Optimization And Application Of TOC Logging Quantitative Prediction Model For Tight Oil Source Rocks

In recent years,a large number of tight oil resources have been found in Ordos Basin,Qaidam Basin and Bohai Bay Basin,and large-scale development has been realized.As the key content of tight oil comprehensive research,TOC evaluation of source rocks has attracted much attention in the process of exploration and development.However,due to the number and limitation of experimental analysis samples of source rocks,the experimental results are highly discrete,which restricts the comprehensive evaluation of tight oil resources.Therefore,it is necessary to study the TOC quantitative evaluation method and its applicability based on logging data,so as to provide technical support for its exploration and development.On the basis of sedimentology,petroleum geology and geophysical logging,based on outcrop profile,core observation,imaging logging and analysis data,the source rocks of Chang 7 member of Yanchang Formation in Jiyuan area of Ordos Basin,lower member of Ganchaigou formation in Zahaquan area of Qaidam Basin and Kong 2member of Cangdong sag of Bohai Bay basin are studied,and the logging response characteristics of source rocks are clarified,The method of qualitative identification and quantitative evaluation of source rocks is established,and the best model and applicable law of log quantitative evaluation of organic carbon content of source rocks in different areas are summarized.The specific understanding is as follows:Ordos Basin,Qaidam Basin and Bohai Bay basin formed depression type basins in the deposition period of Chang 7 member,lower member of upper Ganchaigou Formation and Kong 2 member respectively,which provided good reduction conditions for the development of source rocks and formed a large number of high-quality source rocks.The source rocks of Chang 7 member in Jiyuan area of Ordos Basin are mainly gray black mudstone and carbonaceous mudstone;The lower member of upper Ganchaigou Formation in Zahaquan area of Qaidam Basin is mainly composed of sandstone and mudstone;The source rocks of Kongdian Formation in Cangdong sag of Bohai Bay Basin are mainly gray black mudstone and light gray siltstone;The source rocks in Ordos Basin,Qaidam Basin and Bohai Bay Basin are characterized by high organic matter abundance,I-II₁ type and high maturity,which have good hydrocarbon generation potential.According to the analysis of experimental data of source rock samples,combined with conventional logging curve,conventional logging curve overlap method and imaging logging data,the qualitative identification of source rock is carried out.According to the conventional logging curves,the source rocks of Chang 7 member of Yanchang Formation in Ordos Basin have the characteristics of high natural gamma,high acoustic time difference,high compensated neutron,high resistivity and low density;There are two kinds of lithology in the lower member of Ganchaigou hydrocarbon source rock in Zahaquan area.One is calcareous mudstone.The logging curve shows high uranium free gamma,high uranium value,high acoustic time difference and low density;One is marl,which shows high resistivity,high uranium value,low uranium free gamma and low acoustic time difference;The logging curves of the source rocks in the second member of Kongdian Formation in Cangdong sag show high acoustic time difference,high natural gamma,high resistivity and low density.In the conventional logging curve overlapping method,the resistivity curve and the acoustic time difference curve in the pure mudstone section basically overlap,while the two curves in the source rock development section have the characteristics of certain amplitude difference to identify the source rock.The imaging logging data can be used to identify the source rocks according to the bright pattern of the source rock development section.?log R method,multi parameter statistical method and neural network method can realize the quantitative calculation of organic carbon content in source rocks.?log R method has strong applicability in Ordos Basin,but not in Qaidam Basin and Bohai Bay basin;The neural network method has the highest accuracy and the best applicability in each basin,followed by the four parameter statistical method and the last three parameter statistical method;However,in the multi parameter statistical method,the parameter selection of different blocks is different.Among them,the natural gamma ray,density,acoustic time difference and resistivity are the best parameters for the Chang 7 member of Yanchang Formation in Jiyuan area of Ordos Basin,the Kong 2 member of Cangdong sag in Bohai Bay Basin,and the uranium curve,density,acoustic time difference and resistivity are the best parameters for the lower member of Ganchaigou formation in Zahaquan area of Qaidam Basin The four parameters of resistivity are the best.Comprehensive analysis,compared with the four parameter statistical method?log R method and neural network method are more applicable in sedimentary basins with different geological conditions.