论文部分内容阅读
为了准确认识和预测伊通盆地鹿乡断陷储层敏感性的分布,从实验分析入手,测量不同样品的敏感性、物性和粘土矿物等参数,结合铸体薄片、压汞、扫描电镜等实验方法,从宏观和微观2个角度分析了储层敏感性与孔隙度、渗透率与各类粘土矿物相对含量之间的关系,分析了储层敏感性与储层的孔喉类型和粘土矿物产状之间的关系,建立了不同微相控制下的孔隙度、渗透率、粘土矿物含量、石英和长石含量的解释模型.最后,选取孔隙度、渗透率、石英含量、长石含量、伊利石含量、高岭石含量、绿泥石含量、伊/蒙混层含量8个参数,采用Elman神经网络方法分别建立了速敏、水敏、酸敏和碱敏的预测模型.结果表明:采用神经网络方法预测的储层敏感性指数与实验结果吻合;五星构造带具有强的速敏、酸敏、碱敏和盐敏,鹿乡断陷中部和西北部具有强的水敏性.
In order to accurately recognize and predict the distribution of reservoir sensitivities in the Lvxiang fault depression of the Yitong basin, the sensitivities, physical properties, clay minerals and other parameters of different samples were measured from the experimental analysis. Combined with the experiment of casting thin film, mercury intrusion and scanning electron microscope Method, the relationship between reservoir sensitivity and porosity, permeability and the relative content of various clay minerals is analyzed from macroscopic and microscopic perspectives. The relationship between reservoir sensitivity and reservoir pore throat type and clay mineral production Porosity, permeability, clay mineral content, quartz and feldspar content under the control of different microfacies.Finally, the porosity, permeability, quartz content, feldspar content, The content of kaolinite, the content of chlorite and the content of i / mulched layer, the prediction models of hypersensitivity, water sensitivity, acid sensitivity and alkali sensitivity were established respectively by Elman neural network method.The results showed that neural network The reservoir sensitivity index predicted by the method agrees well with the experimental results. The five-star tectonic belt has strong fast sensitivity, acid sensitivity, alkali sensitivity and salt sensitivity. The central and northwestern parts of the Luxiang fault depression have strong water sensitivity.