Calculation formulas of geological strength index(GSI) and jointing parameter(JP) of rock strength index(RMi) are derived, and the GSI-JP modified analysis method for coal fracture in wellbore direction is given. Considering wavelet neural network, coal principal stresses with reservoir plane is analyzed, consequently coal seam plane fracture classification method is presented based on Hoek-Brown criterion. According to GSI-JP value, coal seam fracture coefficient size and energy method, four grades of A,B,C,D are divided for coal fracture degree at wellbore direction and reservoir plane, and coal fracture classification stan-dard is determined. Appling the classification method and standard, the distribution regularity of coal fracture is studied for ZP-1 Well and JinCheng region in QinShui basin. The research results show that coal GSI value of ZP-1 well is among 30–50, and JP value of RMI is 0.1–0.25, so coal fracture of ZP-1 is considered as grade B. Coal seam fracture degree of YangCheng, DingDian, FanZhuang, FanZhuang, ChangYe, ZhangZi in JinCheng block is defined as grade A, but northwest region coal seam is belong to grade D. The prediction results agree well with the measured values, therefore the classification method and standard are proved. Depending on research results, computer program is corresponding developed.
Existed rock mass classification standards for home and abroad are determined based on qualitative descriptions and subjective experience of individuals to achieve rock mass classification. This classification standard is not applicable for coal seam with depth of a few hundreds meters, thereby the distribution of coal fracture prediction results is affected. Because coal classification method and standard are yet been established for the current domestic to guide coal seam fracture zones forecast, internal factors and external factors of coal fracture is not comprehensive enough, and the safety requirements cannot be met for coal seam drilling.
