized by an increase in electrical resistivity.
Fig. 5

Saturation and resistivity ratio acquired at (a) BH5 and (b) are BH7 positions (Fig. 4(f)). The water saturation of the background soil was used for comparative analysis

Fig. 6

Soil diesel measured by drilling sampling 16 h after completion of injection of diesel (experiment 2)

In experiment 2, it was proposed that the resistivity of the pore water did not change before and after the pollution, and the increase in the saturation exponent was inferred. However, we had not actually taken measurements to confirm the pore water resistivity. In previous studies, many scholars believed that the intrusion of insulating hydrocarbons washed out the salt originally attached to the soil so that the concentration of ions in the pore liquid increased, and the resistivity of the pore liquid decreased, resulting in a briny halo around the hydrocarbons (Andres and Canace 1984; Atekwana et al. 2004; Delgado-Rodríguez et al. 2006; Sauck 2000; Shevnin et al. 2003; Waxman and Thomas 1974). Therefore, the reasons for the decrease in the resistivity of the fresh pollution source area may be liquid distribution state, wettability, and briny halo around the hydrocarbon.

Estimation of diesel saturation

In experiment 2, the diesel saturation and the corresponding resistivity ratio of all sampling point positions were made as a scatter plot, as shown in Fig. 7. The whole pollution source zone was divided into three parts according to the initial water saturation in the vertical direction: L1 (0–0.4 m), L2 (0.4–0.6 m), and L3 (0.6–0.8 m) corresponding to the vadose zone, capillary fringe, and aquifer, respectively (Fig. 7).
Fig. 7

Scatter plot of measured diesel saturation versus resistivity ratio. L1, L2, and L3 represent data for the vadose zone, capillary fringe, and aquifer, respectively

Under the condition of low diesel saturation (0–0.2), the resistivity ratio range of L1, L2, and L3 was about (0.5–1.2). Due to the intrusion of a large amount of diesel to displace different amounts of water in different positions, the difference in initial water saturation did not reflect the obvious regularity of resistivity in the measurement. In the range of diesel saturation (0–0.2), most of the resistivity data decreased, and a small part of the resistivity data did not change or slightly increased. In the fresh pollution source zone, the increase in resistivity can be clearly recognized as the existence of high-resistivity organic matter, and the cause of the decrease in resistivity was more complicated, and it is difficult to directly confirm this region as a contaminated area.

Under the condition of high diesel saturation (0.3–0.55), most of the resistivity data of L1, L2, and L3 increased considerably. This part of the data is mainly the main migration channel of diesel with high residual diesel saturation and the contaminated lens of the saturated zone. Comprehensive analysis of the measured diesel saturation and measured resistivity data indicated a region of the unsaturated zone with diesel saturation greater than 0.3, showing a strip-like high-resistivity anomaly in the range of 0.05–0.2 m in the resistivity ratio profile (Fig. 3(f)). Contaminated lens of LNAPLs in the saturation region below 0.6 m can also be identified from the resistivity ratio profile (Fig. 3(f)).

Since the experimental porous medium was mainly coarse sand, the surface conductivity was very small (Gelius and Zhong 2010), and it was assumed that the saturated zone was > 0.6 m underground. Based on the above conditions, Archie’s second equation was used to calculate the saturated oil saturation (Sn) (Archie 1942):
$$ {S}_n=1-{\left(\frac{\rho_0}{\rho_t}\right)}^{1/n} $$
(5)
where ρ0 is the resistivity of the soil layer under the condition that the background soil layer is completely saturated by water, ρt is the resistivity measured in the saturated region, and n is the saturation exponent. For the clean sand sample without clay soil, the value is close to 2 (Archie 1942; Deng et al. 2016; Liang et al. 2016). The diesel saturation value calculated according to Eq. (5) and plotted on the contour plot is shown in Fig. 8(a), where ρt is the resistivity value measured 16 h after the completion of the injection of diesel. The measured diesel saturation and the calculated diesel saturation in the region below 0.6 m are shown in Fig. 8.
Fig. 8

(a) Calculated value of the diesel saturation of the saturated zone using the Archie formula. (b) The diesel saturation value of the saturated zone measured by the borehole sampling

It can be seen from Fig. 8 that the calculated diesel saturation value generally reflected the position of the lens. The calculated diesel contamination area was smaller than the actual pollution area. In diesel-contaminated soil layers, the saturation exponent is not always a fixed value and is related to the wettability of the soil. In the saturated sand layer, the influence of the change in the saturation exponent is still reflected in the calculated saturation values. When the calculated saturation value has a negative value (Fig. 8(a)), the soil resistivity is reduced compared to the background soil layer. The diesel saturation in this area cannot be correctly reflected. Based on clean sand and no clay, under full saturation conditions, we used n = 2.0; however, the real conditions are different, the presence of clay particles and soil wettability change, and thus calculated value would be lower than the measured value.

