Assessment of Groundwater Contamination Risk in Oilfield Drilling Sites Based on Groundwater Vulnerability, Pollution Source Hazard, and Groundwater Value Function in Yitong County

Abstract

Oilfield drilling sites are the potential dispersive pollution source of groundwater, especially to shallow groundwater. The pollution risk assessment in these areas is an important reference for effective groundwater management and protection. The vulnerability assessment alone is not sufficient for groundwater contamination risk assessment. In this study, we developed a comprehensive groundwater pollution risk assessment method for oilfield drilling sites that combine groundwater vulnerability, pollution source hazard, and groundwater value function to produce a more comprehensive result. Consider the oilfield drilling area in Yitong County of Jilin Province, China, as an example. Thematic maps of the three aspects (groundwater vulnerability, pollution source hazard, and groundwater value function) were generated in ArcGIS environment to assess the contamination risk of groundwater in quaternary pore unconfined aquifer. The results show that 9.92% of the study area is characterized as being at high risk. These areas are mainly distributed around the center position of the oil drilling site, floodplains, and the reservoir. The moderate risk area accounts for 21.04% of the total area. It is distributed in the first-level terrace, mainly because of the high function value of groundwater. The remaining 69.04% of the study area is characterized as none and mild risk, mainly distributed in the valleys and terraces. This integrated groundwater contamination risk assessment method is suited for comparative assessment of multiple-point sources of contamination at a regional scale. Finally, the groundwater contamination risk grade distributed in this area provides a reference for effective protection and sustainable supply of groundwater in the oilfield drilling area.

1. Introduction

Petroleum is one of the most important resources in the world. Oilfield drilling and exploitation distribute widely in the area where oil has been extracted. The oilfield drilling sites are generally considered as the multiple-point sources of contamination to groundwater. Groundwater pollution incidents occur frequently in oil exploitation, transportation, and processing [1]. Oil leakage in the process of oil recovery would produce great harm to groundwater quality, especially carcinogenic, teratogenic, and mutagenic petroleum pollutants [2]. They are greatly harmful to the ecological environment and human health [3]. It is essential to evaluate oil drilling hazards to the environment, especially to groundwater. Therefore, the groundwater pollution risk assessment is strongly necessary in this area.

Groundwater risk assessment is an essential tool for prevention and control of oil leakage [4,5]. It is highly associated with ecological, agricultural, industrial, and human activities. According to former studies, the influence factors of groundwater contamination risk can be categorized into three aspects: the vulnerability, the hazard caused by pollutants, and the degree of groundwater development and utilization [6,7,8]. Groundwater vulnerability research is a key step in assessing pollution risk [9,10]. To deeply understand groundwater vulnerability, various methods are proposed, such as the process-based simulation method [11], statistical method, and index overlay methods (DRASTIC, SINTACS, GOD, SI, etc.) [12,13,14,15,16,17]. At present, the DRASTIC model is the most widely applied one in various countries [18,19,20,21,22,23,24]. The DRASTIC model was first developed by the United States Environmental Protection Agency (USEPA). It then became increasingly mature with the intrinsic vulnerability indices of groundwater and dynamic factors such as anthropogenic activities [22,25,26,27]. Due to distinctive hydrogeological characteristics and diverse human activities, different evaluation methods for intrinsic aquifer vulnerability have been developed and applied [28]. In addition, there are several methods (LM-BP, AHP, PCA, etc.) applied to improve the value of indicators, which were usually subjective [29,30,31].

