基于细菌觅食算法的砂砾岩岩性识别方法

砂砾岩储层岩性复杂多变,母岩成分变化大,孔隙结构复杂,难以精确划分岩性并建立准确的解释模型,导致储层参数计算精度不高。针对松辽盆地梨树断陷砂砾岩储层特点,选择多组分体积模型作为该地区的测井解释模型,将该地层看成由局部均匀的孔隙、泥质、石英、长石、岩屑等5部分组成?根据多组分体积模型建立相应的测井响应方程,引入细菌觅食算法作多组分矿物模型的优化算法,将优化结果与岩芯分析孔隙度及全岩矿物分析的体积分数进行对比,结果验证了细菌觅食算法反演砂砾岩储层多组分矿物模型的可靠性。采用该方法对松辽盆地砂砾岩储层测井资料进行处理,取得了较好的结果。

Identification Method of Sandy-conglomerate Lithology Based on Bacterial Foraging Algorithm

Sandy-conglomerate reservoir lithology is complex, composition variation of parent rock is large, pore structure is complex and strong heterogeneity, so that it is difficult to accurately divide lithology and build accurate interpretation model, resulting in low reservoir parameter calculation accuracy. Based on the characteristics of sandy-conglomerate reservoir in Lishu fault of Songliao Basin, a multi-component volume model was established for well logging interpretation, and the stratum was taken as the combination of local homogeneous pore, muddy, quartz, feldspar and rock debris. According to the multi-component volume model, the corresponding log response equation was built, and the bacterial foraging algorithm was taken as the optimal solution of multi-component mineral model, and then the optimized results were compared with the porosity by core analysis and the volume fraction by whole-rock mineral analysis. The results verify that the bacteria foraging algorithm is reliable for the inversion of sandy-conglomerate multi-component mineral model. Based on bacteria foraging algorithm, the result is good for the well logging interpretation of sandy-conglomerate reservoir in Songliao Basin.