Advancing Sustainable Geothermal Energy: A Case Study of Controlled Source Audio-Frequency Magnetotellurics Applications in Qihe, Shandong

Abstract

Geothermal energy is a key part of sustainable and renewable energy strategies, especially for clean heating in northern regions. This study focuses on Qihe County in Shandong Province, applying a controlled source audio-frequency magnetotellurics (CSAMT) method to investigate deep karst geothermal reservoirs. This research addresses the complex geological conditions and electromagnetic interference in the region, aiming to improve sustainable geothermal resource development. The findings indicate that the geothermal reservoir in the study area primarily consists of Ordovician limestone, characterized by moderate burial depth, high water volume, and elevated water temperature. Integrating CSAMT with vertical electrical sounding (VES) and radiometric surveying has clearly defined the deep aquifer layers and major water-controlling fault structures. Drilling verification results demonstrate the significant effectiveness of the integrated geophysical methods employed, providing reliable technical support for deep geothermal exploration in similar regions. This study makes a significant contribution to the scientific and technical foundation necessary for the sustainable development and utilization of geothermal resources, supporting the broader goals of environmental sustainability and renewable energy.

1. Introduction

Geothermal energy is a widely distributed, renewable, and clean energy source. China possesses abundant geothermal resources, accounting for approximately 7.9% of the global total. However, only 5.8% of these resources are currently being utilized, indicating significant potential for further development. Due to changes in China’s energy structure and the urgent need to control air pollution, geothermal resources are extensively used for clean heating in northern regions [1,2,3]. The study area is located in the urban region of Qihe County, Shandong Province, a key city in China’s “2 + 26” atmospheric pollution control initiative. Existing drilling indicates abundant geothermal resources in this area. The reservoirs are primarily composed of Ordovician limestone, characterized by moderate burial depth, high water volume, and elevated water temperature, suggesting promising prospects for exploration and development. Consequently, geothermal exploration in the study area faces two main challenges due to its reservoir characteristics and geographical location: proximity to mature communities within the county, leading to significant human and electromagnetic interference; and faults with considerable vertical displacements controlling the geothermal reservoir, with the deepest parts exceeding 2000 m.

Geophysical methods are the primary tools for geothermal resource exploration. Almost all geophysical methods have been utilized in geothermal exploration, including gravity detection, magnetic soundings, electrical sounding, and seismic exploration [4,5,6,7,8,9,10]. Previous studies have demonstrated the effectiveness of these methods in identifying geothermal reservoirs and delineating subsurface structures. However, geothermal exploration in urban areas presents unique challenges and opportunities for various geophysical methods. Each method has its strengths and limitations, making it essential to select appropriate techniques based on specific site conditions and exploration objectives. Gravity detection is advantageous for identifying density variations related to geothermal reservoirs. It is non-invasive and cost-effective. However, its resolution is limited, making it challenging to distinguish closely spaced geological features in complex urban environments [11,12]. Magnetic methods are effective for mapping subsurface structures and detecting changes in rock magnetism associated with geothermal activity. They are quick to deploy and efficiently cover large areas. Their main drawback is sensitivity to man-made magnetic noise, which is prevalent in urban settings. Electrical sounding methods, such as resistivity and magnetotelluric (MT) surveys, excel in identifying conductive geothermal fluids and providing detailed resistivity profiles of the subsurface. They are highly effective in urban areas with significant electromagnetic interference. However, they are time-consuming and require extensive data processing and interpretation [13,14,15,16,17]. Additionally, they are susceptible to external environmental interference, particularly in urban areas with significant human and electromagnetic noise [18,19]. High-precision 3D and 2D seismic exploration provide detailed subsurface images and aid in detecting faults and fractures that act as conduits for geothermal fluids [20,21,22]. However, this method is limited by high costs, long durations, and noise from traffic, buildings, and other urban infrastructure.