Groundwater vulnerability assessment, integrated with groundwater pollution source hazard assessment, was gradually applied to appraise the groundwater pollution risk since the 1980s [32,33]. Groundwater pollution source hazard assessments were usually evaluated on water quality by samples, but rarely be mentioned for future pollution forecasts [34]. Therefore, it has been gradually agreed upon that the risk should be expressed by integrating the probability of contamination occurrence with the expected damage of groundwater pollution [35,36,37]. Moreover, the groundwater function value is an important reference for groundwater protection. The value function assessment of groundwater was included in the pollution risk of a groundwater system which focuses on the pollution risk receptors [38]. The purpose of groundwater assessment is to make people aware of the present and future value of groundwater exploitation [39]. Various factors have been illustrated to quantify the value of groundwater as a means to measure the total value of groundwater function through its exploitation and in situ value by many researchers [40]. The groundwater function value assessment system was established by selecting the factors about groundwater ecological service, health service, and socio-economic service functions [41,42]. Generally, it was assessed by groundwater quantity and quality, and water supply significance [43]. In the karst area, the evaluated indices included population density, per capita GDP, groundwater level, average annual precipitation, groundwater quality, and modulus of groundwater resources [44]. All these factors chosen in existing research are mainly related to human activities, and the quality and quantity of groundwater. Groundwater function values can be classified into not only an ex situ (exploration) value and an in situ value but also groundwater quantity and quality [43]. The groundwater function values are difficult to calculate because of their highly complex relationships, such as water quality, water richness, and the capture zone of pumping wells, etc. [35,45]. These indices need to be selected and integrated based on their relevance to the study area.

In this study, we assess groundwater contamination risk in the oilfield drilling site in Yitong County of Jilin Province, China. In particular, we combine groundwater intrinsic vulnerability with pollution source hazard and groundwater function value to produce a more objective result for groundwater contamination risk. This integrated groundwater contamination risk assessment method is suited for comparative assessment of a large number of dispersive pollution sources at a regional scale, especially in the countries with the most impacted groundwater resources such as Iran, Pakistan, India, USA, and North China, etc. The results above provide a good approach for accurately assessing groundwater risk. The mapping can be applied for the effective protection and sustainable supply of groundwater.

2. Study Area

The study area is located to the west of the Yitong River in Yitong Manchu Autonomous County, Jilin Province of China, with a total area of 197.5 km² (Figure 1). The geomorphological types are mainly alluvial plains and erosion platforms. The terrain is roughly inclined from south to north. This study area is located in the Mori-qing fault depression [46]. The climatic condition of this area is the continental monsoon that is influenced by the east monsoon during the summer. The average annual precipitation is 651.7 mm. The amount of potential evaporation is about 800 mm. The sediments in the unsaturated zone and the lithology of aquifers are mainly sand, gravel, conglomerate, sandstone, and others. The western boundary of the study area is the northwest Songliao watershed. The quaternary unconfined aquifer thickens on both sides of the watershed and thins in the plain area. The depth of groundwater is between 0 m and 9.6 m. Moreover, the Quaternary unconfined aquifer and Neogene confined aquifer are in unconformity contact. The main recharge of groundwater is infiltration of precipitation. There is then the lateral runoff, river, and irrigation infiltration in local areas. The main discharge of groundwater is seepage into the rivers, and the vertical evaporation is the main discharge in the valley area. The hydrogeological conditions are different with obvious characteristics from the erosion platform areas in the northwest and southeast to the central plains. For example, groundwater is mainly stored in carbonate rock fissures and fault veins in the northwest and southeast. It is stored mainly in the porous medium in the central alluvial plains. Most of the area is cultivated land with corn crops. The irrigation water is mainly from groundwater except the land near the surface water [47].

The drilling wells of an oilfield mining plant in the study area are located in the quasi-protected area of the drinking water source in the Yitong area. The site selection scope includes the distribution of centralized water supply source wells in Yitong County. At present, 20 underground water source mining wells supply water to the urban settlements in Yitong County for domestic and production. An oil field company intends to drill wells in Yitong County, and 51 oil drilling wells are preliminarily designed in the quasi-protected zone (Figure 1) [47]. Therefore, it is urgent to assess groundwater pollution risk to find out whether and how the oil drilling activities influence the groundwater resource. Effective decisions must be made for the protection and management of groundwater in this area.

