苏里格气田不同沉积相建模方法及空间结构特征评价

摘要

With the Su-10 Block in the Sulige Gasfield as an example, target-based modeling techniques have been deployed, together with sequential indicator simulation method and multi-point geological statistics to construct 3 models for sedimentary facies with different features of spatial structures.Markov chain model has been used to characterize spatial structure features of various types.In addition, impacts of such models on restoration of spatial structure of data have been analyzed.Research results show the Markov chain model can satisfactorily represent spatial features of modeling results to facilitate assessment over modeling performances.Compared with the transcendental geologic phase maps, the multi-point geological statistics can effectively represent features of complicated sedimentary structures.The sequential indicator simulation can effectively characterize spatial structure of large-scale data;Compared with above 2 techniques, target-based modeling method can hardly characterize spatial structure with significant changes in amplitudes.Relevant research results may provide necessary theoretical supports for future development of the Su-10 Block.%以苏里格气田苏10区块为例,运用基于目标的建模方法、序贯指示模拟法和多点地质统计学方法建立3种不同空间结构特征的沉积相模型,利用马尔科夫链模型表征不同类型的空间结构特征,并评价模型对数据空间结构的恢复效果.结果表明,马尔科夫链模型能够很好地再现建模结果的空间特征,评价建模效果.对比先验地质相图,多点地质统计学能够再现复杂沉积体的特征;序贯指示模拟能够表征大尺度数据的空间结构;相比前2种方法,基于目标的建模方法不能很好地表征空间结构,变化幅度较大.该研究可为苏10区块后续开发提供一定的理论依据.