程立华, 吴胜和, 贾爱林, 杨海长, 宋春刚. 综合多信息进行地震约束储层随机建模——以准噶尔盆地庄1井区J1s22砂组为例[J]. 海洋地质与第四纪地质, 2008, 28(3): 127-131.
引用本文: 程立华, 吴胜和, 贾爱林, 杨海长, 宋春刚. 综合多信息进行地震约束储层随机建模——以准噶尔盆地庄1井区J1s22砂组为例[J]. 海洋地质与第四纪地质, 2008, 28(3): 127-131.
CHENG Li-hua, WU Sheng-he, JIA Ai-lin, YANG Hai-zhang, SONG Chun-gang. A NEW METHOD OF RESERVOIR STOCHASTIC MODELING CONSTRAINED BY SEISMIC DATA: A CASE STUDY OF J1S22 FORMATION OF ZH1 BLOCK IN ZHUNGGAR BASIN[J]. Marine Geology & Quaternary Geology, 2008, 28(3): 127-131.
Citation: CHENG Li-hua, WU Sheng-he, JIA Ai-lin, YANG Hai-zhang, SONG Chun-gang. A NEW METHOD OF RESERVOIR STOCHASTIC MODELING CONSTRAINED BY SEISMIC DATA: A CASE STUDY OF J1S22 FORMATION OF ZH1 BLOCK IN ZHUNGGAR BASIN[J]. Marine Geology & Quaternary Geology, 2008, 28(3): 127-131.

综合多信息进行地震约束储层随机建模——以准噶尔盆地庄1井区J1s22砂组为例

A NEW METHOD OF RESERVOIR STOCHASTIC MODELING CONSTRAINED BY SEISMIC DATA: A CASE STUDY OF J1S22 FORMATION OF ZH1 BLOCK IN ZHUNGGAR BASIN

  • 摘要: 针对地震约束储层随机建模中地震反演波阻抗数据体与岩相概率关系差的问题,提出综合多信息、逐级约束、多级建模的思路方法。首先以地震反演数据体约束建立地球物理特征曲线随机参数场,再以每一个地球物理特征曲线参数场为约束建立相模型,最后以优选的相模型和地球物理特征曲线参数场为约束进行储层参数建模。该方法在井距较大的开发前期评价阶段具有较好的应用效果。

     

    Abstract: As a case study of ZHI block in Zhunggar basin, a new method of multistep constraint stochastic modeling was presented, which integrates together geology, well logging and seismic data. The multistep modeling method can properly solve the problem that simulation target value and restriction value have no perfect relations, for example, facies and impedance. The method of multistep modeling has three steps. Firstly, the random field of typical geophysics curve, constrained by impedance data, is built by Collocated Cokriging simulation. Secondly, based on the geophysics curve field realizations the facies models are built. Finally the reservoir parameters model is built using Sequential Gauss simulation, constrained by the reasonable facies model which is picked through facies simulation realization analysis. This method has good performance in the phase of pre-development.

     

/

返回文章
返回