3. Materials and Methods

3.1. Groundwater Vulnerability Assessment

Groundwater intrinsic vulnerability assessment indicators include Depth to water (D), net Recharge (R), Aquifer medium (A), Soil medium (S), Terrain Slope (T), Impact of the vadose zone (I), and hydraulic Conductivity (C) [24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,48]. Considering the rapid development of a social economy in this area, the groundwater vulnerability is highly affected by human activities. The types of land use in this area are industrial land, dry land, paddy field, residential land, and woodland. The land is irrigated by groundwater in a large area, and it is closely relevant to the quantity and quality of groundwater. Moreover, there are several rivers across the study area. According to the field survey, the cultivated land near the rivers is irrigated by the surface water. The distance to the river should not be neglected. Therefore, land-use type (L) and distance to rivers (D_R) are selected as the supplemental indices of the intrinsic vulnerability assessment system. The unconfined aquifer is much easier to be polluted by the activities of oil drilling. The Quaternary unconfined aquifer is considered to be the target aquifer in this study. According to the conditions mentioned above, the DASTIC-LD_R evaluation system was developed to assess the groundwater vulnerability in this area. The index rating applied for parameter quantification is listed in

Table 1. Groundwater vulnerability assessment index rating.

Rating	Depth to Water (m)	Aquifer Media	Soil Medium	Terrain Slope (%)	Impact of the Vadose Zone	Hydraulic Conductivity (m/d)	Land Use Types	Distance to Rivers (m)
10	0–1.0	Fan-shaped sandy gravel	Rubbish	0–0.01	Gravel	≥100	Industrial land	-
9	1.0–3.0	Ice water sedimentation with kaolin soil gravel	Sand gravel	0.01–0.05	Sand gravel	-	-	0–100
8	-	River valley plain gravel	-	0.05–0.10	Coarse sand	80–100	Dry land	-
7	3.0–5.0	Ice water stacked gravel	Loess, Sand loam	-	Fine sand	-	-	100–300
6	-	Wavy platform gravel	Fine sand	0.10~0.30	Silty sand	50–80	Surface waters	-
5	5.0–10	Waxy platform sand, fine sand	Silty sand	-	Sub-sand	-	-	300–500
4	-	Sandstone	Sub-clay	-	Muddy sub-sand	30–50	Paddy field	-
3	10.0–15	Medium coarse sandstone	-	0.30–0.50	Sub-clay	20–30	Residential land	500–800
2	15–20	Gravel fine sandstone	Muddy clay	0.50–0.80	-	4–20	Woodland	-
1	≥20	Bedrock	Silt	≥0.80	Bedrock	<4	-	≥800

.

Based on the analysis above, the groundwater vulnerability can be calculated by the Equation (1):

D_I = Dw × Dr + Rw × Rr + Aw × Ar + Sw × Sr + Tw × Tr + Iw × Ir + Cw × Cr + Lw × Lr + D_Rw × D_Rr

(1)

where D_I is the groundwater vulnerability index; r is the rating value of factors; w is the weight.

Groundwater vulnerability was divided into five grades: low, relatively low, medium, relatively high, and high. The range of each grade for groundwater vulnerability is shown in

Table 2. The range of each grade for groundwater vulnerability.

Grade	Low	Relatively Low	Medium	Relatively High	High
Range	2.5–3.0	3.0–3.5	3.5–4.0	4.0–4.5	4.5–5.5

. The higher the DASTIC-LD_R indicator is, the greater the groundwater vulnerability is.

The Principal Component Analysis (PCA) method was applied to determine the weight of each evaluation index [31]. As a multivariate statistical method, PCA is applied to produce a set of values of linear uncorrelated variables. The principal components reserve as much as possible of the important alteration variation present in the data set [49].

$$\mathrm{Y}=\mathrm{AX}$$

(2)

where Y is the principal components; X is the evaluating indicator; and A is the transformation matrix.