Conclusions

This study used CHERT to observe the migration of LNAPLs and the evolution of contaminated areas. In the sand medium, experiments were carried out on the two different intrusion rates of diesel to observe the migration of diesel in the sand and the initial evolution of the contaminated area. The results show that the resistivity of the contaminated area decreased under a low intrusion rate, and the resistivity of the evolution process increased with time. In contrast, LNAPLs rapidly migrate to the surrounding area under high-rate intrusion conditions. During the migration process, water is concentrated in the transport direction of LNAPLs, and the resistivity of the water accumulation area is reduced. LNAPLs migrated in the form of multiple oil masses at the beginning of the infiltration process, and then aggregated into a larger oil mass as a whole and migrated to the water level surface, eventually forming a contaminated lens with high resistivity characteristics on the water level surface. A large amount of LNAPL invading the soil will cause the wettability of the sand medium to change, and the saturation exponent n will increase, which will affect the measurement results. CHERT can identify contaminated areas with LNAPL saturation greater than 0.3 in the unsaturated zone and quantitatively evaluate the degree of diesel contamination in the saturated zone. When the LNAPL saturation is in the range of 0–0.2, the contaminated area may exhibit low resistivity features, which are easily overlooked in field ERT surveys.

This experimental study demonstrates the use of CHERT to monitor the evolution of fresh LNAPL pollution source zones with a time-lapse method. However, in order to improve the monitoring accuracy of CHERT and establish the relationship between reliable resistivity and saturation, the effects of hydrological processes and soil physical properties on electrical resistivity should be considered.

Notes

Acknowledgments

We thank LetPub (www.letpub.com) for its linguistic assistance during the preparation of this manuscript. We thank Yizhen Niu and Xiangyi Shao for their great support in the research.

Funding information

This research work is supported by the National Natural Science Foundation of China for development project of major scientific research instrument (Grant No. 41427803), National Natural Science Foundation of China (Grant No. 41772307), and National Key Research and Development Projects (Grant No. 2017YFC0307701).

