Gas coal delamination recognition method based on the Fisher discriminant function of trace element characteristics

Abstract

Abstract Numerous geological factors will affect determining the well bottom position by well depth during the construction of coalbed methane wells, resulting in inaccurate results. By using the trace element characteristics of coal samples to determine the source layer of coal samples, it is possible to avoid making incorrect assumptions. The groove method was used in this study to collect 40 gas coal samples from seven gas coal seams in Dingji Mine, which were then sealed and stored. The contents of 40 different trace elements were analyzed in the samples, and the types and content characteristics of trace elements in each coal seam were determined preliminarily through content distribution map analysis. Simultaneously, the SPSS analysis software was used to analyze the 1600 trace element test data from various coal seams, and cluster analysis was performed using the average link method between classes. The clustering tree diagram was created, and the discrimination was calculated using the distance between classes and samples. This study analyzed the typical explanatory variables and established the Fisher discriminant function for coal seam identification. Accordingly, identifying gas coal stratification requires only measuring the concentrations of nine trace elements (typical explanatory variables) in the coal seam being discriminated against. This method can be used in Dingji mine to determine the final hole horizon of surface coalbed methane wells and eliminates the risk of accidentally exposing coal seams during the excavation of an underground roadway.