Study on the Hydrogeological Structure of a Karst Subterranean River and Seepage of a Karst Reservoir: A Case Study of the Yibasan Reservoir in Yunnan Province, China

Abstract

Karst groundwater resources are rich, and they have important water supply significance. A karst reservoir is a means of exploiting and utilizing groundwater resources, but because of the complex hydrogeological structure and underground river pipeline in a karst area, the seepage problem of the reservoir is extremely serious. Therefore, the Yibasan Karst Reservoir (YKR) was selected as the focus of this study. The hydrogeological structure of the subterranean river system of Yutang (SRSY) was identified and the hydraulic connections between the subterranean river conduits were determined using tracer experiment and groundwater dynamic monitoring. Furthermore, the development location and depth of the karst seepage zone of the YKR were determined using geophysical exploration. The results showed that there were three subterranean river conduits in the SRSY, and there was no hydraulic connection. The northern and southern pressure and torsion faults on the sides constitute the impervious boundary of the SRSY, which provided good catchment conditions for the formation of the YKR. Additionally, the northern and main conduits of the SRSY provide a sufficient groundwater source for the YKR. Moreover, the development width of the karst seepage zone of the YKR ranges between 40 and 60 m, and the elevation ranges between 1275 and 1355 m. The research results not only provide an effective basis for the treatment of the karst seepage problem of the YKR but also provide an important reference for the development and utilization of groundwater resources in similar karst areas.

1. Introduction

A large-scale underground karstic network formed by the karst process provides excellent space for the storage and runoff of karst water, which makes the karst groundwater resources in this area extremely rich [1,2,3,4]. Therefore, karst groundwater is a significant and important water supply [5,6,7,8,9]. A karst reservoir serves as a means of exploiting and utilizing karst groundwater resources, thereby addressing the water needs of the local population and facilitating local economic growth [5,10,11]. However, the seepage phenomenon of karst reservoirs is extremely frequent due to the development of karst pipes and cracks in karst areas [12,13]. For instance, the Xiaohewei Karst reservoir, located in Wenshan City, China, has had serious seepage problems since its establishment in 2007 [14]. In addition, the Luotan Karst reservoir, located in Guilin City, China, has been unable to store water normally since 1975, and there have been multiple karst seepage pathways [15]. The seepage of karst reservoirs is a common disaster that results in the depletion of karst groundwater resources, a decrease in the effectiveness of water resource development and utilization, and significant security threats to water conservancy infrastructure [14,15]. Hence, the management of karst reservoir leakage holds significant importance. Numerous scholarly investigations have demonstrated a strong correlation between the seepage of karst reservoirs and the evolution and interconnection of subterranean river systems within karst regions [4,14]. For example, Bakalowicz et al. [16] conducted an analysis of the structural characteristics of the karst groundwater system in Lebanon, offering substantial evidence to support the advancement and utilization of karst groundwater. Wang Xiaoxiao et al. [17] developed a numerical model employing Modflow to investigate the groundwater flow of the Xiedalan Reservoir in Chongqing, examining the prominent seepage pathways within the reservoir.

Despite the favorable outcomes attained in previous studies on the seepage of karst reservoirs, variations in geographical location and hydrogeological conditions give rise to discrepancies in the seepage phenomenon of such reservoirs [15,18]. Consequently, it is imperative, given the distinctive nature of water conservancy engineering, to identify the seepage pathway of karst reservoirs and address the seepage issue through appropriate methodologies.

The Yibasan Karst Reservoir (YKR), situated in Pojiao Town, Maguan County, Yunnan Province, China, is a representative karst reservoir that serves as a medium-sized water supply and irrigation system. However, the reservoir’s effectiveness has been significantly compromised due to seepage issues, posing considerable safety risks to its operation. Consequently, identifying the seepage pathway within the reservoir emerges as the foremost concern in addressing its seepage problem. This paper is based on the engineering conditions of the YKR and utilizes hydrogeological mapping, tracer tests, and groundwater dynamic monitoring methods to investigate the hydrogeological structure of the subterranean stream system and the YKR. Consequently, this paper analyzes the engineering conditions of the YKR, utilizing hydrogeological mapping, the tracer test, and groundwater dynamic monitoring to discover the hydrogeological structure of the subterranean river system Yutang (SRSY) and the YKR. The hydrogeological structure disparities between the subterranean stream system and the YKR are also examined, subsequently establishing the extent of the karst seepage zone in the YKR. Based on the aforementioned research, geophysical exploration was employed to ascertain the precise whereabouts of the karst seepage zone within the YKR, thereby furnishing valuable data for the implementation of an infiltration control initiative. This investigation not only yields fundamental information for the exploitation and utilization of karst groundwater resources in the region but also effectively addresses the issue of karst seepage in the YKR, thereby assuming a pivotal role in the advancement of the local economy and society.