From the above analysis, it is clear that selecting a geophysical method for geothermal exploration in urban areas requires careful consideration of the specific challenges and advantages of each technique. Combining multiple methods often yields a more comprehensive understanding of the subsurface, resulting in a more accurate identification and characterization of geothermal resources. Therefore, this study uses controlled source audio-frequency magnetotellurics (CSAMT) as the primary method to tackle the challenges of exploring deep karst thermal reservoirs in urban areas. CSAMT is an advanced geophysical exploration technology developed from MT. Unlike natural source magnetotellurics (MT), which relies on naturally occurring electromagnetic fields, CSAMT uses a controlled source, providing a higher signal-to-noise ratio and better data quality in urban environments with significant electromagnetic interference, making it a promising method for exploring deep geothermal reservoirs in urban areas. The controlled source also allows for better control over the frequency range, enabling deeper penetration and a more accurate characterization of deep geothermal reservoirs. In recent years, CSAMT has played a significant role in the exploration of mineral resources, environmental geological surveys, and hydrogeological and geothermal investigations [23,24,25,26].

In this study, we aim to explore and develop sustainable geothermal resources in Qihe County, Shandong Province. To address the complex geological conditions and significant electromagnetic interference of the region, we employ the CSAMT method as the primary method for exploring deep karst thermal reservoirs. Additionally, by integrating multiple geophysical methods, we can enhance our understanding of the subsurface, which facilitates more precise identification and characterization of geothermal resources. Therefore, we employed a limited number of vertical electrical sounding (VES) and radiometric surveys within our comprehensive electrical exploration framework to supplement CSAMT. This approach enabled the delineation of deep aquifer layers and major water-controlling fault structures. Subsequent drilling verification confirmed that the employed comprehensive geophysical methods yielded significant results in deep geothermal explorations within the study area. The findings of this research provide valuable scientific and technical insights that support the sustainable development and utilization of geothermal resources.

2. Geological and Geophysical Characteristics of the Study Area

2.1. Geological Characteristics

2.1.1. Structural Characteristics

The study area is located on the alluvial plain of the lower Yellow River, featuring a gently sloping terrain from southwest to northeast, with an elevation ranging from 17 to 30 m above sea level. Structurally, the area is lies on the Luxi Uplift, characterized by well-developed NE and NW trending faults. Figure 1 shows the regional structural map of the study area.

In the geological settings of Qihe County, four major faults have been identified: the Jiaobintun Fault, the Manziying Fault, the Sang Zidian Fault, and the Chang Qing Fault. The Jiaobintun Fault, a key fault in the regional structural unit, is centrally located in the study area. It trends NNE with a dip angle of 40–70°. It is a normal fault, characterized by an uplifted eastern block and a subsided western block. The fault zone exhibits significant vertical displacement, penetrating deep into the strata, with the top of the Middle Ordovician displaced by up to 1000 m. This characteristic makes it the primary heat and water controlling fault in the area. It intersects with three other faults, providing an upward channel for hot water flow.

The Manziying Fault, located in the southern part of the study area, trends NE with a dip angle of 75–80°. It is another normal fault, with the southern block uplifted and the northern block subsided. This fault transects the Ordovician strata, showing a vertical displacement of approximately 200 m.

These faults play crucial roles in controlling the movement and accumulation of geothermal fluids within the karst systems of the study area. They penetrate the deep Ordovician limestone, inducing fracturing and water-bearing properties, and act as conduits for the upward movement of deep thermal sources. Understanding their influences helps with accurately delineating the geothermal reservoirs and planning effective drilling strategies.

2.1.2. Stratigraphic Characteristics

Based on regional geological data and existing geothermal well data, the strata up to a depth of 3000 m in the study area consist of, from the bottom to the top, the Cambrian Sanshanzi Formation (∈), the Ordovician Majiagou Formation (O2-3M), the Carboniferous-Permian Yuemenkou Group (C2-P2Y), the Permian Shihezi Group (P2-3S), the Jurassic (J), the Neogene Huanghua Group (NH), and the Quaternary (Q). The primary component of the thermal reservoir is the Ordovician Majiagou Formation, which has a total thickness of about 800 m and is buried up to 2000 m deep on the downthrown side of the Jiaobintun Fault. The Carboniferous-Permian strata have a thickness of 400–450 m, whereas the Jurassic strata are less commonly distributed, and are found only in some drill holes and geophysical data interpretations, with a thickness of approximately 800 m. The Neogene strata have a thickness of 200 m, while the Quaternary strata are 300 m thick, with the thickness gradually increasing from southeast to northwest.

