Assessment of the Condition of the Foundations of a Building in a Mining Operations Area at Risk of Sinkholes—A Case Study

Abstract

Sinkholes caused by historical underground mining operations are significant geotechnical and safety hazards for new residential developments. This paper presents a case study concerning the assessment of the condition of the foundations of a planned multi-family residential building located within a former mining operations area in southern Poland, which is exposed to the risk of discontinuous ground deformation. This study aimed to identify potential voids within the rock mass and develop safe structural solutions for building foundations. To this end, a comprehensive site investigation was conducted, including two-dimensional electrical resistivity profiling to detect zones of high-resistivity anomalies. High-resistivity anomalies were identified beneath several building segments, suggesting the presence of voids or loose soil resulting from shallow coalmining operations. Based on these findings, a finite element analysis (FEA) of the reinforced concrete foundation slab was performed to simulate the presence of subsurface cavities. The results indicated local tensile stress in the slab of up to 0.34 MPa, which necessitated subsequent design adjustments. Consequently, the use of additional bottom reinforcement and continuous reinforced concrete ribs was proposed to enhance structural safety. This study highlights the necessity of detailed geotechnical and geophysical analyses of planned development zones located in former mining operation areas to address the risks related to sinkholes and ensure the long-term safety of new buildings.

1. Introduction

The rapid expansion of urban and residential developments in suburban areas, as observed in recent years, as well as issues caused by climate change, have increased the frequency of sinkholes [1]. Sinkholes may result from natural processes (e.g., karstification, related to the dissolution of soluble carbonate rocks, such as salt rock, gypsum, or limestone) or human activities (such as mining operations and the related changes in groundwater levels, water permeation through leaky systems, and excessive surface loads) [2,3,4,5]. The risk of karst sinkhole formation is a global issue, affecting Mexico [6], the United States [7], Iran [8], the United Arab Emirates [9], and Türkiye [10]. Among human activities, underground mining operations are the most impactful. Underground mining significantly degrades the environment [11,12] in terms of continuous and discontinuous deformations, including sinkhole formation [13,14,15]. Mining causes degradation of underground structures [16,17], formation of voids, and subsidence [18]. Mining operations affect the soil [19,20] and both underground infrastructure [21] and above-ground developments [22,23,24]. Mining damages are commonly reported in countries where underground operations are conducted, such as China [25], Italy [26], and Poland [27,28]. However, it should be noted that discontinuous deformations, including sinkholes, are often impossible to predict, and their effects may not manifest until much later [29,30]. In recent years, this issue has become apparent in Poland [31] as a result of shallow mining operations. Discontinuous deformations are particularly dangerous to human life and health. An example of such a hazard is sinkholes in residential districts, which pose a tangible threat to the inhabitants’ safety and might force them to evacuate at some point.

Therefore, even though the scale of mining operations is considerably smaller than in the 1950s, sinkholes have become a significant problem which requires careful study by geologists, physicists, and construction research and development specialists. In the early 21st century, several valuable publications appeared that significantly improved the state of knowledge on sinkholes [32,33,34]. Szajna and Gontaszewska [35] reported the results of field studies of Quaternary deposits around sinkholes resulting from former coal mining operations in the vicinity of Zielona Góra. The authors presented their analyses, which indicated that the mechanical properties around the sinkhole areas had changed. Strzałkowski [36] assessed the risk of sinkholes and supported his conclusions with several analyses. Other relevant publications on the subject concern monitoring areas which are at risk of sinkhole formation. For instance, Ferentinou et al. [37] discussed sinkhole detection methods using both satellite imaging and observations on the ground. Malinowska et al. [38] described the use of Persistent Scatterer Interferometry Synthetic Aperture Radar (PS-InSAR) for monitoring the surface above the decommissioned coal mines. This method is well-known and broadly applied in studies conducted in areas which are at risk of sinkhole formation, as well as in areas where sinkholes are already present. The risk of sinkhole formation in mining operation areas can also be analysed using electrical resistivity tomography (ERT) [39].

