Study On Log Interpretation Method Of Fractured Tight Sandstone

Tightsand oil and gas is the world's oil and gas exploration and development research hot at present, rich in resources, exploration potential. However, the complex causes of dense sandstone, making the reservoir has a dense lithology, low porosity, low permeability characteristics. The pore structure of the tight sandstone reservoir is complicated, showing two or more complex pore structures. The logging characteristics of tight sandstone are more complex,and it is difficult to explain and evaluate the logging.In this study, the 8th Yangtze River Oilfield is taken as an example to study the logging interpretation method of fractured tight sandstone. The sandstone in this area is denser,porosity and permeability are lower,and there are two kinds of reservoirs in the formation. The existence of cracks in the tight sandstone formation makes the reservoir have effective porosity and permeability. In this paper, the fracture identification method of the tight sandstone is studied by combining the geological characteristics, lithology, porosity, permeability, logging response characteristics and core observation data of the target layer in the study area. The logging response characteristics of the fractures are studied, Crack porosity calculation method, the comprehensive probability index method which indicates the development of fractures,calculates the comprehensive indication of fracture porosity and fracture development,and realizes the effective identification of reservoir fractures. Then, the distribution characteristics and pore size distribution of porosity were obtained by a series of experiments such as porosity experiment, nuclear magnetic resonance experiment and its results. The porosity,saturation and permeability of the study area were constructed.Explain the model. In addition, considering the effect of bound water and fractures on the saturation, the rock mass is divided into matrix and fracture according to the theory of double porosity media. The matrix saturation model based on water coupling index and the new saturation of matrix-fracture are deduced Explain the model. The physical and experimental results of rock physics confirm the reliability of the model. Finally,the identification of reservoir fluid and the quantitative prediction of logging parameters are realized by the application of the machine. The engine combines three BP neural networks, support vector machine (SVM) and adaptive - neural network fuzzy reasoning system (ANFIS). Firstly,the three intelligent systems are trained separately to find out the optimal model parameters, And then the results of the three output, and then in some way in the Commission to deal with, and finally get the only solution. In this study, the experimental data of the test data of the Honghe Oilfield and the relevant conventional logging data are used to train and test the machine model of the classification committee and the regression committee respectively. The minimum error parameter model is obtained, and then the trained committee machine is applied to New data, the final realization of the reservoir fluid class prediction, and quantitative prediction of porosity, permeability, water saturation.In this paper, the calculation model of logging parameters is constructed by a variety of methods. The logging interpretation method of fractured tight sandstone strata is studied, and the research and application of the interpretation method of tight sandstone logging are realized.