含泥质夹层油藏蒸汽注采过程的自适应网格算法研究

根据泥质夹层的低渗特性及空间分布,本文提出了一种含泥质夹层油藏网格渗透率的粗化计算方法,并在此基础上,将自适应网格算法应用于含泥质夹层油藏的数值模拟,提升其计算效率.在计算过程中,网格的动态划分仅依据流体物理量的变化,泥质夹层区域不全部采用细网格,仅针对流动锋面处的泥质夹层采用细网格,其余泥质夹层处采用不同程度的粗网格.相较于传统算法,网格数大幅下降.数值算例表明,自适应网格算法的计算结果精度与全精细网格一致,能够准确模拟出泥质夹层对于流体的阻碍作用,同时计算效率得到大幅提升,约为全精细网格算法的3~7倍.

Adaptive Mesh Refinement Method for Steam-Injected Oil Recovery Process in the Reservoir with Muddy Intercalation

Considering the low permeability and spatial distribution of muddy intercalation, a coarsening algorithm for grid permeability is proposed in this article. Based on this, the adaptive mesh refinement (AMR) algorithm is applied to the numerical simulation for the reservoir with muddy intercalations. In the AMR method, refining the dynamic grids or not only depends on the local gradient magnitude of the relevant physical quantities. It means that though fine grids are used, but not for all the muddy intercalation area. Only at the muddy intercalation area where the flow front passes over, the fine grids will be used locally. At the other muddy intercalation area, different levels of coarse grids are retained. In this way, the total number of grids is greatly reduced. Numerical examples show that the calculation results of the modified AMR method are consistent with the full fine grid method. The hindering effect of muddy intercalation on the flow can be presented accurately, and at the same time, the computational efficiency is improved by approximately 3~7 times compared to the traditional full fine grid method.