低孔砂岩孔隙度计算方法及裂缝识别技术

在低孔砂岩储层中，测井计算出的孔隙度与岩心孔隙度相关系数仅为０．５，原因是孔隙度测井与岩心孔隙度的相关性差。计算出响应对孔隙度的关联度和权重，根据权重大选取输入曲线，用ＢＰ神经网络建立起计算机隙度的非线性模型，计算结果与岩心孔隙度的相关系数提高到０．７５。

Porosity Calculation and Fracture Identification for Low Porosity Sandstone

In low porosity sandstone reservoir, correlation coefficient of the calculated porosity and the core analysis porosity is only 0.5 because porosity log is badly correlated with the core analysis porosity. The correlativity and weight of log response to porosity are calculated. Based on the weight, the input curves are selected, and BP neural network is used to set up a nonlinear model for porosity calculation, which increases the correlation coefficient to 0.75. Fracture features reflected on the electric logging images are given and the conventional logging respone features of fractures are recounted. After the fracture samples and weight identified by conventional log data are obtained and the distance between samples calculated, the samples are optimized. Then the fractures are identified by distance judgement method. The agreement rate of fracture intervals from conventional logging with that from imaging logging reaches 75%. Fracture density obtained from FMI data is converted into fracture development. And the development of fracture computed from conventional log data agrees well with that from FMI data.