However, geological and geophysical investigations often fail to produce a comprehensive image of the risks that may emerge when former mining operation areas are repurposed and developed [29,30]. Examples of sinkholes in developed areas are provided in Section 2.1. However, this issue does not only affect existing structures. For areas designated for new investment projects, if the designers are fully aware of the geological conditions, it is possible to anticipate the risks and design the planned developments accordingly. Therefore, this study is a case study of a planned development in an area which is at risk of sinkhole formation due to historic mining operations. In the area of the planned multi-family residential development, an investigation of the geological and mining conditions indicated a risk of voids in the rock mass and, consequently, discontinuous ground deformations in the form of sinkholes. Since such deformations could compromise the safety of the planned structure, it was necessary to carefully identify the geological and mining conditions in terms of the sinkhole formation risk, and to design the structure in a way that would secure it against potentially hazardous ground deformations. For this purpose, a geophysical investigation of the development area was conducted, followed by a numerical analysis of the foundation of the planned building. This study also proposes a concept for securing planned multi-family residential developments.

2. Building Description—Subsoil

2.1. Description of the Object of Analysis (Sinkholes)—Selected Examples

Sinkholes resulting from mining operations threaten not only the site directly affected by ground deformation but also adjacent areas. The area of effect of sinkholes may be exceptionally large, posing a considerable risk not only to buildings but also to urban infrastructure.

Upper Silesia is the region most commonly associated with the mining industry [40]. As evidenced by land surveying (GPS measurements, precise levelling, measurements of subsidence of buildings and technical structures) and satellite imaging (including InSAR–Interferometry Synthetic Aperture Radar, InSAR), the area that is most exposed to sinkholes in Upper Silesia includes nearly 512 sq. km of subsided land, of which 312 sq. km is an area with rapidly increasing ground deformations [41,42].

A prime example of sinkholes in Upper Silesia is the ground deformation caused by shallow mining operations in Jaworzno (on the site of the former Jan Kanty mine). Investigations conducted using ground-penetrating radar (GPR) and microgravimetry revealed the presence of voids under the foundations [39]. The identified gravimetric anomalies and strong radar signals confirmed the voids, which had likely been formed as a result of secondary sinkholes related to former mining operations. The sinkhole appeared in December 2024 at the intersection of Moniuszki and Martyniaków streets; the area is owned by the Chrzanów Forestry Service Office (Figure 1a,b). Another example of ground deformation identified by the Central Mining Institute (GIG) was the sinkhole in Jaworzno, which appeared in December 2017 in S.F. Mazur Street in the Podwale district. The 10 × 10 m, 4 m-deep sinkhole formed in the middle of a large residential area, between apartment buildings (Figure 1c,d). Similarly to the previous example, this deformation resulted from shallow mining operations in the area.

Another example is the sinkhole formed in the Dańdówka district of Sosnowiec in 2013 (Figure 2a). The sinkhole appeared in the immediate vicinity of apartment buildings, destroying part of a children’s playground (Figure 2b). As such, it is a significant hazard to the safety, health, and life of children. The local inhabitants were evacuated. The sinkhole was shaped as a 10–15 m-deep funnel with a diameter of 15 m. The cause of the deformation was the ventilation shaft of the former Niwka-Modrzejów mine, located nearby [44].

2.2. Description of the Structural System of the Planned Multi-Residential Building

In accordance with the local development plan, a five-story apartment building with an underground parking lot was constructed on the site (Figure 3a). The overall dimensions of the building are 85 m × 60 m, and the height of the structure is up to 20.4 m. The building has an irregular layout plan and is divided into 10 segments (A–J, Figure 3b), each with a length of less than 45 m. Structural expansion joints were made between the segments, and the width of the joints ranged from 5 cm (foundation slab) to 10 cm (superstructure) (Figure 3c). The main frame of the building is constructed using reinforced concrete columns arranged at 5.0–7.5 m spacing. The building is founded on a reinforced concrete slab with a foundation slab thickness of 0.55–0.6 m. The building foundation depth was approximately 4.0 m. Under the columns, the foundation slab was thickened by 0.3 m. Along the structural expansion joints, additional strip footings were provided under the foundation slab; the strip footing thickness was 0.3 m (Figure 3c).

2.3. Mining Conditions

The development site is located within the former “Katowice I” mining operations area, where shallow mining was conducted. The area is currently at risk of discontinuous deformation. Furthermore, the local Carboniferous overburden is small, and there is a natural gas hazard.