2.1.3. Hydrogeological Conditions

Based on the characteristics of water-bearing media and the movement and storage of groundwater in the aquifers, the study area can be divided into two major water-bearing rock groups: the loose rock pore water-bearing rock group and the carbonate rock fissure karst water-bearing rock group. These groups exhibit significant differences in water-bearing properties due to variations in lithology and geomorphological settings.

In the shallow layers of the study area, loose rock pore water is prevalent, with aquifer thickness significantly influenced by the Yellow River breach fans. In the upper and middle parts of the breach fans, the aquifer primarily comprises medium to fine sand, with coarse particles, substantial single-layer thickness, and strong water-bearing properties. At the margins and distal ends of the breach fans, the aquifer consists of multiple layers of fine sand and silt, with thin single-layer thickness, interbedded with clay and silty clay, leading to weaker water-bearing properties.

The primary water-bearing layer in the area is the carbonate rock karst aquifer. This aquifer consists of Ordovician limestone, which is light grey to grey-brown, medium–thick bedded, and interbedded with dolomitic limestone and thin-bedded grey-green mudstone containing chert nodules. Karst features, including caves and fissures, are prevalent, often filled or semi-filled with calcite. Small karst caves, usually 2–5 mm in diameter, are common in the shallow parts, with some areas exhibiting a honeycomb structure. The development of karst fissures is significantly influenced by structural features and burial depth, resulting in the clear vertical zoning of water-bearing properties within the Ordovician limestone. The shallow section features well-developed karst fissures and a weathering crust. These facilitate rapid groundwater flow and strong water-bearing properties, which diminish with increasing depth.

2.2. Geophysical Characteristics

The strata in the study area can be categorized into three electrical layers (see as

Table 1. Electrical parameters of the study area.

Strata	Main Lithology	Resistivity ($\mathsf{\Omega }·\mathbf{m}$)
Quaternary	Sand, clay, medium-coarse sand	5–50
Neogene	Mudstone, siltstone, fine sandstone	10–20
Jurassic	Feldspar sandstone	5–570
Permian	Sandstone, shale	5–70
Carboniferous	Shale, sandstone, limestone	5–80
Ordovician	Limestone	50–2000
Cambrian	Shale, limestone	50–500

): low-resistivity Quaternary and Neogene strata, medium-resistivity Carboniferous-Permian and Jurassic strata, and high-resistivity Ordovician strata. The Neogene–Quaternary strata show relatively low resistivity, creating a low-resistivity shield above the thermal reservoir. Fault structures disrupt the continuity and integrity of the original rock, often becoming filled with materials that lower resistivity and form low-resistivity zones. The fracture zones between faults and surrounding rock exhibit significant low-resistivity characteristics. Therefore, the study area offers favourable geophysical conditions for exploration using CSAMT and vertical electrical sounding methods.

3. Geophysical Exploration Methods

3.1. Controlled Source Audio-Frequency Magnetotellurics (CSAMT)

The CSAMT survey lines were arranged in two east–west profiles (L1, L2) and two north–south profiles (L3, L4), forming a cross pattern. The layout of CSAMT survey lines is shown in Figure 2. This arrangement aimed to investigate fault structure development and to delineate the distribution of deep geothermal reservoirs. Based on preliminary survey results from the four profiles, an additional east–west profile (L5) and a north–south profile (L6) were added near inferred fault zones and target well locations. This aimed to conduct focused research on delineating areas of low electrical resistivity indicative of fault development.

During the CSAMT acquisition parameter setup, the minimum transmitter–receiver distance was increased to enhance exploration depth and reduce near-field penetration frequency. To improve the signal-to-noise ratio and suppress noise, several strategies were implemented: (1) using high-current transmission to enhance effective signal strength; (2) increasing the receiving electrode spacing to boost electric field strength; (3) controlling the maximum transmission–reception distance, as it is inversely proportional to the signal-to-noise ratio; and (4) employing large electrode spacing for transmission. Through theoretical calculations and field experiments, the following field data acquisition parameters were determined: a transmitter–receiver distance (R) of 10–16 km, transmitter electrode spacing (AB) of 1.8 km, receiver dipole spacing (MN) of 50–100 m, maximum transmitter current (IMAX) of 18 A, and a working frequency range of 9600–0.125 Hz, with a total of 46 frequency points.

During data processing and interpretation, invalid recording points were removed. Files with frequency-domain apparent resistivity and impedance phases were generated for each recording point. The data were then subjected to static correction, spatial filtering, and other processing before conducting two-dimensional geophysical inversion.