2. Study Area

2.1. Geographic Position

The YKR is a medium-sized reservoir for agricultural irrigation, industrial water supply, and downstream cascade power station regulation and power generation, with a total storage capacity of 11,412,600 m³. It (23°5′23.85″–23°7′35.50″ N, 104°14′6.23″–104°19′7.64″ E) is located in Pojiao Town, Maguan County, Yunnan Province, China (Figure 1), located at the southern margin of the Yunnan-Guizhou Plateau and mainly in a carbonate area. As a result of the lithology and faults, the landforms are different. The overall elevation gradually decreases from west to east, with the elevation, ranging between 1400 and 1600 m, and the overall topography is high in the south, low in the north, high in the west, and low in the east. The region belongs to the central subtropical low-latitude plateau monsoon climate, and the climate is characterized by a dry winter and spring and humid summer and autumn, with no severe cold in winter and no extreme heat in summer.

2.2. Geological Background

The YKR is mainly located in a carbonate area, where the outcropping beds are mainly Devonian limestone, Permian dolomite, dolomitic limestone, and Carboniferous limestone. In addition, there are also Ordovician and upper Devonian clastic rock strata, whose lithology is mainly sandstone and mud shale. There are two syncline sites in the study area. Due to the late tectonic transformation, the structural traces are relatively broken and the influence on the groundwater system is small. The faults are the main factors that affect the groundwater system in this area. The main faults are F1, F2, F3, F4, and small faults, all of which are compression-torsional reverse faults, with the water-insulating faults having closed fault surfaces and poor permeability. Among them, F2 is the largest fault that affects the reservoir area.

2.3. Hydrogeological Background

Based on the composition of the soluble rocks, lithologic combinations, karst development of the carbonate aquifers, discharge and quantity of the water-falling caves, depressions, karst caves, subterranean rivers, and large karst springs, and the water-rich grade in the study area, the water-bearing rocks were divided into the following groups: the water-bearing rock group of the carbonate fissure cave water and the water-bearing rock group of the carbonate rocks containing clastic rock caves and dolomite cave karst fissure water. Atmospheric precipitation is the main source of groundwater recharge in the area, passing through the Quaternary loose layer or directly through the pores, fissures, and dissolution pipes (sinkholes, kart windows, and shafts) of the rock. Part of the groundwater recharges the deep groundwater and part of it is discharged to the surface in the form of fissure springs, karst springs, or subterranean rivers. The study area is controlled by the lithology, geological structure, geomorphic conditions, and hydrological network. Macroscopically, the groundwater in this area is mainly stored in the subterranean river conduits, which constitute the whole SRSY. According to the hydrogeological characteristics of the study area, it was divided into three pipelines, i.e., the main, northern, and southern conduits of the subterranean rivers. As fault F1 is a compression-torsional fault and the fault plane is tightly closed, the fault gouge can be seen, and its thickness reaches 10 cm in some parts. Fault F1 has been identified as an impervious fault. Furthermore, it can be concluded that the groundwater in the depression in the northern part of the fault does not cross fault F1 and has no hydraulic connection with the northern conduit because of the drainage direction of the drainage hole that has developed around the northern part of the fault. Therefore, fault F1 constitutes the northern impervious boundary of the SRSY. Fault F4 has caused contact between the carbonate rocks and clastic rocks, which constitutes the underground boundary of the southern part of the subterranean river system but not necessarily the southern groundwater boundary of the YKR. In addition, the development of the main conduit of the SYRS is affected by fault F2. Thus, the groundwater runoff in the study area flows from southwest to northeast along the main, northern, and southern conduits of the subterranean river, and it is finally discharged at the outlet of the eastern subterranean river. A hydrogeological map of the study area is shown in Figure 2.

3. Methods

3.1. Tracer Test

3.1.1. Tracer Test Principle

Tracer tests can identify the development of the subterranean river conduits and the connectivity and hydraulic connections among the conduits in karst areas [2,19,20]. The principle of a tracer test is to detect the fluorescence intensity of the water samples using fluorescence meters, which are characterized by fluorescence after dissolution in water [20]. This was used to support the groundwater connectivity between the release point and the test point and determine the hydraulic connection between the groundwater conduit. The tracers that were used in the test were uranine and sulforhodamine. Uranine is easily dissolved in water, and the solution is yellow-red with a strong yellow-green fluorescence. Sulforhodamine is a synthetic material with a bright peachy color that fluoresces intensely in solution. Based on the release point and acceptance point in the field, the tracer dosage was calculated.

