基于电成像测井图像的井壁压裂缝视宽度计算

摘要

Accurately predict the fractures width around the wellbore is a key to protecting the reservoir and control leakage . Aiming at the limitation of the existing method , this paper calculates the visual width of the fractures from the point mode and the segment mode .In point mode ,the visual width of the fracture around the current point is extracted by using the mouse . In the segment mode ,the average visual width of the section of fractures is calculated by selecting a section of imaging logging image .The histogram of the fracture image does not show the two peaks ,so the two‐the maximum variance method is presented to automatically extract the fracture in the segregates mode .The maximum variance method is used by twice to determine the optimal multi‐class classification thresholds ,w hich segregates the fracture region and the other regions . The automatic fracture results are accurately extracted by combining with human‐computer interaction to achieve the best segmentation effect .Measured data processing results show that the two modes can better extract the fractures ,and calculate the fracture width ,which has good practicability .%准确计算井壁附近压裂缝宽度是控制泥浆漏失和保护储层的关键。裂缝宽度的计算方法存在各种局限性。依据电成像测井图像，利用点模式与段模式2种方式分别计算压裂缝的视宽度。在点模式方式中，提取当前点压裂缝视宽度。在段模式方式中，通过选择一段压裂缝计算其平均视宽度。针对实测压裂缝电成像测井图像直方图不存在双峰情况，在段模式方式下自动提取压裂缝目标时，提出了采用双类间最大方差法，确定最佳分类阈值，对压裂缝的区域与其他区域进行分割；辅助人机交互方式对自动提取的压裂缝目标进行修正，达到最佳分割效果。实测资料处理结果表明，2种方式均能够较好地提取压裂缝，计算压裂缝的视宽度，点模式的计算速度更快，段模式的计算精度更好。