The weight coefficient and the normalized weight were calculated by SPSS 16.0. The results for the indices are shown in

Table 3. Weight of each indicator in groundwater vulnerability assessment.

Assessment Indicator	Weight Coefficient	Weight Value
D	0.443	0.17
A	0.206	0.079
S	0.229	0.088
T	0.242	0.093
I	0.451	0.173
C	0.412	0.158
L	0.242	0.093
DR	0.381	0.146

.

3.2. Groundwater Pollution Source Hazard Evaluation

The result of groundwater pollution source hazard reveals the degree of damage caused by pollution sources to groundwater [50]. During the oilfield drilling process, oily sewage such as ground crude oil, drilling wastewater, and waste mud penetrates into the groundwater through the aeration zone, which are potential pollutants. Benzene, toluene, ethylbenzene, and xylene are the most common components of groundwater petroleum pollutants [51,52]. Therefore, the benzene series was selected as the simulation indicator, and its transport range in groundwater was forecasted in certain scenarios. According to the government standard for groundwater quality in China (GB/T 14848-2017), the concentration of benzene series in groundwater should not exceed 0.05 mg/L in the producing state of the oilfield. Groundwater reaches the III class of drinking water standard, which can be regarded as no risk. When the concentration of benzene series exceeds 0.3 mg/L, it would seriously contaminate groundwater and would severely exceed the standard range [53]. The concentration of petroleum pollutants was set to four pollution level ranges: none (0, 0.05), mild (0.05, 0.15), medium (0.15, 0.30), and severe (0.30,$\infty$). The grade of groundwater pollution source hazard was divided into four classes accordingly: no, mild, medium, and high. In order to analyze the transport characteristics of pollutants in the porous medium, groundwater numerical simulation was built by applying the finite difference method (FDM). The software Groundwater Modelling System (GMS, Version 10.4) was chosen to find out the contamination transfer law, such as the range and the concentration of pollutants in a certain time. GMS has become one of the most widely used groundwater simulation software in the world with a good user interface, powerful pre-processing and post-processing functions, and excellent 3D visual effects [54,55]. The modules MODFLOW and MT3DMS are applied separately to simulate groundwater flow and the influence range of benzene series when pollutants are released [56]. The effects of drilling on the unconfined water near each water source well are forecasted by simulation.

3.3. Groundwater Function Assessment

The selection of assessment indices for groundwater function is according to the main influence facts in the ecological and environment of groundwater as well as the economic and social development of water [57]. The oilfield drilling and centralized water supply wells for human habitation and groundwater function value were assessed by groundwater quality and groundwater storage. The calculation formula is as follow:

$$\mathrm{V}={\mathrm{G}}_{\mathrm{Q}}{\mathrm{G}}_{\mathrm{S}}$$

(3)

where V is groundwater function value; G_Q is groundwater quality; and G_S is groundwater storage. The grade of groundwater function value was divided into four classes: no, mild, medium, and high. The high value means the area contains abundant storage and high groundwater quality, so it has greater sensitivity and susceptibility of groundwater to pollution [57,58,59].

3.4. Comprehensive Assessment of Groundwater Pollution Risk

The groundwater contamination risk is not only related to the groundwater vulnerability, but also closely connected to the groundwater pollution source hazard and groundwater function value. The three aspects mentioned above were appraised separately by applying the geographic information system (GIS). The groundwater vulnerability assessment map, groundwater pollution source risk assessment map, and groundwater value function assessment map were counted by software ArcGIS (Version. 10.2). The attributes of the three layers were extracted for statistics. The weights of groundwater contamination risk were calculated by the principal component factor analysis (PCA) function using the software SPSS (Version. 16.0). Finally, the groundwater contamination risk level map was drawn by software ArcGIS. The flow chart of the study is shown in Figure 2.