3.1.2. Tracer Test Design

The conduits of the SRSY were divided into the main, southern, and northern conduits. To further determine the hydraulic connections between the conduits and the main groundwater sources in the seepage zone of the YKR, a tracer test was conducted. As shown in Figure 2 and Figure 3, the release points for the tracer were all located in the Damagu Reservoir (A1, A2, and A3), and there were five acceptance points for the tracer (B1, B2, B3, B4, and B5). Because the water in Damagu Reservoir is shallow in the dry season, only A1 and A2 had water inflow, while A3 had no water inflow. Therefore, the tracer test was completed in two stages. In the first stage, the tracer test was carried out on the northern conduit. In the dry season, the inlet flow of the A1 subterranean river was maintained at 200 L/s, and the inlet flow of the A2 subterranean river was maintained at 50 L/s. Based on the inlet flow of the subterranean river and the distance from the acceptance point, 17 kg of uranine was released into the subterranean river inlet A1 and 6 kg of sulforhodamine was released into the subterranean river inlet A2. In the second stage, the tracer test was carried out on the southern conduit. The A3 inlet of the southern conduit was about 21.5 km away from outlet B5 of the SRSY, and the inlet flow was about 480 L/s in the wet season. Similarly, the tracer dosage at this point was 20 kg of uranine.

3.2. Groundwater Dynamic Monitoring Principle and Design

Groundwater dynamic monitoring refers to the changing processes of the groundwater level, water temperature, water quantity, and water quality under the influence of natural factors, such as precipitation, surface runoff, and human exploitation. It can provide basic information and a guiding basis for the exploitation, utilization, and management of groundwater [21,22,23]. Moreover, the dynamic change in the groundwater level can effectively reflect the groundwater recharge source, groundwater system, hydraulic connection of each monitoring point, and groundwater type [23]. In addition, since the karst groundwater is mainly composed of a subterranean river and conduit flow, the variation in the groundwater level in each conduit can fully reflect the hydraulic connections and hydrogeological conditions among the conduits. Thus, groundwater dynamic monitoring can be used to identify the karst subterranean river and hydraulic connections between the conduits.

Therefore, according to the distribution of the underground pipeline of the SRSY in the study area, four monitoring points were arranged in the south subterranean river conduits of the SRSY, namely, C1, C2, C3, and C4; in the main subterranean river conduits, C5, C6, C7, C8, and C9; and C10, C11, C12, and C13 in the north subterranean river conduits (Figure 3).

3.3. Geophysical Exploration Principle and Design

The geophysical exploration method is based on the physical property differences in density, magnetism, electricity, elasticity, and radioactivity among the underground media. By observing and studying the spatial or temporal distribution laws of artificial excitation or the natural physical field, the structure, material composition, formation, and evolution of the underground media can be determined. The various natural phenomena and changes related to them have previously been clarified to identify geological structures and to solve hydrological and engineering geological problems [24]. In terms of electrical properties, intact limestone and siliceous rocks have a high resistance, while sand, mudstone, bedrock fractured zones, and strong karst development zones have a low resistance. In the study area, karst fissures, caves, and pipes are common in the Devonian limestones. When a medium, such as air or water, fills the karst fissures, caves, and pipes, its conductivity, dielectric constant, elasticity, and density are significantly different from that of the surrounding rock, so it can be used to determine the karst leakage zone. The electrical parameters of some of the most common media are shown in

Table 1. The relative dielectric constants and resistivity of different media.

Medium	Relative Dielectric Constant (εr)	Resistivity (Ω·m)
Air	1	+∞
Freshwater	81	10–100
Sandy shale	5–15	10–1000
Siliceous rock	3.5–10	500–20,000
Intact limestone	4–8	>5000
Fractured limestone	7–10	500–5000
Wet clay	15–40	10–200

.

The audio-frequency magnetotelluric method is suitable for detecting a low resistance zone and geological structure, and it also has high accuracy in verifying the plane position of a subterranean river conduit. When the audio-frequency magnetotelluric method is used to detect low-resistivity bodies, such as karst fractured zones or water-filled caves and conduits, a low-resistivity abnormal trap or low-resistance zone will appear above the anomaly in the One-dimensional inversion resistivity contour. According to these characteristics, the development locations of karst underground conduits, karst caves, fault-fractured zones, and karst-fractured zones can be inferred. So, to ascertain the location of the karst seepage zone in the YKR the study area, one geophysical exploration line is laid out in the potential karst seepage zone and groundwater flow direction perpendicular to the YKR (Figure 3). Then, the audio geo-electromagnetic method was selected.

