Search

Spontaneous coal combustion is a serious hazard that affects mining safety. Since the coal seams are thin and their resistivity dramatically changes during spontaneous combustion, locating the burned cavities or caving zones using electromagnetic (EM) methods is challenging. The conventional transient electromagnetic (TEM) method with a loop transmitter is ineffective, as the main targets are mostly horizontal and resistive. We propose in this paper to apply TEM with an electrical source to detect the resistive targets of spontaneous coal combustion. Considering that its areas generally have very complex geometry, we adopt the unstructured finite-element method for accurate EM forward modeling. For the time discretization, we choose the unconditionally stable backward Euler scheme. Numerical experiments for typical spontaneous combustion models show the high sensitivity of TEM for an electrical source to the burned cavities, and based on this we further use the limited-memory BFGS method to do 3D inversions. To obtain high-resolution imaging results, we bound the resistivity based on the measurements to the coal mine. The inversion results show that the 3D TEM inversions can effectively recover the locations and resistivities of the burned cavities, implying TEM is a valid tool for detecting spontaneous coal combustion. [ABSTRACT FROM AUTHOR]