References

  1. Aguilera RF (2004) A triple porosity model for Petrophysical analysis of naturally fractured reservoirs. Petrophysics 45:157–166Google Scholar
  2. Andres K, Canace R (1984) Use of the electrical resistivity technique to delineate a hydrocarbon spill in the coastal plain deposits of New Jersey. In: Proceedings of the NWWA/API Conference on petroleum hydrocarbons and organic chemicals in ground water-prevention, detection and restoration, Houston. pp 188–197Google Scholar
  3. Arato A, Wehrer M, Birã B, Godio A (2014) Integration of geophysical, geochemical and microbiological data for a comprehensive small-scale characterization of an aged LNAPL-contaminated site. Environ Sci Pollut Res Int 21:8948–8963.  https://doi.org/10.1007/s11356-013-2171-2 CrossRefGoogle Scholar
  4. Archie GE (1942) The electrical resistivity log as an aid in determining some reservoir characteristics. Soc Pet Eng 146:54–62.  https://doi.org/10.2118/942054-g Google Scholar
  5. Atekwana EA, Atekwana EA (2010) Geophysical signatures of microbial activity at hydrocarbon contaminated sites: a review. Surv Geophys 31:247–283.  https://doi.org/10.1007/s10712-009-9089-8 CrossRefGoogle Scholar
  6. Atekwana EA, Sauck WA, Werkema DD, Meju MA (2000) Investigations of geoelectrical signatures at a hydrocarbon contaminated site. J Appl Geophys 44:167–180CrossRefGoogle Scholar
  7. Atekwana EA, Atekwana EA, Rowe RS, Werkema DD Jr, Legall FD (2004) The relationship of total dissolved solids measurements to bulk electrical conductivity in an aquifer contaminated with hydrocarbon. J Appl Geophys 56:281–294.  https://doi.org/10.1016/j.jappgeo.2004.08.003 CrossRefGoogle Scholar
  8. Baker R, Moore J (1998) The application of time-lapse electrical tomography in groundwater studies. Lead Edge 17:1454–1458.  https://doi.org/10.1190/1.1437878 CrossRefGoogle Scholar
  9. Barker R, Rao TV, Thangarajan M (2001) Delineation of contaminant zone through electrical imaging technique. Currentence 81:277–283Google Scholar
  10. Bellmunt F, Marcuello A, Ledo J, Queralt P (2016) Capability of cross-hole electrical configurations for monitoring rapid plume migration experiments. J Appl Geophys 124:73–82.  https://doi.org/10.1016/j.jappgeo.2015.11.010 CrossRefGoogle Scholar
  11. Cardarelli E, Filippo GD (2009) Electrical resistivity and induced polarization tomography in identifying the plume of chlorinated hydrocarbons in sedimentary formation: a case study in rho (Milan - Italy). Waste Manag Res 27:595–602.  https://doi.org/10.1177/0734242x09102524 CrossRefGoogle Scholar
  12. Cassiani G, Bruno V, Villa A, Fusi N, Binley AM (2006) A saline trace test monitored via time-lapse surface electrical resistivity tomography. J Appl Geophys 59:244–259.  https://doi.org/10.1016/j.jappgeo.2005.10.007 CrossRefGoogle Scholar
  13. Cassiani G, Binley A, Kemna A, Wehrer M, Orozco AF, Deiana R, Boaga J, Rossi M, Dietrich P, Werban U, Zschornack L, Godio A, JafarGandomi A, Deidda GP (2014) Noninvasive characterization of the Trecate (Italy) crude-oil contaminated site: links between contamination and geophysical signals. Environ Sci Pollut Res 21:8914–8931.  https://doi.org/10.1007/s11356-014-2494-7 CrossRefGoogle Scholar
  14. Chen L, Liu Q, Li Y, Lu R, Wu S, Li X, Hou T (2016) Leakage diffusion of underwater crude oil in wind fields. Springerplus 5(1875):1875.  https://doi.org/10.1186/s40064-016-3457-x CrossRefGoogle Scholar
  15. Daily W, Owen E (1991) Cross-borehole resistivity tomography. Geophysics 56:1228–1235.  https://doi.org/10.1190/1.1443142 CrossRefGoogle Scholar
  16. Daily W, Ramirez A, LaBrecque D, Nitao J (1992) Electrical resistivity tomography of vadose water movement. Water Resour Res 28:1429–1442.  https://doi.org/10.1029/91WR03087 CrossRefGoogle Scholar
  17. Delgado-Rodríguez O, Shevnin V, Ochoa-Valdés J, Ryjov A (2006) Using electrical techniques for planning the remediation process in a hydrocarbon contaminated site. Rev Int Contamin Ambien 22:157–163Google Scholar
