多因素水淹强度指数在剩余油评价中的应用

Multi-factor Water-out Index Application In Residual Oil Evaluation

针对常规测井解释方法确定水淹层剩余油饱和度时参数难以确定、测井资料缺失等问题,选用合理的水淹评价因素,利用基于椭圆基函数的动态模糊神经网络方法对水淹强度指数进行多因素预测,得到由多因素水淹强度指数划分的水淹级别和剩余油平面分布规律。实例证明,水淹级别预测正确率为100%,剩余油饱和度计算相对误差在10.7%以内,具有现场应用意义。

It is difficulty to gain the residual oil saturation in water- out intervals by conventional well- logging interpretations due to the shortage of well- logging data. Appropriate water- out evaluation factors are determined,and the dynamic neutral network based on ellipse primary function is used to perform multi- factor prediction of water- out index. This multi- factor water- out index can be used to divide water- out level and identify plane distribution of residual oil. Practical application demonstrates that the prediction accuracy of water- out level can reach 100%,and the relative error of residual oil calculation is less than 10. 7%,which is favorable for field application.