4. Results and Discussion

4.1. Groundwater Vulnerability Assessment

The groundwater vulnerability assessment indices were calculated with Equation (2) by applying SPSS (Version 16.0), and the assessment results were calculated by Equation (1). The raster maps of each indicator for the DASTIC-LD_R model were mapped separately by the software ArcGIS (Version 10.2) (Figure 3a–h). The value of DI was between 2.5 and 5.0. It was divided into five grades according to the equal interval method. The weight of each evaluation indicator was between 0.079 and 0.173, which showed a clear difference to vulnerability. The aeration zone exhibited the highest impaction weight, followed by depth to water, hydraulic conductivity, and distance to rivers. High groundwater vulnerability was observed in areas where the groundwater was shallow [60]. In comparison, aquifer medium, soil medium, terrain slope, and land use forms exerted less of an impact on the vulnerability distribution.

According to the results of the groundwater vulnerability assessment shown in Figure 3i, the area with relatively low vulnerability accounted for 25.57% of the total area. It is mainly distributed in the U-shaped valleys and first terraces, where the lithology is loess and sandy loam. The medium and relatively high vulnerability areas accounted for 69.85% and are concentrated in the floodplain and first terraces. Mainly, the lithology is sand gravel alluvium and sub-sand, the permeability coefficient is high, and the aquifer is shallow. In addition, human activity is an important factor in the expansion and intensification of groundwater vulnerability [57]. Human activities are relatively intense with industrial activities and extensive pollution emissions. The comprehensive results between groundwater quality analysis and groundwater vulnerability assessment show that the valley area and first terrace are the place with poor groundwater quality and high groundwater vulnerability. The U-shaped valleys show good water quality and where the low groundwater vulnerability distributes [60].

4.2. Groundwater Pollution Source Hazard Evaluation

The hydrogeological conceptual model was established based on hydrogeological conditions (Figure 4a). The seepage movement is basically in accordance with the Darcy’s law. The groundwater flow is limited by the accuracy of the data and is treated as a planar two-dimensional flow. The elements change over time and are generalized as unsteady.

The western boundary of the study area is the Songliao watershed, which can be generalized to the impervious boundary. The eastern boundary is the Yitong River, which can be generalized to river boundaries as the recharge of the unconfined aquifer. The northern and southern borders are generalized to the Neumann boundary. According to the hydrogeological conceptual model, the groundwater flow in the study area is generalized as a pore unconfined water in loose rocks. The hydrogeological parameters, such as infiltration coefficient of precipitation, phreatic evaporation coefficient, irrigation leakage coefficient, and specific yield were calculated. They were input to the model as the initial parameters. The groundwater recharge such as precipitation infiltration, lateral inflow, river seepage, and irrigation infiltration were all calculated and input to the model, so was the groundwater discharge such as evaporation, lateral outflow, and artificial mining.

After the model had been calibrated, the observed water level and the simulated one fit well. The final value of parameters was confirmed, which is shown in

Table 4. Calculated value of sources-sinks and the final hydrogeological parameters.

Sources-Sinks	Calculated Value	Total	Parameter	Calculated Value
Area (km2)	-	197.51	Area (km2)	197.51
Precipitation infiltration	11.38~373.07	1014.61	Hydraulic conductivity (m/d)	4~102
Lateral inflow	0~187.08	196.16	Specific yield	0.03~0.25
River infiltration	0~61.40	157.42	Irrigation leakage coefficient	0.05~0.15
Irrigation flow	0.80~12.88	71.1	Evaporation coefficient	0~0.1
Evaporation	0~−228.98	−340.81	Longitudinal diffusion coefficient (m2/h)	3.98
Lateral outflow	0~−63.40	−63.4	Lateral diffusion coefficient (m2/h)	0.313
Artificial mining	−9.13~−302.61	−971.87

. The scatter plot in Figure 4c shows good compatibility between observed and calculated groundwater levels by the model.

The mobility of pollutants was assessed based on the migration characteristics in groundwater [56]. The impact of petroleum pollutants on groundwater is predicted by applying the groundwater solute transport model of 20 years to the leakage of polluted water from drilling wells in the process of oil mining.

