Study On Lithologic Seismic Inversion Methods For Coal And Gas Outburst Prediction

Coal and gas outburst accidents considered as "the first killer" of the coal mine,has resulted in huge loss in China. With complex structure and largeground stress,coaland gas outburst easily occur so it is urgent to deeply study the geology of the mineand the gas occurrence to provide the reliable geological protection and scientificbasis. Deformed coal is necessary condition for coal and gas outburst and as ahigh-risk coal it is the significant basis to evaluate the gas accidents risk.In this paper, the development of deformed coal in15#coal layer has studied bythe lithologic seismic inversion methods which include PNN, elastic and simultaneousinversion methods, then gas outburst risk can be assessed and predicted from the pointof view the deformed coal’s distribution.Firstly, the significant differences of the physical and chemical propertiesbetween the deformedcoal and bituminous coal are described as well as petrophysicalparameters characteristic of the deformed coal. Petrophysical conversion formulasabout common physical quantities are summarized and ordered and it also providesthe basic theoretical basis to evaluate deformed coal.Secondly, the probabilistic neural network inversion technique is chosen to applyto complete the porosity inversion which avoids the inherent limitations of theconventional inversion. The theory of attributes inversion methods is analyzed fromthe introduction of attribute technology and model-based inversion theory and finallythe chapter focuses on the theory of probabilistic neural network inversion.Thirdly, the theory foundation of pre-stack inversion is Zeoppritz equation andapproximate formula based on the equations. Based the approximate formulas, elasticimpedance inversion is elaboratedas well as simultaneous inversion which is the otherindependent pre-stack inversion method.Fourthly, the risk evaluation of coal and gas outburst hazard has completed bythe lithologic seismic inversion method in Yang Quan coal mine. The pre-stack anglegathers and common post stack gathers are prepared after the fidelity processing forinversion and PNN inversion，acoustic inversion，elastic inversion and simultaneousinversion methods are applied to obtain porosity,AI, EI, ZP, Zs, and μ*ρ volumesin order to interpret the deformed coal distribution.Fifthly, the porosity volume is used as the qualitative explanation, and the crossplot between EI values(no more than0.17) and AI values(no more than0.2) as well as the cross plot between λ*ρvalues(no more than15) and μ*ρvalues(no more than10) is considered as quantitative results to evaluate the distribution of the deformedcoal.Finally, the author has proposed the new concept—comprehensive evaluationfactor X, which is the linear combination of all kinds of lithologic data volumes, todivide the15#coal into five types: no deformed coal area, almost no deformed coalarea, less likelydeformed coalarea, likelydeformed coal area, high-risk deformed coalarea. Based on the information of comprehensive evaluation factorX, the occurrenceof deformed coal is predicted and it is the reliable basis for the evaluation of gasoutburst risk.