摘要
煤系地层富含有机岩石,导致测井响应特征复杂。以实测TOC含量数据为基础,分析了煤系烃源岩中煤岩、炭质泥岩和泥岩有机碳含量的测井响应特征,发现电阻率、声波时差、自然伽马、密度和中子孔隙度等测井参数与有机碳含量相关性较高,并基于此建立了针对不同岩性煤系烃源岩有机碳含量计算模型,即多元回归模型和Δlog R模型。结合研究层段,分别建立了一元、二元、五元线性回归模型及重叠法模型,进行TOC含量计算并做了误差分析。结果表明:五元回归模型的计算误差最小,但在没有实测TOC含量数据层段计算误差较大;重叠法计算误差最大,但其整个层段的TOC含量预测值更符合岩性特征。煤系地层有机碳预测研究旨在为后续煤系烃源岩综合评价提供参考。
Coal-bearing strata are rich in organic rocks,which can result in complex logging response. Based on the determined TOC data,the logging response characteristics of organic carbon content in coal,carbonaceous mudstone and mudstone were analyzed. It was found that the organic carbon content and parameters of the electrical resistivity log,acoustic log,natural gamma log,density log and neutron porosity log were well correlated. According to previous analysis,multiple regression and Δlog R models for calculating the organic carbon content of source rocks with different lithologies in coal measures were obtained. Combined with the study formation,single factor model,binary regression model,five variables regression model and the overlapping model were respectively established. Moreover,the models mentioned above were used for calculating the TOC content and the error analysis were conducted. Results show that the calculation error of five variables regression model is the smallest,but is larger at the intervals where there are no measured TOC data; the calculation error of overlapping method is the largest,while the predicted TOC content value is more consistent with the lithological characteristics over the whole interval. The study of organic carbon content prediction in coal measures aims to provide a reference for the comprehensive evaluation of the geochemical parameter in coal-bearing source rock,which also has some referential value for related application research in the future.
出处
《煤炭学报》
EI
CAS
CSCD
北大核心
2016年第4期954-963,共10页
Journal of China Coal Society
基金
国家重大科学仪器设备开发专项资助项目(2012YQ030126)
国家自然基金煤炭联合资助项目(U1261203)
中国地质调查局资助项目(12120115102101)
关键词
煤系地层
烃源岩
有机碳含量
测井
多元回归
Δlog
R
coal-bearing measures
hydrocarbon source rocks
TOC
well logs
multiple regression
Δlog R