  18. Demirel C, Candansayar ME (2017) Two-dimensional joint inversions of cross-hole resistivity data and resolution analysis of combined arrays. Geophys Prospect 65:876–890.  https://doi.org/10.1111/1365-2478.12432 CrossRefGoogle Scholar
  19. Deng Y, Shi X, Xu H, Sun Y, Wu J, Revil A (2016) Quantitative assessment of electrical resistivity tomography for monitoring DNAPLs migration—comparison with high-resolution light transmission visualization in laboratory sandbox. J Hydrol 544:254–266.  https://doi.org/10.1016/j.jhydrol.2016.11.036 CrossRefGoogle Scholar
  20. Donaldson EC, Siddiqui TK (1989) Relationship between the Archie saturation exponent and wettability. SPE Form Eval 4:359–362.  https://doi.org/10.2118/16790-pa CrossRefGoogle Scholar
  21. Franco RD et al (2009) Monitoring the saltwater intrusion by time lapse electrical resistivity tomography: the Chioggia test site (Venice Lagoon, Italy). J Appl Geophys 69:117–130.  https://doi.org/10.1016/j.jappgeo.2009.08.004
  22. Gelius LJ, Zhong W (2010) Modelling production caused changes in conductivity for a siliciclastic reservoir: a differential effective medium approach. Geophys Prospect 56:677–691.  https://doi.org/10.1111/j.1365-2478.2008.00720.x CrossRefGoogle Scholar
  23. Godio A, Arato A, Stocco S (2010) Geophysical characterization of a non aqueous-phase liquid-contaminated site. Environ Geosci 17:141–162.  https://doi.org/10.1306/eg.04261010003 CrossRefGoogle Scholar
  24. Grumman DL Jr, Daniels JJ (1995) Experiments on the detection of organic contaminants in the vadose zone. J Environ Eng Geophys 1:31–38.  https://doi.org/10.4133/JEEG1.A.31 CrossRefGoogle Scholar
  25. Halihan T, Sefa V, Sale T, Lyverse M (2017) Mechanism for detecting NAPL using electrical resistivity imaging. J Contam Hydrol 205:57–69.  https://doi.org/10.1016/j.jconhyd.2017.08.007 CrossRefGoogle Scholar
  26. Hamzah U, Ismail MA, Samsudin AR (2009) Geoelectrical resistivity and ground penetrating radar techniques in the study of hydrocarbon-contaminated soil. Sains Malaysiana 38:305–311Google Scholar
  27. Hu LM, Meegoda JN, Lo IMC, Liu Y, Gao SY, Wu ZQ (2010) Gasoline contaminated sites: pollutant transport and remediation. Adv Environ Geotech 221–235.  https://doi.org/10.1007/978-3-642-04460-1_15
  28. Illman WA, Berg SJ, Liu X, Massi A (2010) Hydraulic/partitioning tracer tomography for DNAPL source zone characterization: small-scale sandbox. Experiments. Environ Sci Technol 44:8609–8614.  https://doi.org/10.1021/es101654j CrossRefGoogle Scholar
  29. Karaoulis M, Tsourlos P, Kim JH, Revil A (2014) 4D time-lapse ERT inversion: introducing combined time and space constraints. Near Surf Geophys 11:25–34.  https://doi.org/10.1021/es101654j CrossRefGoogle Scholar
  30. Keller GV, Frischknecht FC (1966) Electrical methods in geophysical prospecting. Pergamon PressGoogle Scholar
  31. Liang C, Zheng X-L, Zhang J-J, Yue F (2011) Testing of water content in oil-contaminated porous medium. J Univ (Earth Sci Edn) 41:826–830.  https://doi.org/10.3969/j.issn.1671-5888.2011.03.029 Google Scholar
  32. Liang L, Abubakar A, Habashy TM (2016) Reservoir property mapping and monitoring from joint inversion of time-lapse seismic, electromagnetic, and production data. Geophysics 81:WC33–WC44.  https://doi.org/10.1190/geo2015-0620.1 Google Scholar
  33. Loke MH (1999) Constrained Time-Lapse Resistivity Imaging Inversion Sageep:EEM7.  https://doi.org/10.4133/1.2922877
  34. Loke MH (2004) Tutorial: 2-D and 3-D electrical imaging surveysGoogle Scholar
  35. Loke MH, Barker RD (1996) Rapid least-squares inversion of apparent resistivity pseudosections by a quasi-Newton method1. Geophys Prospect 44:131–152.  https://doi.org/10.1111/j.1365-2478.1996.tb00142.x CrossRefGoogle Scholar
  36. Lowe DF, Oubre CL, Ward CH (1999) Surfactants and Cosolvents for NAPL remediation: a technology practices manual. Crc PressGoogle Scholar
  37. Mungan N, Moore EJ (1968) Certain wettability effects on electrical resistivity in porous media. J Can Pet Technol 7:20–25.  https://doi.org/10.2118/68-01-04 CrossRefGoogle Scholar
