应用渗流反问题理论计算非均质油气藏多参数分布

Calculation of multi-parameter distribution in heterogeneous reservoir based on the inverse problem theory

将地质统计学方法引入到油气藏参数识别中,以井点已知信息为条件,确定油藏参数初始分布的先验模型,建立了初始油气藏孔隙度、渗透率参数和压力拟合的目标函数。目标函数中考虑了动、静态信息。采用改进的Gauss-Newton法,利用后验协方差矩阵估算油藏参数,实现了一种有效反求非均质低渗透油藏参数的方法。计算结果表明,该算法收敛性好,确定的渗透率场比孔隙度场更准确。

The geo-statistical method was used to identify the parameters of oil and gas reservoir. A prior model for initial distribution of reservoir parameters was determined on the basis of information around a well. The dynamic and static information was integrated to formulate the objection functions including initial permeability, porosity and pressure matching. The modified Gauss-Newton algorithm was used to calculate the multi-parameter of reservoir using post-covariance matrix. A fine characterization method for conditioning to static and dynamic data in a heterogeneous low-permeability reservoir is proposed. Some field cases show that the algorithm is convergent. The permeability field is more accurate than porosity field in some applications.