Reservoir assisted history matching method with a covariance localization EnKF using fast marching method.zip

As per the static parameter field information of reservoir geology modeling and the functional equation, FMM is employed to realize the rapid tracking of the pressure wave transmission duration of each well, determine the sensitivity region of an individual well, and build the localized matrix. In combination with the covariance localized EnKF approach, the gradient modification of data assimilation approach is achieved, and the false correlation is decreased. Through progressively matching and renewing the reservoir parametric model, the optimal model is finally obtained.

基于油藏地质建模的静态参数场信息与函数方程，采用快速行进法（Fast Marching Method，FMM）实现各井压力波传播时长的快速追踪，确定单井敏感区域并构建局域化矩阵。结合协方差局域化集合卡尔曼滤波（Ensemble Kalman Filter，EnKF）方法，实现数据同化方法的梯度修正，降低虚假相关。通过逐步匹配与更新油藏参数模型，最终得到最优模型。