The seam that was exploited closest to the surface, directly below the development area, was seam 510, which was 5.0 m thick. The seam was exploited from 1929 to 1931 at ca. 75–85 m below the ground using a hydraulic backfill system. Furthermore, two mining operations were conducted in the 1930s at a site directly adjacent to the southwestern border of a building plot. Seam 504 (thickness: 1.8 m, ca. 60 m below the ground) was exploited in 1937 using a hydraulic backfill system, and seam 501 (ca. 50 m below the ground) was exploited in 1936 using mine galleries. Currently, no coal mining operations are planned for this region.

Historical shallow coal extraction conducted in the early 20th century caused local loosening and partial collapse of the roof strata, resulting in increased rock mass discontinuity and karst-like voids. These conditions lead to a reduced bearing capacity of the soil and pose a risk of sinkhole formation, affecting the foundations of modern structures. Consequently, all new construction projects in the area must include geophysical and numerical analyses to ensure their safety and resilience.

2.4. Geological and Engineering Conditions

The thickness of the Quaternary overburden above the Carboniferous deposits in this area is 10–15 m. The subsoil structure was as follows:

• From the surface to ca 0.5–1.9 m below the ground level: Quaternary rock formations, represented by Holocene deposits.
• From 0.5–1.9 m to 4.2–6.3 m below the ground level: Pleistocene fluvioglacial deposits, represented by silt, loamy sand, clay, silty clay, sandy loam, and fine and silty sand.
• From 4.2–6.3 to the geological boring depth (10.0–12.0 m): glacial deposits, that is, sandy loam, sandy loam with gravel, loamy sand with gravel, and fine sand inserts.

At the foundation depth (ca. 4.0 m below ground level), the subsoil is composed mainly of hard sandy loam with local additions of gravel. The assumed modulus of deformation of the soil is E₀ = 30 MPa.

Groundwater is present in the subsoil (encountered at 4.2–6.5 m below the ground level; stabilised level between 2.8 and 4.6 m below the ground level). The presence of groundwater was related to fluvioglacial and glacial deposits in the area.

3. Methods

A geophysical survey was conducted to investigate this subsoil (Section 3.1). A building foundation model was then developed in accordance with the identified subsoil structure. Details of the applied method are provided in Section 3.2.

3.1. Geophysical Survey

To obtain information on the current state of the rock mass and make it possible to determine the risk of discontinuous deformations, a geophysical survey was conducted. The choice of a specific methodology and measuring equipment depends on the local geological structure of the investigated structure and its position in space. The data obtained using two conventional electrical resistivity methods, that is, electrical resistivity profiling and vertical electrical sounding, are one-dimensional. This means that the electrical resistivity differences in the geological medium for electrical sounding are determined vertically, whereas in electrical resistivity profiling, electrical resistivity differentiates horizontally [39]. The profiling method is used when the borders between rock formations are either steep or vertical. The electrical sounding method was used when the borders were either horizontal or slightly tilted. The results of the electrical resistivity tests are presented as vertical profiles, sections, and maps of electrical resistivity, with geophysical, geological, and engineering interpretations.

In this study, two-dimensional electrical resistivity profiling was applied in a medium-gradient system. This method involves a series of measurements taken using a specific measurement system along the profile line, with specified sampling steps, while maintaining a constant distance between the electrodes [45]. Changes in the apparent resistivity were measured at a specific depth along the profile line. The depth range of the method depends mainly on the distance between the current electrodes, the resistivity of the relevant stratum through which the current flows, and the soil water content in the area. Different electrode spacing arrangements were used to investigate strata located at various depths. In this study, measurement spacings of 90 m and 150 m were used. The theoretical depth ranges of such spacings are approximately 20–22 m and 35–40 m. The measurements were performed within a 5 × 5 grid. Based on the results of the electrical resistivity profiling, the curve of the apparent resistivity changes along the profile line is drawn. Interpretation of profiling curves provides data on subsoil parameters, tectonics, location of lithologic complexes, shape and location of fractures, location of interfering objects, and areas of high- and low-resistivity formations [39].

In theory, multielectrode resistivity profiling allows the estimation of anomaly depth by analysing the vertical gradient of the apparent resistivity. However, in this case, the exact depth of the anomaly could not be determined because of the high electrical heterogeneity of the near-surface layers, interference from buried utilities, and limited electrode spacing. This constitutes a limitation of the applied method [46] that should be acknowledged when assessing sinkhole risk.

