Research on Forecast of Coal Spontaneous Combustion in Goaf of Coal Mines

Abstract

The risk of coal spontaneous combustion in goaf of fully mechanized face with high dip and hard roof is high, and forecast of coal spontaneous combustion is difficult. The division method of coal spontaneous combustion “three-zone” by temperature rise and CO concentration in goaf was presented in this paper, and the BP neural network model was used to forecast the temperature in goaf. Tubes and temperature probes were buried in goaf at NO.7162 fully mechanized coal face in Longdong Coal Mine to monitor the temperature and gas concentration. According to the temperature rise and CO concentration in goaf, the range of spontaneous combustion “three-zone” was determined. On the basis of the gas monitoring system with tubes in goaf, the BP neural network model forecasting the temperature of coal spontaneous combustion was established. Taking advantage of the values of CO concentration and CO2 concentration, the coal temperature was forecasted successfully by the neural network model. The results show that the division method of coal spontaneous combustion “three-zone” by temperature rise and CO concentration and the BP neural network model can improve the forecast accuracy of spontaneous combustion, and provide scientific basis for spontaneous combustion prevention in goaf.