Study On Hydrogeological Control Of 15~# CBM Drainage In Southern Qinshui Basin

In the south of Qinshui Basin,the 15^#coalbed methane resource is larger than that of 3^#coal seam,while 15^#coalbed methane drainage is difficult due to large water production during the production process,and there is no gas production or very few gas production.In order to realize the large-scale mining of 15^#coal bed gas,it is necessary to strengthen the study of the hydrogeological conditions and related processes for the coal seam gas production of 15^#coal seam.Based on the study of Shizhuang in southern Qinshui Basin,this paper systematically studies the hydrogeological conditions of 15^#coalbed methane drainage through data collection,sample collection,theoretical analysis,field test and numerical simulation,and the conclusion is as follows:The 15^#coal seam in the study area is one of the main coal seam.It has good coal bearing property,in addition,its roof is K₂ limestone aquifer and the bottom is a mudstone aquifer.This paper adopts conventional hydrochemical analysis and hydrogen oxygen stable isotope analysis,and 15^#coal seam water production is mixed water of coal seam and roof,and the water quality is HCO₃-Na type,which is deep underground water quality.The stable isotope composition of hydrogen and oxygen in the water production of the 15^#coalbed methane well is not affected by the annual and seasonal changes,and it is evenly distributed on both sides of the meteoric water line.The water quality of two layers of combined mining water is HCO₃-Na type and Cl-Na type,and the 3^#seam in the replenishment source of two-layer combined water production is the major part.The curve of pressure drop funnel in the process of coal mining in 15^#is simulated by using Visual Modflow.When the roof is the permeability of mudstone,the pressure drop funnel curve shows great difference,which indicates that the main source of recharge of the effluent from 15^#coal seam production is the K₂ limestone of the roof.It is suggested to carry out the fracturing test of coal seam roof and the quarrying of quarry wells to improve the productivity of 15^#coalbed methane gas.Based on the idea of time series prediction,a BP neural network model which is suitable for predicting the water production of CBM wells is constructed by using MATLAB software.The results of the prediction of the TS-006 well in Shizhuang,southern Qinshui Basin show that:this BP Neural Network model can accurately predict the water production of the coalbed methane wells in the next 7 days.