Although high-resistivity anomalies may result from air-filled cavities, loose soil zones, or anthropogenic backfills, distinguishing between these scenarios based solely on resistivity data is challenging. In practice, integration with borehole testing, seismic profiling, or ground-penetrating radar (GPR) is recommended [47]. In this study, the most conservative interpretation, assuming the presence of voids, was adopted for the numerical simulation to ensure safety.

3.2. Numerical Model of the Foundation Slab in the J Segment

The foundation slab on the Winkler elastic foundation was used for the calculations, in which a void (corresponding to a sinkhole) was simulated. The slab dimensions conformed to the design, as shown in Figure 3b. The columns and walls based on the foundation slab were not included in the model. A void (representing a sinkhole) was simulated, and its geometric dimensions and location corresponded to the largest high-resistivity anomaly found in the geophysical survey.

The calculations were performed in AutoCAD ROBOT Structural Analysis 2024 software using the finite element method (FEM). A discretisation grid was applied with a 0.2 × 0.2 m cell size. The number of calculation nodes was 9255. Elastic foundations were assumed under the entire foundation slab area, excluding the zone representing the void (Figure 4). The modulus of subgrade reaction, k_s = 9.2 MN/m³ was applied in accordance with [48] for the assumed soil parameters (E₀ = 30 MPa, see Section 2.3) and load σ = 200 kPa.

The Winkler foundation model does not explicitly consider soil arching; therefore, the obtained slab deformation represents the upper bound of possible displacements. The calculations performed for the 2D model assumed the worst-case scenario of a void with an undefined depth. In reality, the depths of the voids that occurred are finite, and the results for the 3D model would have been more favourable.

The Winkler foundation model was adopted due to its simplicity and suitability for simulating local loss of subgrade stiffness. Although it neglects shear interaction between adjacent soil springs, it allows efficient evaluation of stress redistribution in the slab. This approach is appropriate for a preliminary risk assessment, while more advanced continuum-based soil models could be applied in future research for more precise prediction.

The model assumed the self-weight load of the foundation slab and (evenly distributed) operational load of p = 3.0 kN/m² representing the load of a cargo-carrying truck on the parking lot floor, as per EN 1991-1-1 [49]. Other loads, e.g., of the superstructure, were not included.

The calculations were performed on characteristic loads. The formation of the void (sinkhole) should be considered an exceptional event. In such cases, as per EN 1990 [50], using calculation factors is not required.

The numerical model did not explicitly include column and wall stiffness and loads. This simplification was adopted to focus on the local response of the slab to subgrade loss and to maintain computational efficiency. In reality, the superstructure provides additional restraint, which would likely reduce the maximum tensile stresses obtained. Therefore, the present results provide a conservative estimate of the additional stress state of the slab that results only from the formation of a void under the foundation.

4. Results and Discussion

4.1. Results and Interpretation of Geophysical Survey

Figure 5 and Figure 6 show the results of the geophysical survey. The electrical resistivity recorded during the survey was low (mostly in the range of 7–30 [Ω m]). This confirms the presence in the subsoil of clay and loam deposits—see Section 2.3. Furthermore, on the basis of the apparent resistivity values, the low- and high-resistivity anomaly zones were defined. The anomaly maps for the two analysed measurement spacings are shown in Figure 6. High-resistivity anomalies were found in the central and north-western sections of the development. Those sections correspond to the J segment (Figure 3b, Figure 5a and Figure 6a) and segments A, E and H (Figure 3b, Figure 5b and Figure 6b), respectively.

The dimensions of the anomaly located in the central section of the building (J segment) are 8.0 × 4.0 m. The anomalies within the north-western edge of the foundation slab of the building (segments A, E and H) are 3.0 × 0.5 m: 3.3 × 3.5 m and 9.0 × 2.5 m, respectively. A geophysical survey was not conducted under the existing building (marked in black on Figure 6). The existing building, which will be demolished, is located within the J segment of the planned structure. Therefore, it is unknown if the high-resistivity anomaly zone is actually shaped as shown on Figure 6a, or if its range is in fact much broader, extending underneath the existing building.

Furthermore, a small high-resistivity anomaly was identified in the southern section of the development area—see Figure 6. High-resistivity anomalies might indicate the presence of a dry zone, loose soil, voids in the subsoil, or local replacement of soil [51].

