用测井资料计算碳酸盐岩渗透率

碳酸盐岩渗透率大小取决于孔隙大小、喉道宽度和裂缝宽度等因素。根据孔隙和裂缝的发育情况及其在储集和渗滤中所起的作用 ,将碳酸盐岩储层划分为孔隙型、裂缝 -孔隙型和裂缝型 3种储集类型。从岩心分析、成像测井资料中划分出含 3种储集类型的储层段。从常规测井资料中提取出指示孔隙和裂缝的参数形成识别样本 ,对样本做非线性映射处理 ,得到不同类型储层在 X- Y空间上的分布。用 BP网络建立识别储集类型模型 ,识别结果与成像测井识别结果和岩心分析结果吻合。用灰色静态模型计算裂缝密度 ,对孔隙度做泥质、有机质校正 ,计算喉道大小及弯曲程度 ,合成孔隙结构参数。用非参数回归方法按储层类型建立计算渗透率的数学模型。经实际资料处理 ,计算渗透率的精度有了明显提高。

Calculation of Carbonate Formation Permeability with Log Data

Carbonate formation permeability depends on pore size, throat and fracture width, etc.. In view of pore and fracture abundance along with their functions in reservoir and percolation, carbonate reservoir may be divided into three types: porous, fractured porous and fractured. The reservoir intervals can be picked up by core analysis and FMI data. Recognition samples from conventional log data indicate the 2 D distribution of the reservoir by non linear mapping processing. The BP neural network based reservoir model provides the same result as that of FMI and core analysis. Fracture density is determined from the grey static model; Pore structures parameters are derived from calculating the throat sizes and pore bending deflection and correcting porosity for clay and organic matter. Math model of permeability is set up by non parameter regression in accordance with the reservoir types, with which calculation accuracy of the permeability has been apparently improved.