  38. Newell CJ, Acree SD, Ross RR, Huling SG (2011) Ground water issue: light nonaqueous phase liquids, vol 44, pp 1294–1295Google Scholar
  39. Oldenborger GA, Knoll MD, Routh PS, Labrecque DJ (2007) Time-lapse ERT monitoring of an injection/withdrawal experiment in a shallow unconfined aquifer. Geophysics 72:F177–F187.  https://doi.org/10.1190/1.2734365 CrossRefGoogle Scholar
  40. Orlando L, Renzi B (2015) Electrical permittivity and resistivity time lapses of multiphase DNAPLs in a lab test. Water Resour Res 51:377–389.  https://doi.org/10.1002/2014wr015291 CrossRefGoogle Scholar
  41. Pan Y, Jia Y, Wang Y, Xia X, Guo L (2018) Study on diesel vertical migration characteristics and mechanism in water-bearing sand stratum using an automated resistivity monitoring system. Environ Sci Pollut Res 25:1–11.  https://doi.org/10.1007/s11356-017-0698-3 Google Scholar
  42. Perri MT, Cassiani G, Gervasio I, Deiana R, Binley A (2012) A saline tracer test monitored via both surface and cross-borehole electrical resistivity tomography: comparison of time-lapse results. J Appl Geophys 79:6–16.  https://doi.org/10.1016/j.jappgeo.2011.12.011 CrossRefGoogle Scholar
  43. Peter A, Miles B, Teutsch G (2008) Estimation of emission from an LNAPL contaminated zone considering groundwater recharge. Environ Geol 55:321–337.  https://doi.org/10.1007/s00254-007-0978-1 CrossRefGoogle Scholar
  44. Power C, Gerhard JI, Tsourlos P, Soupios P, Simyrdanis K, Karaoulis M (2015) Improved time-lapse electrical resistivity tomography monitoring of dense non-aqueous phase liquids with surface-to-horizontal borehole arrays. J Appl Geophys 112:1–13.  https://doi.org/10.1016/j.jappgeo.2014.10.022 CrossRefGoogle Scholar
  45. Quyum A, Achari G, Goodman RH (2002) Effect of wetting and drying and dilution on moisture migration through oil contaminated hydrophobic soils. Sci Total Environ 296:77–87.  https://doi.org/10.1016/s0048-9697(02)00046-3 CrossRefGoogle Scholar
  46. Rao PSC, Annable MD, Kim H (2000) NAPL source zone characterization and remediation technology performance assessment: recent developments and applications of tracer techniques. J Contam Hydrol 45:63–78.  https://doi.org/10.1016/s0169-7722(00)00119-4 CrossRefGoogle Scholar
  47. Revil A, Schmutz M, Batzle ML (2011) Influence of oil wettability upon spectral induced polarization of oil-bearing sands. Geophysics 76:A31–A36.  https://doi.org/10.1190/geo2011-0006.1 CrossRefGoogle Scholar
  48. Rosales RM, Martínez-Pagán P, Faz A, Bech J (2014) Study of subsoil in former petrol stations in SE of Spain: physicochemical characterization and hydrocarbon contamination assessment. J Geochem Explor 147:306–320.  https://doi.org/10.1016/j.gexplo.2014.10.006 CrossRefGoogle Scholar
  49. Sadowski R (1988) Clay-organic interactions MSc Thesis, Dept of Geochemistry, Colo School of Mines, Golden, CO 209 pGoogle Scholar
  50. Sasaki Y (1992) Resolution of resistivity tomography inferred from numerical simulation. Geophys Prospect 40:453–463.  https://doi.org/10.1111/j.1365-2478.1992.tb00536.x CrossRefGoogle Scholar
  51. Sauck WA (2000) A model for the resistivity structure of LNAPL plumes and their environs in sandy sediments. J Appl Geophys 44:151–165.  https://doi.org/10.1016/S0926-9851(99)00021-X CrossRefGoogle Scholar
  52. Schroth MH, Istok JD, Selker JS (1998) Three-phase immiscible fluid movement in the vicinity of textural interfaces. J Contam Hydrol 32:1–23.  https://doi.org/10.1016/s0169-7722(97)00069-7 CrossRefGoogle Scholar
  53. Schubert M, Paschke A, Lau S, Geyer W, Knöller K (2007) Radon as a naturally occurring tracer for the assessment of residual NAPL contamination of aquifers. Environmental Pollution 145:920–927.  https://doi.org/10.1016/j.envpol.2006.04.029
  54. Shevnin V, Delgado O, Mousatov A, Nakamura E, Mejía A (2003) Oil pollution detection using resistivity sounding. Geofis Int 42:613–622Google Scholar
  55. Shima H (1992) 2D and 3D resistivity imaging reconstruction using cross-hole data. Geophysics 55:682–694CrossRefGoogle Scholar
