Sedimentary Facies Analysis Method And Software Development,

The research of underground is harder than that of celestial bodies, so thepetroleum indusny heavily depends on high technique than others. It is veryimportant to predict the distribution of sedimentare facies, especially tinysedimentaIy sub-facies in the research of petroleuxn exploration and helpful ofancient physiognomy rebuild, sedimentary condition identify distribution research ofstrata, evolving, storing, cariying and gathering analysis of petroleum.Subjected to actual measure method and techaical level, it is difficult to predictthe distribution of sedimentary facies of underground strata by rule and line. Ascontaining many fields of study in petroleurn exploration, it is necessary to carrythrough various analysis and process in order to foretell facies distribution ofunderground strata by use of seismic data, geological data and well logs datatogether.There is no integrated analysis software of sedimentary facies at present both athome and overseas. Limited fimtions are contained in some foreign software. Theyare deficient both in function and method practically. Identify of sedimentary faciesis mainly by manual work in actual exploration. There are great differences betweendifferent people in analysis as different experiences and criterions.According to the associated training scheme of Institute of Software of ChineseAcademy of Sciences and Daqing Oil Field Corporation, this dissertation dealsdeeply with several importan and difficult problems existed in actual petroleumexploration. Some new algorithIns are presented and integrated analysis software ofsedimentary facies is developed. Various data is fully used in character analysis fromdifferent sources, such as seismic data, geological data and well logs data. Artificialeffects are decreased and realize auomatically identify of sedimentary faciespreliminarily. They provide bases for evaluation of oil and gas reservoit, predictionof reserves and design of well position. The following is the main research works:l. High resolution proce8s technique of sei8mic dat8Due to the high coSts, usually only several wells are driIled during petrOleumexPloration, so few direct information is available, such as well core and log. Seismcdare are indirect dare measurd on the ed's surfaCies, but we can get dare body ofthree dhoensions about Undmpund infOrmation and Strata smictUre. The precisionirnProvement of seismic data is base ofentir works of petroleum exploration.According to seisndc wave tranSmission theory this dissertaion uses fractalmethOd to combine seisndc with sonic log. A tecboque of high-resolution process isdeveloped to recover high frequency of seismic tha has been attenuated by strataabsorphon. Processed seisndc data can shOw the real undergrund thetUre in detall,and enhance the resolution and increasing the signal-noise ratio of seisndc data.2. FuZZy reasoning technique of geologySpecial Anowledge is needed in analysis and differentiating of sedimentaryfacies, and makes choice between severai kinds of knowledge. Many predictions areimpeffect and unsure. Conclusions to the sarn strta from differeni sources andfields ndgh suPport each other or in opposite. Different people usuaJly get differentresultS as manual analysis of sediInentary faCies depending on expeence withdifferent criterion in actual exPloration.ttis dissertation reaIhes integrated reasoning of geology with fuzzy tecboque.Dare StrUtUre of sedimentary facies and repository are contricted. Geologic, welllog and seismic pattems are detennind with different kinds of sediment undertalfOrm differentiating standar. Pattems space of sedAnent forms is denOted byattributes tw. Sedimentary facies is predicted with input adribute charaCters ofgeological, seisndc data and well log thOugh tw reasoning. FUZZy matching ofgeological attributes frame is processed in the entire pattems space with the mostpossibility principle.3. Clustering technique of...