Identification Of Sedimentary Microfacies In The Block Of L-S Of Ordos Basin By Logging Data

Sedimentary microfacies is a basic and important work in the geological research of oil fields and an important part of reservoir description.The L-S block is located on the eastern edge of the Ordos Basin.The strata of this area are dominated by tight sandstones with strong diagenesis and strong heterogeneity,belonging to the low-pore special low-permeability reservoirs,which brings difficulties to the exploration and development of the oil fields.Through the study of ancient geography,there was a general understanding and understanding of the sedimentary environment at that time.For dense lithologic reservoirs,the skeleton parameters and pore permeability vary significantly among different lithologic formations.It is difficult to distinguish the strata and identify the sedimentary microfacies based on poor lithology discrimination.Therefore,a priority work area is needed for lithology.,physical property research,and correct lithology identification.However,due to differences in the methods and operations of the thin section identification results,different identification agencies have shown that the differences in lithology and well logging response characteristics are not significant,which leads to the lack of precise lithology according to the composition of the components,and will affect the subsequent deposition of microfacies.For discrimination,priority is given to the selection of particle size attributes,physical properties such as porosity and shale content.After the use of FISHER to discriminate the lithologic granularity in the working area,the effect was not ideal.The particle size distribution and the Pe log curve were directly derived to obtain the median value of particle size.The particle size parameter was very suitable for the identification of the microfacies.Then analyze the types of sedimentary microfacies,study the response characteristics of different microfacies log curves,establish the log phase model from the genesis angle,Analyze Logging Phases in several aspects,including contact relationship,smoothness,top-bottom contact,and amplitude characteristics of logging curves,and select the GR,SP,DEN,and sonic time difference,the relative center of gravity of the array induction curve,tooth rate,mean median,average slope,variance root of variance,and specific amplitude.Including geological constraints,shale content and porosity,a total of 35 parameters.Samples of sedimentary microfacies parameters were collected and cross-validated by SVM support vector machine.It was found that the success rate of SVM operation prediction did not reach expectations,After increasing the number of samples,using optimal decision,univariate feature selection to reducing redundant feature parameters and assigning weights,the success rate of regional sedimentary microfacies quantitative discrimination was improved with repeating attempts which realized the quantitative identification of sedimentary microfacies in the research area...