  56. Slater L, Binley AM, Daily W, Johnson R (2000) Cross-hole electrical imaging of a controlled saline tracer injection. J Appl Geophys 44:85–102.  https://doi.org/10.1016/S0926-9851(00)00002-1
  57. Slater LD, Ntarlagiannis D, Day-Lewis FD, Mwakanyamale K, Versteeg RJ, Ward A, Strickland C, Johnson CD, Lane JW Jr (2010) Use of electrical imaging and distributed temperature sensing methods to characterize surface water&ndash;groundwater exchange regulating uranium transport at the Hanford 300 Area, Washington. Water Resour Res 46:79–93.  https://doi.org/10.1029/2010wr009110 CrossRefGoogle Scholar
  58. Suman RJ (1997) Effects of pore structure and wettability on the electrical resistivity of partially saturated rocks—a network study. Geophysics 62:1151–1162.  https://doi.org/10.1190/1.1444216 CrossRefGoogle Scholar
  59. Thabit JM, Khalid FH (2016) Resistivity imaging survey to delineate subsurface seepage of hydrocarbon contaminated water at Karbala Governorate, Iraq. Environ Earth Sci 75(87).  https://doi.org/10.1007/s12665-015-4880-y
  60. Tomlinson DW, Rivett MO, Wealthall GP, Reh S (2017) Understanding complex LNAPL sites: illustrated handbook of LNAPL transport and fate in the subsurface. J Environ Manag 204:748–756.  https://doi.org/10.1016/j.jenvman.2017.08.015 CrossRefGoogle Scholar
  61. Toumelin E, Torres-Verdín C (2005) Influence of oil saturation and wettability on rock resistivity measurements: a uniform pore-scale approach. SPWLA 46th Annual Logging Symposium. Society of Petrophysicists and Well-Log AnalystsGoogle Scholar
  62. Tsourlos P, Ogilvy R, Papazachos C, Meldrum P (2011) Measurement and inversion schemes for single borehole-to-surface electrical resistivity tomography surveys. J Geophys Eng 8:487–497(411).  https://doi.org/10.1088/1742-2132/8/4/001 CrossRefGoogle Scholar
  63. Waxman MH, Thomas EC (1974) Electrical conductivities in Shaly Sands-I. The relation between hydrocarbon saturation and resistivity index; II. The temperature coefficient of electrical conductivity. J Pet Technol 26:213–227.  https://doi.org/10.2118/4094-ms CrossRefGoogle Scholar
  64. Weber WJ, Huang W (1996) A distributed reactivity model for sorption by soils and sediments. 4. Intraparticle heterogeneity and phase distribution relationships under nonequilibrium conditions. Environ Sci Technol 30:881–888.  https://doi.org/10.1021/es950329y CrossRefGoogle Scholar
  65. Wilkinson PB, Chambers JE, Lelliott M, Wealthall GP, Ogilvy RD (2010) Extreme sensitivity of crosshole electrical resistivity tomography measurements to geometric errors. Geophys J Int 173:49–62.  https://doi.org/10.1111/j.1365-246x.2008.03725.x CrossRefGoogle Scholar
  66. Wu Y, Shuler P, Blanco M, Tang Y, Goddard W (2008) A study of wetting behavior and surfactant EOR in carbonates with model compounds. SPE J 13:26–34.  https://doi.org/10.2118/99612-ms CrossRefGoogle Scholar
  67. Yeh TCJ, Zhu J (2007) Hydraulic/partitioning tracer tomography for characterization of dense nonaqueous phase liquid source zones. Water Resour Res 43:113–119.  https://doi.org/10.1029/2006wr004877 Google Scholar
  68. Zhao J, Rijo L, Ward S (1986) Effects of geologic noise on cross-borehole electrical surveys. Geophysics 51:1978–1991.  https://doi.org/10.1190/1.1442053 CrossRefGoogle Scholar
  69. Zhou Y, Cardiff M (2017) Oscillatory hydraulic testing as a strategy for NAPL source zone monitoring: laboratory experiments. J Contam Hydrol 200:24–34.  https://doi.org/10.1016/j.jconhyd.2017.03.005 CrossRefGoogle Scholar
  70. Zhou B, Greenhalgh SA (2000) Cross-hole resistivity tomography using different electrode configurations. Geophys Prospect 48:887–912.  https://doi.org/10.1046/j.1365-2478.2000.00220.x CrossRefGoogle Scholar

Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature 2019

Authors and Affiliations

  1. 1.College of Environmental Science and EngineeringOcean University of ChinaQingdaoChina
  2. 2.Shandong Provincial Key Laboratory of Marine Environment and Geological EngineeringQingdaoChina

Personalised recommendations