Logging Interpretation Research Of Proximate Analysis Of Coal In Exploration Area Of HLC

Proximate analysis of coal is a very important work in the process of coal explorationand development. That is because the proximate analysis of coal not only is one of theimportant indicators to understand the characteristics of coal, but also can serve as a standardof evaluation of coal quality, being of great significance for preliminary judgging the coalpeculiarity, type, all kinds of coal processing effect and its industrial application.The proximate analysis of coal usually can be determined by the methods of coal samplelaboratory test, logging volume model and probability model. In the past, the proximateanalysis of coal is mainly obtained by coal sample laboratory analysis, but this method hassome inevitable defects. However, in the logging data not only contain rich formation underthe primitive state of physical information, also carry plenty of continuous formation andgeological information as the determination results of logging curve. Relative to proximateanalysis of coal of experimental testing method, the geophysical well logging method issimple, need not disturbing the natural structure of coal sample and the state of nature, itmakes up for the lab test of those defects well.The paper based on actual data of HLC exploration area, taking5#coal industrialanalysis index prediction of exploration area as research purpose By means of mathematicalstatistics, regression analysis method, research was carried out by the following:Firstly, analyzed the inner link between the index parameters of proximate analysis ofcoal and amplitude parameters of well logging curves, demonstrating the use of log amplitudeparameters to predict the feasibility of the coal industrial analysis index, laid the theorygroundwork for the research behind; secondly, we set up the monadic regression predictionmodels and multivariate stepwise regression prediction models between the proximateanalysis of coal and amplitude parameters of well logging curves; finally, by comparing the prediction effect of prediction models established anterior, then choosing the regressionmodel which model significance is highest, fitting degree is greatest and regression residualis minimum as the best prediction model of research area.And the precision of predictionmodel was inspected, eventually obtained a ideal effect. The good results obtained in thestudy area showed that in the process of coal field exploration and development, usinglogging data to predict proximate analysis of coal is a fast, economic and practical means，should promotion and application.