Low-resistivity anomalies have been identified in the eastern section of the investment site, outside the planned building (on the site of the planned service yard)—see Figure 6. This anomaly probably indicates water saturation of the soil due to the presence of groundwater at 6.9 m below the ground level (Section 2.3).

The anomalous zone in the central section of the building (J segment) is located closer to the surface than the other zones, since its presence was identified with smaller measurement spacing. Based on the results of the geophysical survey, it is not possible to determine the exact depth at which the soil resistivity occurs (Figure 5 and Figure 6). Furthermore, in the anomaly zones, the thickness, and parameters of the subsoil to the survey depth were not determined. Therefore, it is not possible to unequivocally determine if the observed anomalies are caused by loose soil, local replacement of soil, or a void in the rock mass. Given the history of the mining operations in the area, it can be assumed that the survey results likely point to the most unfavourable scenario, i.e., the void in the rock mass. This might lead to the formation of sinkholes in the area. Therefore, for further studies, it was assumed that there were voids in the subsoil in the anomaly zones identified by the geophysical survey. Since the locations of the anomalies observed at different measurement spacing do not overlap, it was decided not to extend the area of possible sinkholes. Consequently, the void dimensions identified by the geophysical survey were used for further analysis. Because the largest anomaly was found in the J segment, the numerical calculations shown below were performed for that segment.

4.2. Results and Analysis of Numerical Calculations

Figure 7, Figure 8, Figure 9 and Figure 10 show the numerical analysis of the J segment. The analysis was performed for the combined foundation slab self-weight and operational loads (3.0 kN/m²).

The obtained results indicate that the tensile stress at the bottom of the foundation slab may be up to 0.34 MPa in the middle of the assumed void area (sinkhole). Therefore, to safely transfer the additional extreme impacts caused by the potential void, an additional reinforcement at the bottom of the slab was calculated and proposed in the form of Ø12 bars arranged at 0.25 m spacing. The area of the additional reinforcement should be at least 10.0 × 6.0 m.

Considering the foregoing, a set of recommendations was formulated for the foundation slabs of segments A, E, and H, under which voids might also form (see Figure 6b). It is recommended to design a reinforced concrete rib with an appropriate cross-sectional size to carry additional loads along the entire north-western wall of the building. Figure 11 shows an example of the proposed foundation structure solution in the vicinity of the north-western wall.

Additionally, adding two layers of bitumen felt between the rib and the foundation slab was proposed. It was also pointed out that the rib should not be divided longitudinally. In the proposed solution for securing segments A, G and H, additional reinforcement of foundation slabs will probably not be required.

Although a detailed verification of reinforcement effectiveness was not within the scope of this study, future numerical and experimental investigations are planned to quantify the stress reduction and serviceability improvement resulting from the proposed measures.

5. Conclusions

Building new structures in the Upper Silesia region should always be preceded by an appropriate investigation of the local mining and geological conditions. The conducted geophysical survey revealed the presence of anomalous zones in the subsoil. The locations of high-resistivity anomalies differ depending on the measurement spacing used. However, it is not possible to unequivocally determine the cause of the anomalies. Therefore, the authors assumed the presence of voids in the subsoil as the most unfavourable but most likely scenario given the history of the mining operations in the area. Furthermore, in order to determine the presence of potential voids, it was recommended to conduct another subsoil investigation once the excavation is made. It is therefore recommended that subsequent investigations include targeted boreholes or microgravity measurements to validate the resistivity findings. The void dimensions assumed for numerical calculations corresponded to the dimensions of the anomalies identified by the geophysical survey. On the basis of the performed numerical analysis, it was recommended to provide additional reinforcement of the foundation slab of the J segment and to design an additional reinforced concrete rib under the north-western wall of the building (segments A, E and H) due to the risk of void (sinkhole) formation. Implementing the recommendations listed above should ensure that the building can be safely used both before and after the potential formation of voids in the rock mass.

The long-term evolution of detected anomalies depends on groundwater fluctuations, progressive roof collapse in historical workings, and cyclic loading from the superstructure. Consequently, periodic geodetic and strain monitoring is recommended after construction. Settlement markers and embedded strain gauges in critical foundation sections would allow early detection of abnormal deformations and ensure the long-term resilience of the structure. These measures form part of a sustainable risk management strategy for construction in post-mining areas.