页岩储层标准等温吸附曲线的建立——以鄂尔多斯盆地长7页岩储层为例

Establishment of standard adsorption isotherms for shale reservoirs:a case of Chang 7 shale reservoir in Ordos Basin

页岩的矿物成分复杂、孔隙结构多样,导致页岩吸附性能差异大,同一储层中不同岩样的等温吸附曲线形态变化大,有必要针对页岩建立标准等温吸附曲线。对来自鄂尔多斯盆地长7油层组的页岩储层样品进行了低温液氮等温吸附实验,将实验所得曲线进行初步标准化后按照BET理论中的常数C值取2为底的指数进行分类并标准化。再将所有样品的标准等温吸附曲线使用4种吸附层厚度方程在Matlab数值分析软件中进行参数拟合,用以验证标准等温吸附曲线。按拟合精度从小到大排列的4种吸附层厚度方程的顺序为:Carbon Black 2次方程<Carbon Black 3次方程<Halsey方程<De Bore方程。最后通过参数平均得出了每一类型的页岩标准等温吸附曲线。所得页岩储层的标准等温吸附曲线代表了一系列表面性质相似的储层的吸附特征,可直接用于油藏数值模拟或油藏工程计算,简化了大量的繁琐计算过程,降低了岩样随机选择所导致的数据不确定性,且使计算结果具有较强的代表性。

The composition of shale is complicated and the pore structure is diverse, which results in great differ- ence in adsorption properties of shale. The adsorption isotherm curves of different rock samples in the same reser- voir vary greatly. Thus, we need to establish a standard adsorption isotherm curves for shale. The adsorption iso- thermal experiments of shale samples from Chang 7 reservoir in Ordos Basin were carried out by low tempera- ture liquid nitrogen method. The experimental curves were initially standardized and then classified according to the C value of the BET theory, taking the exponent of 2 as the standardized value. Then, the standard adsorption isothermal curve of all samples was used to fit the parameters of the four adsorption layer thickness equations in Matlab numerical analysis software to verify the adsorption isothermal curve. The fitting accuracy of the four adsorp- tion layer thickness equations is： Carbon Black, square 〈 Carbon Black, cube 〈 Halsey 〈 De Bore. Finally, the standard adsorption isothermal curves of each type of shale were obtained by parameter averaging. The standard of adsorption isothermal curve of shale reservoir, which stand a series of adsorption characteristics of reservoirs with similar surface properties, can be calculated directly in reservoir simulation and reservoir engineering use, reduce the data uncertainly caused by random selection of rock samples, simplify a lot of calculation process, and make the calculation result is representative.