Research On The Log Evaluation Method Of Tuffaceous Reservoirs

With the increasing need of oil and gas resources and the deep going study of development and exploration of oil and gas fields, the complex reservoirs have been the research hotspots in recent years. Of those, the reservoir evaluation of volcanic clastic rock and volcanic clastic sedimentary reservoirs has drawn wide attention, but the research level is still not deep enough. These reservoirs, which are named tuffaceous reservoir in this paper, contain tuff and other volcanic clastic materials. The tuffaceous reservoir is widely distributed, such as in Songliao Basin, Hailar basin, Zhungeer basin of China. Compared with normal shaly sand reservoirs, the reservoir has various sources, complex mineral composition and complicated diagenesis, which makes significant variations in reservoir pore structure, physical property and aeolotropism. Besides, the tuff and shale have similarities in many properties, which increase the difficulty to distinguish. So the log evaluation of tuffaceous reservoirs is a difficult problem. The tuffaceous reservoir logging interpretation method in this paper will provide a reference for this kind reservoir logging evaluation, which has an important significance for the exploration and development. Especially when the traditional reservoir productivity is getting serious, economic significance and prospects for development is more obvious.This study takes the volcanic reservoirs, volcanic clastic reservoirs in the Wangfu fault and the volcanic clastic sedimentary reservoirs, volcanic clastic reservoirs of the Hailar-Tamtsag basin as the examples to research.Lithology identification is the basis of logging evaluation. So the first task is to identify the lithology of the study area. The elemental capture spectroscopy(ECS) logging has the high detection sensitivity and the strongly combination, which is the only way to solve the problem of lithology identification from the angle of rock composition in all well logging techniques. We can get the structure information of rock from imaging logging, and the rock compositions from ECS data. In order to improve the accuracy, the author use conventional logging curve with the recognition of electric imaging logging and ECS logging data to identify lithology.Exploration of the tuffaceous reservoir has shown that conventional logging methods are often problematic in evaluating tuffaceous formations. Due to the presence of tuff, the calculation accuracy of reservoir parameters cannot meet the requirements of well logging evaluation. The presence of tuff results in significant variations in reservoir pore structure and associated physical properties. And the existence of the tuff directly affects the calculation of reservoir parameters and the accuracy of reservoir saturation calculation. Additionally, the tuff contains a certain amount of radioactive elements, negatively affecting natural gamma-rays, which reduces the accuracy of the logging evaluation. So the tuff content is an important parameter in the evaluation of the reservoir.The previous often used the calculation method of the shale content to get tuff content. Owing to the differences in the source and composition between shale and tuff, the calculation of tuff content using the same methods for shale content is cannot meet the accuracy requirements of logging evaluation. In this paper, the differences in the log response characteristics(radioactive elements content, the density and so on) between shale and tuff are theoretically analyzed based on their source and sedimentary diagenetic environment. The present study takes the tuffaceous reservoirs in Hailar-Tamtsag Basin as an example to prove the theoretical conclusion. The tuff is then divided into fine- and coarse-grained fractions, according to the differences in the distribution of radioactive elements, uranium, thorium and potassium. Next, a volume model suitable for tuffaceous sandstone reservoirs is established.The response equation is established according to the volume model of the reservoir. The logging response value of the components can be calculated with the core analysis data and the differences in the logging response. According to the comparison of calculation results with three optimization algorithms, the method which combines bacterial foraging algorithm and particle swarm optimization algorithm is chosen, which improves the calculation accuracy of shale and tuff content and lays the foundation for the logging evaluation of tuffaceous reservoir.The matrix parameters are obtained by ECS logging data, and the porosity was calculated by conventional logging curve. Taking into account that the ECS logging can be used to measure the content of the radioactive elements by point, the variable matrix parameters of the rock can be obtained. Because the mineral composition of the condensate reservoir is complex, the matrix parameters are difficult to be determined. But the ECS logging data can solve this problem well. The calculation method of matrix density and neutron using ECS has been developed, but the method to calculate the matrix acoustic time hasn’t been found. When the borehole condition is poor, the neutron and density logging curve will be affected by different degree, especially the density curve. But the acoustic wave curve is little affected by the diameter. So the author tries to use ECS data to calculate the variable acoustic time for calculating porosity when the wellbore is instable. The author take the volcanic rocks in Wangfu fault as the example to find the relationship between elements(Si、Ca、Fe、K、Ti、Na) and the matrix acoustic time. The data show that these elements have no linear relationship, like the matrix density and matrix neutron, with the matrix acoustic time. So the GRNN nerve network is considered to solve this problem, and get a good application.Due to the conclusion that cation exchange capacity(CEC) can reflect the additional conductivity of shale, the difference of the conductivity between shale and tuff is analyzed with the use of CEC experimental data and core analysis data. And the relationship between CEC and the resistivity is found. The relationship is used in the Poupon method to calculate water saturation. And the method is named improved Poupon method. Based on the W-S model, the author gets another method to calculate water saturation without CEC data.Finally, the crossplots of the conventional logging data and SVM classification algorithm are used to identify reservoir fluid. The logging evaluation system suitable for tuffaceous reservoir is established, the results of the application are encouraging.Through the analysis of shale and tuff of logging response differences, the problem to distinguish shale and tuff is solved. Two methods are proposed to calculate the saturation, one is the improved Poupon method and the another one is based on the W-S model without CEC data. The problems about lithology identification, reservoir parameter calculation and oil content are solved which provides reference value for the exploration and development of the tuffaceous reservoir, and has economic significance.