3.2. Vertical Electrical Sounding (VES)

Vertical electrical sounding (VES) was used to supplement the CSAMT survey, particularly in areas with significant electromagnetic interference. Specific points along the CSAMT survey lines were selected to provide additional data on subsurface resistivity distribution. The DC resistivity method utilizes the electric field generated by artificial sources. By varying the electrode spacing AB, current and voltage data at various positions are measured to calculate subsurface resistivity at different depths. This method effectively resists electromagnetic interference and accurately identifies resistivity curve types. However, it generally detects only shallow underground structures and typically has low spatial resolution. Consequently, it was employed solely to validate the results of the CSAMT survey. Figure 2 illustrates the layout of the survey lines.

A vertical resistivity sounding profile was established along line L5, spanning approximately 250 m and including nine receivers. The maximum transmitter electrode spacing (AB) was 5000 m, maintaining an AB/MN ratio of 5.

3.3. Radioactivity Measurement

Studies have shown frequent radioactive anomalies, particularly significant radon anomalies, above water storage structures, geothermal boundaries, and active fault fracture zones [27,28,29,30]. The electrostatic α-card radon measurement method identifies fault fracture zones by measuring changes in radon and its progeny concentrations.

Underground thermal water often contains high levels of radon. Fractures alter the original formation’s compactness, enhancing regional ventilation. This allows radon to migrate and accumulate, enriching radioactive elements. Differences in geothermal temperature, ground pressure, and groundwater convection create an upward-moving air current above the water-bearing fault zone, leading to high radioactive anomalies.

In this study, the α-card method employs an HFS-1 radiation detector to measure along the L6 CSAMT survey line, consisting of 65 points spaced 50 m apart. Pits are dug to a depth of 0.5 m, with cup burial times exceeding 4 h. Each α-card measurement is taken twice: once for 3 min and once for 5 min.

4. Results

4.1. Profile Analysis

4.1.1. CSAMT Profiles Analysis

The collected field data were preprocessed and subjected to inversion imaging. MT data were primarily used to correct deep information derived from CSAMT, while vertical resistivity sounding was employed to verify the accuracy of the CSAMT profile interpretation. Interpretations of strata and fault zones were conducted based on CSAMT data and geological knowledge. Inferences regarding the distribution patterns of concealed faults and geothermal anomalies in the Ordovician system were made. Figure 3 and Figure 4 display the inversion resistivity profiles and interpretation maps for lines L1 through L6. The electrical properties of the strata transition from low to medium to high resistivity from shallow to deep, with contour lines extending horizontally along the survey lines. Locally, dense stepped zones or abrupt folds with clear cross-sectional morphology are observed.

The profile can be vertically divided into three layers: The first layer is a shallow, surface low-resistivity layer, with a thickness ranging from 500 to 700 m and a resistivity of less than 100 $\mathsf{\Omega }·\mathrm{m}$. In the east–west profiles of L1, L2, and L5 (Figure 3), it appears deeper in the west and shallower in the east. In the north–south profiles of L3, L4, and L6 (Figure 4), it is deeper in the north and shallower in the south. This layer is inferred to consist of Quaternary and Neogene clay and sandstone. The second layer is a medium-resistivity layer with a relatively stable thickness, buried at depths of 800–2300 m, and with resistivity ranging from 10 to 80 $\mathsf{\Omega }·\mathrm{m}$. This layer is inferred to be Carboniferous-Permian strata. The top of the Ordovician system is buried at depths greater than 2000 m, reaching depths of up to 2500 m at its deepest points. The third layer is a high-resistivity layer with a resistivity above 100 $\mathsf{\Omega }·\mathrm{m}$, inferred to be Cambrian-Ordovician strata, primarily composed of limestone.

In the inversion resistivity profiles (Figure 3) of the three parallel east–west CSAMT lines, several instances of distorted resistivity contours were observed, extending into the Ordovician strata and down to the basement. Notable distortions occur at approximately 2100 m on the L1 line, at 1500 and 2500 m on the L2 line, and at 2300 m on the L5 line, all at depths greater than 600 m. Based on interpretation principles for fault development in electromagnetic profiles, it is inferred that the L1 profile develops a single fault, F1; the L2 profile develops two faults, F1 and F2, from west to east; and the L5 profile also develops two faults, F1 and F2, from west to east.

