Hydrogeochemical characteristics of coal mine based on box-plot and its application in water inrush source identification

Abstract

ABSTRACT In order to meet the growing energy needs, more and more people pay attention to the deep coal resource (>1000 m). Compared with shallow coal mining, deep coal mining is facing more complex problems. Among them, mine water inrush disaster is a difficult problem, which needs to be solved urgently. In this paper, 45 groundwater samples from three deep-seated aquifers (the Quaternary, the Coal bearing, and the Limestone ) are collected. Based on the box-plot analysis of conventional ions, a discriminant model is established. When [K++Na+] >500 mg/L, [Cl−]>450 mg/L, [TDS]>1500 mg/L, the Coal bearing can be distinguished; when [K++Na+]<200 mg/L, [HCO3 −]<250 mg/L, [SO4 2-]>35 mg/L, pH>9, the Limestone can be recognized, and the remaining may the Quaternary. The discriminant model accuracy can reach 96%, 100% and 94% for the three aquifers, respectively. In addition, cluster analysis shows that the water-inrushing identification model is still very satisfactory. The results show that the model based on box-plot has a good discriminant effect and can be used to identify water sources in coal mines, and it can also provide theoretical guidance for the prevention and control of water damage in other coal mines around the world.