基于可变临界孔隙度模型的致密砂岩储层参数地震反演方法  

Seismic inversion method for tight sandstone reservoir properties based on a variable critical porosity model

作　　者：巴晶[1] 方志坚 符力耘 郭强 BA Jing;FANG ZhiJian;FU LiYun;GUO Qiang(School of Earth Sciences and Engineering,Hohai University,Nanjing211100,China;School of Geosciences,China University of Petroleum(East China),Qingdao Shandong266580,China;College of Information Engineering,China Jiliang University,Hangzhou310018,China)

机构地区：[1]河海大学地球科学与工程学院,南京211100 [2]中国石油大学(华东)地球科学与技术学院,山东青岛266580 [3]中国计量大学信息工程学院,杭州310018

出　　处：《地球物理学报》2023年第6期2576-2591,共16页Chinese Journal of Geophysics

基　　金：国家自然科学基金项目(42174161,41974123);江苏省杰出青年基金项目(BK20200021)资助。

摘　　要：储层参数地震反演是通过地震观测数据定量评估储层物性及含油气性的重要途径.中国西部致密砂岩油气分布范围广泛,具有巨大的勘探开发潜力,但其孔隙结构复杂,储层参数的定量预测难度较大.本文提出了一种基于可变临界孔隙度模型的储层参数地震反演方法.为有效描述储层的复杂孔隙结构,本方法将临界孔隙度作为变量,使其具有随深度变化的特征,以表征复杂孔隙结构,进而提高致密砂岩储层岩石物理建模的精度.此外,为直接从观测地震数据反演储层参数,结合岩石物理模型和地震反射系数方程建立正演算子;基于贝叶斯框架构建反演问题,采用混合高斯模型构建储层参数先验约束,用以描述储层的岩性(相)差异特征;基于临界孔隙度、泥质含量与孔隙度的统计关系,结合快速模拟退火算法,在迭代优化中使临界孔隙度与储层参数交替更新,用以提高反演结果稳定性.选用中国西部四川盆地某工区实际资料,对本方法进行应用测试与验证,结果表明本方法对于含气储层具有良好的指示性,相比常规固定临界孔隙度方法,本方法可以有效提升反演结果准确性.Petrophysical seismic inversion is an important approach to quantitatively evaluate reservoir properties from seismic observation data.Tight sandstone oil/gas reservoirs are widely distributed in West China,which exhibit a great potential for the exploration and production.The prediction of tight sandstone reservoir properties is quite difficult due to the complex pore structures.This work presents a novel seismic inversion method for reservoir properties based on the variable critical porosity model.To effectively describe the complex pore structures in tight sandstone reservoirs,the critical porosity is considered as a variable,which varies spatially to characterize the complex pore structures,by which the accuracy of rock physics modeling of tight sandstones is improved.In addition,for the purpose of directly extracting reservoir properties from the observed seismic data,the forward operator is established by combining rock physics model and seismic reflectivity equation.The relevant inverse problem is proposed based on a Bayesian framework and the Gaussian mixture model is adopted to formulate the prior constraint on the reservoir properties to characterize the lithological/facies variations.According to the statistical relations between critical porosity,clay volume,and total porosity,the critical porosity is alternatively updated along with the reservoir properties during the fast simulated annealing optimization process,to improve the stability of the inversion results.The developed method is tested in a working area of the Sichuan Basin,west China,of which the results demonstrate that the method can be a good indicator for the gas-bearing zones.Compared with the method with the conventional fixed critical porosity model,the new method effectively improves the precision of inversion results.

关 键 词：致密砂岩 复杂孔隙储层 储层参数地震反演 临界孔隙度 混合高斯模型 非线性优化 贝叶斯框架 

分 类 号：P631[天文地球—地质矿产勘探]