Analysis on typical characteristics and causes of coal mine gas explosion accidents in China
Zhijun Lin,Min Li,Shan He,Deming Wang,Shiliang Shi,Dan Wang
DOI: https://doi.org/10.1007/s11356-024-34890-7
2024-09-04
Abstract:Large-scale coal mine gas explosion (CMGE) accidents have occurred occasionally and exerted a devastating effect on society. Therefore, it is essential to systematically identify the characteristics and association rules of causes of CMGE accidents through analysis on large-scale CMGE accident reports. In this study, 298 large-scale CMGE accidents in China from 2000 to 2021 were taken as the data sample, and mathematical statistical methods were adopted to analyze their general characteristics, coupling cross characteristics, and characteristics of gas accumulation and ignition sources. Moreover, the text mining technology and the Apriori algorithm were used for exploring the formation mechanism of CMGE accidents, during which 46 main causal factors were identified and 59 strong association rules were obtained. Furthermore, an accident causation network was constructed based on the co-occurrence matrix. The key causal items and sets of CMGE accidents were clarified through network centrality analysis. According to the research results, electrical equipment failure, cable short circuit, mine lamp misfire, hot-line work, and blasting spark are the key ignition sources of CMGE. Fan failure, airflow short circuit, and local ventilation fan damage are the main causes of gas accumulation. Besides, the confidence levels of two association rules of "static spark-fan failure" and "blasting spark-airflow short circuit" are higher than 70%, indicating that they are the two dominant risk-coupling paths of gas explosions. In addition, six causes appear frequently in the shortest risk paths of gas explosion and are closely related to other causes, i.e., fan failure, local ventilation fan damage, static sparks, electrical equipment failure, self-heating ignition, and friction impact sparks. This study provides a new perspective on identifying causes of accidents and their complex association mechanisms from accident report data for practical guidance in risk assessment and accident prevention.