3D Modeling for Subsurface Anomaly Detection: Improving Tunnel Remediation

By Dr Silpaja Chandrasekar, PhDReviewed by Laura ThomsonMar 5 2025

A recent article in Applied Sciences investigated the impact of abandoned mining tunnels on surface stability in a planned expressway area in Chenxi County using electrical resistivity tomography (ERT). Five survey lines detected multiple low-resistivity anomalies, which borehole verification confirmed as eight water-filled cavities, including six mining tunnels and two karst caves.

A 3D model was developed to visualize these underground structures, revealing variations in anomaly detection across survey lines. A significant anomaly at K33+260K33+350 was identified as three closely spaced tunnels, while another at K33+50K33+110 appeared differently across lines. The findings aid goaf remediation and enhance tunnel detection methods.

Related Work

Past work highlighted the impact of natural and human-made underground voids on surface stability, such as karst formations and abandoned mining tunnels. With China's expanding expressway network, studies emphasized how residual tunnels disrupt stress balance, causing deformation and collapse risks. Geophysical methods, particularly ERT, were widely used for detecting these voids, though challenges remained due to varying backfill conditions.

ERT Survey Methodology

ERT is based on the conductive properties of porous media and involves injecting direct current into the ground through two current electrodes while measuring the potential difference between two potential electrodes. It allows for the calculation of subsurface resistivity distribution. Apparent resistivity (ρa) is determined using the measured current (I), potential difference (ΔV), and a geometric factor (K) related to the electrode configuration. Various electrode arrays, such as Wenner, Schlumberger, and dipole-dipole, are used depending on the study's requirements. The Wenner array, known for its ability to highlight horizontal subsurface features, was selected for this experiment to identify underground cavities better. 

Five ERT survey lines, labeled L1 to L5, were designed based on geological conditions and previous survey data. These lines were oriented from north-northeast (NNE) to south-southwest (SSW), parallel to the expressway, with 10-meter spacing between them. While the initial design length for each line was 900 meters, site conditions led to minor deviations: Line 1 measured 867 meters, Line 2 was 855 meters, Line 3 was 867 meters, Line 4 was 873 meters, and Line 5 was 858 meters. The WDJD-2 DC resistivity and induced polarization system, developed by the Pentium Numerical Control Research Institute, was used for field data acquisition. The electrode spacing was set at 3 meters to ensure sufficient resolution. 

After data collection, the Res2dinv software was used to perform inversion and interpretation, producing five high-resolution two-dimensional subsurface profiles. The study accounted for variations in the depth of investigation (DOI) influenced by electrode configuration and ambient electrical noise. The Wenner array effectively captured horizontal resistivity variations, aiding in identifying subsurface features.

Subsurface Anomaly Detection

The inversion results for lines L1, L3, and L5 illustrate the subsurface resistivity distribution, revealing the presence of low-resistivity zones that indicate underground cavities. Since the results for lines L2 and L4 were similar, they were excluded from the analysis. In the ERT profile of line L1, six low-resistivity zones (V1–V6) with resistivity values below 70 ohm-meters (Ω·m) were detected. Two of these zones were in Zone-1, while the remaining four were in Zone-2. Given the continuous rainfall a week before the measurements, it was inferred that surface water infiltrated these underground cavities. The geological conditions suggest that the two low-resistivity zones in Zone-1 correspond to water-filled karst caves, while the four in Zone-2 are associated with abandoned mining tunnels. 

The ERT profile for line L3 revealed five low-resistivity zones (F1–F5) with resistivity values below 70 Ω·m. Two zones were located in Zone-1, while the remaining three were in Zone-2. These low-resistivity zones were consistent with those in line L1, with some minor variations. F3 in line L3, which appeared as a single zone, corresponded to two smaller low-resistivity zones (V3 and V4) in line L1. It suggests that the anomalies in this area are not fully connected and can be separated into smaller structures.

The findings reinforce the presence of underground cavities and potential water infiltration in the surveyed region. Similarly, the ERT profile for line L5 identified five low-resistivity zones (K1–K5) with values below 70 Ω·m, distributed in a pattern similar to line L3. The low-resistivity zone between K33+260 and K33+350 appeared as three smaller disconnected anomalies, while in lines L1 and L3, the same region showed a more continuous large low-resistivity zone. This variation suggests subsurface connectivity differences, reinforcing the underground structures' complexity.

The consistency of the findings in the three survey lines supports the interpretation that these anomalies correspond to either karst formations or abandoned mining tunnels. Six boreholes, each 40 meters deep, confirmed a karst cave at ZK1 and abandoned mining tunnels at ZK2, ZK4, and ZK5, while no cavities were found at ZK3 and ZK6. The results indicated multiple tunnels in the large low-resistivity zones, validating ERT's effectiveness in detecting underground voids.

Conclusion

ERT identified eight underground voids along the expressway, two of which were karst caves and six abandoned mining tunnels. Closely spaced low-resistivity zones appeared as a single anomaly in inversion results. Borehole data validated ERT accuracy and enhanced precision. ERT effectively mapped underground voids, aiding future construction and void treatment.

Disclaimer: The views expressed here are those of the author expressed in their private capacity and do not necessarily represent the views of AZoM.com Limited T/A AZoNetwork the owner and operator of this website. This disclaimer forms part of the Terms and conditions of use of this website.