Research on site selection model for underground coal gasification based on fuzzy evaluation and BP neural network:A case study of Fukang Mining Area,Xinjiang
YU Yu
YANG Shuguang
ZHANG Na
LI Xin
WANG Haichao
SANG Shuxun
ZENG Zhiwei
TIAN Jijun
LAI Peng
TAO Jian
LIU Yunxuan
Abstract：As a strategic alternative to enhance natural gas reserves and production in China,Underground Coal Gasifica-tion(UCG)technology relies heavily on precise geological evaluation and site selection for safe and efficient develop-ment.However,traditional UCG site selection methods suffer from strong subjectivity and poor timeliness.To address these issues,this study took 162 sets of coal seam data from 26 coal mines in the Fukang Mining Area,Xinjiang,as the re-search object.Based on a comprehensive review of previous geological factors influencing UCG,an evaluation index sys-tem was constructed,encompassing four primary indicators(e.g.,furnace construction feasibility,process controllability),nine secondary indicators(e.g.,wellbore stability,trajectory controllability),and 30 tertiary indicators(e.g.,rock hardness coefficient,coal seam thickness).Using the Analytic Hierarchy Process(AHP)and fuzzy evaluation method,a compre-hensive geological assessment of each coal seam was conducted to identify the most favorable coal seams with UCG de-velopment potential in each mine.Building on these evaluation results,a fuzzy evaluation-Backpropagation Neural Net-work UCG site selection model was introduced.Finally,to validate the model's reliability and efficiency,a case study was conducted on Kanglong Coal Mine in the Fukang Mining Area.The results indicate that geological factors such as coal seam thickness(weight proportion:17.99％),coal seam dip angle(7.33％),and fold complexity(7.26％)are key indicators affecting UCG site selection.The 35-36# coal constitutes the largest share of UCG-favorable coal seams in the study area,reaching 19.23％.The model can rapidly generate site suitability assessment results by inputting standardized geological parameter data,enabling quick and accurate UCG site selection in blank evaluation areas with complex geological condi-tions.These findings are expected to enrich the UCG site selection system and methodologies,providing insights for the clean development of hard-to-mine coal resources in Xinjiang.
Keywords：underground coal gasificationfuzzy comprehensive evaluationBP neural networkFukang Mining Areamachine learning
Publication Date：2025-08-31
Online Publishing Date：2025-09-29(First online date of this platform, not the publication date of the document)
Pages：13( 3830-3842 )