In the inversion resistivity profiles (Figure 4) of the three north–south CSAMT lines, resistivity contour distortions are observed at depths greater than 600 m, specifically at 2000 m on the L3 line and 2500 m on the L4 line. Below 1500 m, the high-resistivity layer contours exhibit abrupt discontinuities, clearly caused by faulting. It is inferred that both the L3 and L4 lines develop the F3 fault, which places the southern strata at the footwall and the northern strata at the hanging wall. Since the F3 fault intersects the Ordovician interface at the endpoint of the L6 line, no significant fault zone anomalies are observed in the L6 inversion resistivity profile.

However, some anomalies shown in the figures are not related to geothermal or mineral resources but are influenced by electromagnetic interference from urban areas. In the shallow sections of the L1 line (points 2200 to 2900), the L4 line (points 1000 to 1500), and the L6 line (points 1600 to 2100), secondary high-resistivity anomalies extending downward are observed. Field investigations revealed power lines, streetlights, and other facilities at these locations, suggesting that these anomalies are likely caused by environmental electromagnetic interference.

4.1.2. VES Profiles Analysis

The vertical electrical sounding (VES) of apparent resistivity is a direct current (DC) resistivity method influenced by volumetric effects. The measured apparent resistivity values reflect the integrated response of the strata above the depth corresponding to the electrode spacing. By analyzing the variation in apparent resistivity values at different electrode spacings, and combining this with geophysical characteristics, the strata and faults along the profile were interpreted. By integrating Figure 3b and Figure 5, both methods show good consistency in identifying fault zones and stratigraphic changes. This highlights the accuracy and reliability of the CSAMT profile interpretation and validates it through a successful comparative experiment.

4.1.3. Radioactivity Curve Analysis

After preprocessing α-card data to remove random noise, an α-ray intensity pulse reading profile was created for the L6 survey line. In the radioactivity curve (Figure 6), it is observed that the pulse background values along the entire survey line are relatively low. Using the 2100 m point as a boundary, the α-ray intensity radiation values are lower in the south and higher in the north, suggesting that the profile’s stratigraphic depth is shallower in the south and deeper in the north. The section between 1200 and 1250 shows a higher radioactive peak, presumed to be related to the F2 fault.

4.2. Planar Interpretation

Based on the interpreted results from the inversion resistivity profiles of lines L1 to L6, horizontal slices of apparent resistivity at depths of −1000 m, −1500 m, −1800 m, −2000 m, −2200 m, and −2500 m were extracted (Figure 7). These slices were used to comprehensively assess the distribution of faults and the characteristics of geothermal anomalies.

Analyzing the available data, the primary goal of this geothermal resource exploration is to identify the fracture water-bearing structures within the Ordovician limestone strata. Therefore, the focus of geothermal reservoir exploration should be on identifying low-resistivity anomaly zones within the deep high-resistivity layers. Significant low-resistivity anomaly zones are identified at depths of −1500 m, −1800 m, −2000 m, and −2200 m. These anomalies are inferred to result from fractured lithology, with well-developed fissures forming water-rich zones, which provide essential conditions for geothermal reservoir formation. Within the deep high-resistivity basement layers, a distinct NNE-trending low-resistivity anomaly zone in the western part is identified as the F1 fault, which is characteristic of a typical faulted and fractured zone. The central part features a NW-trending low-resistivity anomaly zone, reflecting the F2 fault. The north-central part features an ENE-trending low-resistivity anomaly zone, indicative of the F3 fault. On the planar map, the F1 fault acts as a boundary with lower resistivity to the west and higher resistivity to the east. Similarly, the F3 fault demarcates a boundary, with lower resistivity to the south and higher resistivity to the north. Near the F2 fault and south of the F3 fault, a prominent low-resistivity anomaly is observed. This anomaly results from the combined effect of the F3 and F2 faults, indicating an electrically conductive faulted and fractured zone. The F1, F2, and F3 faults extend deeply into the basement, facilitating deep groundwater circulation and serving as critical conduits for geothermal mineralization.

4.3. Geothermal Well Design

According to the electrical prospecting theory, geophysical characteristics, and hydrogeological data, geothermal resource-rich areas are generally located near fault zones. The subsurface rocks near fault zones are often fractured, providing favorable conditions for heat and water conduction, which facilitates the formation of geothermal resources.

