基于马蒙算法地质统计学反演的缝洞储集体预测

The prediction research on fracture-cavity reservoirs by geostatistical inversion based on Markov Chain and Monte-Carlo algorithm in the Tahe oilfeild

塔河油田缝洞储集体空间非均质性强,基于地震资料的确定性预测方法表征缝洞储集体时存在多解性。采用一种基于马蒙算法的地质统计学反演方法实现了缝洞储集体的定量预测并作出不确定性评价。首先从地质、地球物理角度获取目标区储集体发育的先验信息,再进行地质统计学参数的实验调试,最终利用地质统计学反演获得储集体岩相类型结果及波阻抗结果的多个实现,并对预测结果作出不确定性评价。该方法提供了一种地震数据主导的缝洞储集体定量预测手段,能够较好地解决溶洞储集体纵向深度归位、地震预测成果不确定评价问题,对于该类油藏的地质建模表征具有指导意义。

The spatial heterogeneity of carbonate fractured-vuggy reservoir is very strong in the Tahe oilfeild.Interwell reservoir prediction mainly depends on the application of seismic data.However,the results of deterministic seismic interpretation have multi-solutions.A geostatistical inversion method based on Markov Chain and Monte-Carlo algorithm is adopted for fractured-vuggy reservoir quantitative prediction and uncertainty evaluation.Firstly,geological prior information of the target zone can be obtained by geological regularity and deterministic interpretation results from well logging and seismic survey.Then,controlled by suitable geostatistical simulation parameters of the study area,the geostatistical inversion can be done,yielding many equal probability results of reservoir type volume and impedance data volume.Based on these results,the uncertainty of inversion can be analyzed.The method provides an effective seismic-driven way for quantitative prediction of fractured-vuggy reservoir,giving better solutions to the problems of the cave vertical depth positioning and the uncertainty evaluation of seismic prediction results. It is of great guiding significance for the modeling and characterization of fractured-vuggy reservoir.