According to relevant statistics, 359 oil wells already existed in the Yitong area. Assorted indicators of oil pollution or the oil content, will reach 2–10% after 20 years [47]. Therefore, the concentration of the benzene series in the model is set to be 85.0 mg/l. The maximum amount of return water outside the oil pipe is 20 m³/d for 15 d.

The forecasted results are shown in Figure 5. The total range of the benzene series on the unconfined aquifer in 20 years is 3.572 km², and the total area that exceeds the standard range is 0.350 km². The maximum migration distance of the benzene series for single oil drilling is 0.824 km on the phreatic water. The influence range of the benzene series in oil drilling wells (Y2, Y4, and Y6) is relatively close to the water source wells (SY4-SY12). The pollution risk to the water source well (SY4) is high, with the nearest distance of 75 m. The map suggests that the areas where they have oil drilling wells were the most harmful potential contamination sources [39].

4.3. Groundwater Function Value Assessment

With the sensitivity analysis of groundwater function value assessment in the study area (Figure 6), which was calculated by Equation (3), the relatively low function value areas accounted for 40.71% of the total area. It is mainly distributed in the northeast and southwest, with sparsely populated areas, dispersed water supply, and relatively good groundwater quality. The medium function value accounted for 18.88% and is mainly distributed in the north and south, with small population density, scattered water supply, and a poor groundwater utilization rate. The rating of relatively high and high includes 40.41%. They are mainly distributed in the central and western areas, due to densely populated villages and large population density. There are groundwater sources for concentric supply, and the groundwater quality is medium. As such, there is medium or low groundwater quality in more populated areas where the groundwater function is high.

4.4. Comprehensive Assessment of Groundwater Pollution Risk

The groundwater pollution risk assessment rating mapping can be obtained using the Raster Calculator of the ArcGIS by further analysis of the range values of the risk assessment. The range [1.8, 3.8] was divided into four classes according to the equal interval method: none, mild, medium, and high (Figure 7). First, the results of groundwater pollution risk assessment show that the groundwater in the region with no risk and mild risk include the area of about 136.35 km², which is 69.04% of the total area. Because there is no water source well in these areas, the risk of groundwater pollution in U-shaped valleys and first terraces is low. The groundwater vulnerability in these areas is also low or medium. Furthermore, the medium pollution risk area is about 41.55 km², which includes 21.04% since it is mainly located around the first terrace and the oil drilling site. The vulnerability of groundwater and the hazard from groundwater pollution sources is high. Lastly the high-risk area is about 19.60 km², which includes 9.92%. It is mainly distributed in the floodplain and oil drilling site, with high groundwater vulnerability and a high hazard from groundwater pollution sources.

5. Conclusions

In this study, we developed a comprehensive groundwater pollution risk assessment method for areas with dispersive pollutants such as the oilfield drilling sites. The results of groundwater pollution risk assessment show that groundwater at no risk and mild risk accounted for 69.04% of the total area and are mainly distributed in U-shaped valleys and first terraces. Medium risk accounted for 21.04% and is mainly distributed around the first terrace and the oil drilling sites. High risk accounted for 9.92% and is mainly distributed in the floodplain and oil drilling sites. In high risk areas, the water manager should take relevant prevention and control measures, carry out groundwater protection and purification work according to local conditions, and strengthen groundwater quality monitoring. When the water source is not polluted by petroleum pollutants in oil drilling sites, long-term monitoring is still needed and relevant treatment measures should be taken.

The study proposes a new method to assess groundwater pollution risk which combines groundwater vulnerability, pollution source hazard, and groundwater function value. This integrated groundwater contamination risk assessment method is suited for comparative assessment of a large number of dispersive pollution sources at a regional scale. The results above provide a good approach for assessing groundwater risk accurately. The mapping can be applied for the effective protection and sustainable supply of groundwater.