To find out the hydrogeological structure of the SRSY and the conditions for constructing reservoirs in karst depressions in the study area, the characteristics of the YKR leakage zone were comprehensively analyzed. A tracer test was carried out and several groundwater level dynamic monitoring points were set up in each conduit. Then, a geophysical exploration line was laid out at the location of the potential leakage zone; a distribution map of each working point is shown in Figure 3.

4. Results and Discussion

4.1. Hydrogeological Structure of the Karst Subterranean River

4.1.1. Analysis of the Upstream and Downstream Connectivity among the Karst Conduits

Uranine was monitored at B1, B2, and B3 for the subterranean river inlet A1. Since the groundwater at B2 and B3 was in an approximately static state, it was assumed that the water only flowed in the lower part, so no tracer was detected during the monitoring process in B2 and B3. The distance between the A1 subterranean river inlet and the B1 monitoring point was about 2.75 km. Uranine was detected at point B1 14.74 h after release, and it peaked at 46.35 μg/L after 18.92 h. After 29.49 h, the uranine dropped below the background value. The recovery of the uranine was 6.45 kg and the recovery rate was 37.9% (Figure 4). The distance between the A1 subterranean river inlet and the B4 monitoring point was about 12.81 km. Uranine was detected at point B4 63.5 h after release, and it peaked at 39.22 μg/L after 82.5 h. After 124.7 h, the uranine dropped below the background value. The recovery of the uranine was 5.81 kg and the recovery rate was 35.88% (Figure 5). Therefore, the above results prove that A1, B1, and B4 all had connectivity.

In the analysis of the sulforhodamine monitoring results for A2, the distance between the A2 subterranean river inlet and the B1 monitoring point was about 3.22 km. Sulforhodamine was detected at point B1 24.81 h after release, and it peaked at 2.92 μg/L after 29.01 h. After 33.49 h, the uranine dropped below the background value. The recovery of the sulforhodamine was 2.35 kg and the recovery rate was 39.167% (Figure 4). Therefore, the above results prove that A2 and B1 had connectivity.

Then, in the analysis of the uranine monitoring results for A3, uranine was detected at point B4 39.75 h after release, and it peaked at 21.3 μg/L after 48 h. Also, 7.28 kg of sodium fluorescein was recovered with a recovery rate of 36.4%. Uranine was detected at point B5 62.25 h after release, and it peaked at 15.37 μg/L after 109 h. The recovery of the uranine was 8.12 kg and the recovery rate was 40.6%. Furthermore, uranine was detected at the other points (see Figure 6).

In summary, A1, B1, and B4 all had connectivity, and A1 and A2 also had connectivity. In addition, the distance between A2 and A1 was only 520 m, so the tracer data for the two tracer routes were similar. However, because A1–B1 was shorter than A2–B2, the peak time for the sulforhodamine was later than that for the uranine. Also, A3 had connectivity with B4 and B5. Therefore, it was concluded that the subterranean river in the southern conduit flows into the main conduit from B4 and flows out from B5. Thus, the results provide a basis for the determination and containment of the seepage zone of the YKR.

4.1.2. Hydraulic Connections between the Karst Conduits

Groundwater dynamics are the result of the combined influence of natural and human actions, representing the dynamic index of the groundwater recharge or downward discharge status. The main influencing factors of groundwater dynamics include atmospheric precipitation, submersible evaporation, irrigation, and artificial mining [25,26,27]. The monitoring points for the groundwater level were mainly located in the carbonate fissure karst water-bearing rock group, namely, C1, C2, C3, and C4, which were located in the southern conduit of SRSY; C5, C6, C7, C8, and C9, which were located in the main conduit; and C10, C11, C12, and C13, which were located in the northern conduit. The monitoring data for the groundwater level and precipitation from 10 June to 2 August 2019 were analyzed to determine the dynamic groundwater level and daily precipitation at each monitoring point, as displayed in Figure 7. Firstly, the water level at each monitoring point in each conduit in the SRSY changed with rainfall. The results showed that atmospheric precipitation was the main source of groundwater recharge in the study area but the response time to rainfall and regularity of the groundwater at each monitoring site were different. Secondly, the dynamic curve of the groundwater level at the monitoring points of each conduit had a similar trend, indicating that the monitoring points at each conduit had hydraulic connections and belonged to the same groundwater flow system. Thirdly, the variation in the groundwater level of the monitoring points of the main, northern, and southern conduits was very different. This indicates that there was no hydraulic connection between the three conduits and they belonged to different subterranean river conduits.

