3D Localization of Coal Fires Based on Self-Potential Data: Sandbox Experiments

Location of coal fires is indispensable in firefighting planning and engineering. The self-potential method has been used for decades in the qualitative determination of coal fires. However, papers about the 3D inversion of coal fires in terms of self-potential data are scarce. In this paper, a 3D inversion algorithm is adopted to localize coal fires. This algorithm is first benchmarked on three synthetic cases to test the peculiar electrode configuration used in this study. Five sandbox experiments are performed with an electric heater buried at different depths and directions. The first experiment is designed to investigate the temperature variation during the heating process. The other four experiments (experiments #1–4) are conducted to obtain the self-potential anomaly under different conditions. The 3D inverted results show that the position of the heater is well-retrieved. A sandbox experiment with a portion of the sandbox containing burning coal is also carried out. Both resistivity and self-potential data are collected and inverted to obtain the 3D distribution of the causative source current density responsible for the observed self-potential signals. The distribution of the inverted source current density is consistent with the true position of the coal fire.