应用随机模拟技术建立夹层模型

APPLICATION OF STOCHASTIC SIMULATION TO INTERLAYER MODELING

崇仁杰 宋春华 程立芝

Chong Renjie Song Chunhua Cheng Lizhi

针对块状底水油藏的储层具有非均质性较为严重,夹层分布变化大,采用传统地质建模方法难以描述夹层大小及延伸长度,因而不能很好反映储层非均质性的实际情况,建立了先用随机模拟方法建立夹层预测模型,再利用此模型对储层属性模型进行校正而建立储层属性模型的“二步建模”方法,从而解决了传统建模方法中存在的问题,使得油藏模拟研究变得快捷和准确。该方法的实质是既利用了钻井资料的高分辨率;又充分利用了地震资料的高覆盖性。平湖油田H2油藏花港组储层为三角洲水下分流河道砂体,泥质夹层很发育。用传统的“一步建模法”,历史拟合需做 10次以上的参数调整,经 15天以上方能完成,效果还不理想;采用“二步建模”法,历史拟合仅需 5～6次参数修改,3～4天拟合就可完成。计算结果与实际生产结果吻合程度很高。

Pinghu oil gas field in East China Sea is a massive,bottom water drive light crude oil pool.The reservoir heterogeneity is relatively serious,and the distribution of muddy interlayers is a big variety.It is difficult to describe the size of muddy interlayers by traditional geologic model building method.It takes about 15 days and more than 10 times modifying the parameters of history fitting and reservoir simulation on H2 resevoir,because the geological model by traditional method can not describe clearly the reservoir heterogeneity,and the result of history fitting is not good.Use stochastic simulation to built prediction model,then,3-D reservoir attribute model is revised by the prediction model.It only takes 3～4 days and 5～6 times modifying the parameters of the history fitting and reservoir simulation.The result is very good and exact.

10.11743/ogg20020119