4.2. Analysis of the Construction of the Reservoir in the Karst Depression

4.2.1. Catchment Conditions and Boundary Conditions

The strata in the SRSY are mainly carbonate rocks, and the average annual precipitation in the study area is 1710 mm, with abundant rainfall. In addition, the landform is mainly a dissolution depression, and karst phenomena, such as drainage caves, sink holes, and karst windows, are well developed in the region, which makes it easy for the surface water to recharge groundwater, resulting in extremely abundant karst groundwater. There are three subterranean river conduits in the study area, namely, the northern, main, and southern conduits, and there was no hydraulic connection among them. Based on this, a section of the SRSY was plotted (Figure 8). Figure 8 shows that F1 and F4 constitute the northern and southern impervious boundaries of the SRSY, forming the entire catchment range of the SRSY. However, since the northern, main, and southern conduits flowing through the YKR area have no water contact, the northern catchment boundary of the YKR is the same as the northern catchment boundary of the SRSY, which is an F1 compression-torsional fault. Moreover, the southern catchment boundary is the impervious boundary between the main and southern conduits. Based on the above results, the impervious boundary between the main and southern conduits of the SRSY is the F3 compression-torsional reverse fault, so F3 is the southern catchment boundary of the YKR area. However, since the entrance elevation of the subterranean river of the southern conduit was higher than the water level elevation of the YKR, seepage in the southern conduit was not considered, and the confluence of the northern and main conduits became the main focus.

4.2.2. The Factors That Supported the Construction of the Reservoir in the Karst Depression

Carbonate rocks, such as Devonian limestone, Permian dolomite, dolomitic limestone, and Carboniferous limestone, provided good material conditions for the formation of the YKR. As shown in Figure 9, karst depressions and sinkholes in the study area have developed, providing good channels for the surface water and atmospheric rainfall to rapidly collect and recharge the groundwater. Furthermore, the geological structure of the study area was the dominant factor that determined the construction of the reservoir in the karst depression, especially the fault structure, which plays a controlling role in the development of the karst and forms the water boundary between the SRSY and YKR, creating a large catchment range. The flow rate of the northern and main conduits in the dry season can reach 250 L/s, providing sufficient water for the YKR. Therefore, the large catchment area and water supply provide the hydrodynamic conditions for the YKR, which are also important factors for constructing a reservoir in a karst depression.

4.2.3. Identification and Characteristic Analysis of the Karst Seepage Zone

In Figure 10, the measuring section at 1160–1220 m with an elevation of 1275–1355 m contained low-resistivity abnormal sections, which are presumed to be the development location of the karst conduit, and these local deep low-resistivity abnormal sections are presumed to be caused by iron in the rocks. The measuring section at 1320–1420 m with an elevation of 1235–1375 m contained low-resistivity abnormal sections, which is presumed to be the development location of the fault fractured zone. Combined with the hydrogeological structure and boundary conditions of the YKR, it was inferred that the karst seepage zone is located at the eastern side of the YKR, at the junction between the main and northern conduits. The karst seepage zone is 40–60 m wide and 1275–1355 m deep, so it is suggested that a seepage prevention project should be carried out at this site.

5. Conclusions

(1)

There are three subterranean river conduits in the study area, the northern, main, and southern conduits, and there was no hydraulic connection among them. The compression-torsional faults F1 and F4 constitute the northern and southern impervious boundaries of the SRSY, while the compression-torsional faults F1 and F3 constitute the northern and southern impervious boundaries of the YKR. In addition, the karst seepage of the YKR was mainly caused by the main conduit, which is located on the east side of the YKR. The karst seepage zone is 40–60 m wide and 1275–1355 m deep.

(2)

Reservoir seepage in a karst area is common, which threatens the safe operation of water conservancy projects. In this study, the hydrogeological conditions of the study area, subterranean river conduit distribution, hydraulic connections, and karst seepage channels were determined using a field investigation, tracer test, geophysical exploration, groundwater dynamic monitoring, and other means. The hydrogeological structure and boundary conditions of the FSPRSR and YKR were also ascertained, providing a foundation for the management of karst seepage in the YKR and a reference for the development and utilization of groundwater resources in similar karst areas.