Study On Synthetic Identification Method Of Tight Sandstone Logging

In the exploration of subtle reservoirs in continental strata, sedimentary microfacies research is a very important foundation work. Traditional research methods principal integrated cores, sedimentary facies sedimentary microfacies of logging information for qualitative identification, however the core information is limited and costly. Therefore, identification of sedimentary microfacies of logging data quickly and accurately is the urgent problems in the process of exploration and development of oil and gas fields. In this paper, based on research of support vector machine theory, established a set of precise quantitative identification using conventional logging data logging methods. First, through select more mouth key take heart wells paragraph with traditional of geological method for deposition micro-phase fine divided, select most can reflect deposition micro-phase changes of measuring wells curve form features, and top end of contact relationship and the curve smooth degree, for quantitative description, and based on points shaped theory introduced most can reflect geological conditions of two a parameter-box dimension number mud mass content on formation for geological constraint, established measuring wells phase model; last uses support vector machine (SVM) for automatically classification recognition. A block of sulige for braided River sedimentary environments, are of low porosity and low permeability and tight sandstone reservoirs of low gas saturation, strata heterogeneity, and pore structure of the complex. The practical application of this method in the study of block is good, sedimentary facies recognition accuracy up to90%, added improved logging facies of quantitative methods in the study area, and who have some promotional value.