关键词: 非线性贝叶斯反演算法, 地表沉降, 油田储层参数, 单源Okada模型, 双Okada模型

Abstract: It is of great significance for real-time monitoring of reservoir safety to use surface subsidence information to quickly acquire reservoir status information based on geophysical inversion algorithm.At present, there are few studies on reservoir parameter inversion at home and abroad. Therefore, this paper takes Panjin area of Liaohe which is a typical area of surface subsidence caused by oil exploitation as the research area,InSAR subsidence monitoring results as observation data, Okada model as inversion model, and introduces the nonlinear Bayes inversion algorithm into the inversion of reservoir parametersin oilfield for the first time.During the experiment, more than one reservoir was found in the oilfield, so the double Okada model was introduced into the inversion of reservoir parameters in oilfield for the first time.Experiments show that:(1) The nonlinear Bayes inversion algorithm can not only obtain the optimal inversion parameters, but also explain the uncertainty of inversion results. (2) Compared with the single-source Okada model, the double Okada model is more in line with the mapping relationship between reservoir parameters change and surface subsidence, and the reservoir parameters obtained from double Okada model are more reliable.

Key words: the nonlinear Bayes inversion algorithm, surface subsidence, oilfield reservoir parameter, single-source Okada model, double Okada model