Based on the low-resistivity anomalies from various geophysical profiles, we identified one geothermal water-rich anomaly zones. This anomaly is located in the western parts of the exploration area, extending in a north–south direction, with an approximate length of 450 m and a width of 400 m. The interpretation results reveal that the geothermal area’s cap rock consists of Quaternary, Neogene, Carboniferous, and Permian strata. The Ordovician strata’s top layer is buried at around −2200 m. The geothermal reservoir is composed of fractured Ordovician limestone, which facilitates hydrothermal exchange and can be a key area for geothermal exploration and development.

The Q1 geothermal well, located 10 km northeast of the anomaly area, was drilled to 1601.57 m, targeting the Ordovician Majiagou formation. However, it did not reach the Ordovician strata due to insufficient designed depth. The TY1 geothermal well, approximately 8 km southeast of the anomaly area, was drilled to 2200 m, also targeting the Ordovician reservoir, but similarly did not reach the Ordovician strata. Since the three geothermal wells are within the same structural unit of the Qihe Sub-uplift, the depth of the ZK1 well will be determined by referencing the drilling experiences of the other two wells. Significant low-resistivity anomalies are observed at depths of −2000 m, −2200 m, −2500 m, and −3000 m, based on the planar interpretation of the area, and the interpreted depth of the Ordovician strata is approximately 2400 m. Consequently, the designed drilling depth for well ZK1 is 2700 m.

4.4. Drilling Verfication

Drilling site ZK1 was strategically positioned near the F1 fault on line L5 (refer to its location in Figure 7), and verification drilling was conducted. Borehole ZK1 reached a depth of 2700 m. The encountered strata, from shallow to deep, included the Neogene Pingyuan Formation (to 330 m), the Quaternary Minghuazhen Formation (to 660 m), the Jurassic (to 940 m), and the Carboniferous-Permian (to 2600 m). The drilling results closely matched the CSAMT detection results, particularly in the stratification of the low-resistivity Neogene Pingyuan Formation and the Quaternary Minghuazhen Formation, confirming the accuracy of the CSAMT processing. The top of the high-resistivity Carboniferous-Permian strata also aligned well with the CSAMT interpretation, verifying the accuracy of the inversion for deep strata.

Chemical analysis of water samples from the ZK1 well showed a geothermal water salinity of less than 1 g/L, a hydrochemical type of SO₄-Na-Ca, a temperature of around 60 °C, and a yield of about 2000 m³/d. The geothermal water in this area is deeply buried, derived from Ordovician karst fissure water, and is nearly free of organic contamination. This geothermal fluid can be used for winter heating. Plans for the complete reinjection of the geothermal tailwater ensure no impact on the ecological environment.

5. Conclusions

The geothermal resource survey in the urban area of Qihe County encountered numerous challenges. After comparing various traditional geophysical exploration methods, this study employed the controlled source audio-frequency magnetotelluric (CSAMT) method and vertical electrical sounding (VES). The CSAMT method is known for its high detection accuracy, strong resistance to interference, and significant detection depth (30 m to 3 km). It is user-friendly, and after near-field correction with MT data, it provides a broad frequency band for inversion processing, making it highly effective in this context. The results indicate three inferred faults within a depth of 3000 m in the study area. The F1 fault trends NNE, dipping west at an angle of 80°; the F2 fault trends north–south, dipping west at an angle of 80°; and the F3 fault trends east–west, dipping south at an angle of 70°.

Based on the comprehensive interpretation of electromagnetic methods, geothermal wells were planned and drilling verification was conducted. The drilling results closely matched the CSAMT detection results, especially in the stratification of the low-resistivity Neogene Pingyuan Formation and the Quaternary Minghuazhen Formation, confirming the accuracy of the CSAMT processing. The top of the high-resistivity Carboniferous-Permian strata also aligned well with the CSAMT interpretation, verifying the accuracy of the inversion for deep strata. In geothermal surveys of northern Chinese cities, the investigation environment is often complex, and many traditional geophysical methods face limitations. In similar areas, it is necessary to combine the strengths of multiple methods and conduct comprehensive geophysical exploration, as this is an effective means to improve the investigation of geothermal resources in urban settings.