长裸眼非均质储层酸压主因素优选方法研究

Study on optimal method of main acid fracturing factors for long open- hole wells in heterogeneous carbonate reservoir

塔里木盆地奥陶系碳酸盐岩油藏产层厚度大,完井井段长,非均质性严重。分别运用均值和均方差方法优选酸压井段物性参数,运用最优BP神经网络结构编制酸压决策软件,对塔里木盆地酸压施工的20井次进行酸压前决策模拟。实例应用对比分析表明,均值方法符合率75%,均方差方法符合率90%。对于长裸眼非均质碳酸盐岩储层,采用均方差方法优选储层物性参数,可以较准确地描述不同深度的物性差异给酸压造成的影响,定量地指导酸压选井选层工作及增产措施的制定。

Tarim Ordovician carbonate reservoir is characterized by bigger thickness of producing pay,long completion in- terval and serious heterogeneity. The mean value and square error methods were applied to optimize the physical parameters of acid fracturing segment, respectively. Using the optimal BP neural network structure, an acid fracturing decision - making software was compiled, to simulate the decision before acid fracturing of 20 wells in Tarim Basin. By analyzing application ex- amples ,the mean value method and the mean square error method can reach a coincidence rate of 75% and 90% ,respective- ly. Therefore,for long open- hole wells in heterogeneous carbonate reservoir,the mean square error method is used to opti- mize reservoir parameters accurately to describe the impact of the property difference at different depth on acid fracturing, and guide effectively and quantitatively the work of well - layer selection and other simulation treatment designs.