基于多点地质统计方法的岩相建模研究  

Stochastic Simulation of Lithofacies Distribution Based-on Multi-point Geostatistics

作　　者：潘少伟[1] 王家华[1] 杨少春[2] 秦国伟[3] 

机构地区：[1]西安石油大学计算机学院,西安710065 [2]中国石油大学(华东)地球资源与信息学院,青岛266555 [3]中国石油大庆油田有限责任公司勘探开发研究院,大庆163712

出　　处：《科学技术与工程》2012年第12期2805-2809,共5页Science Technology and Engineering

基　　金：国家自然科学基金项目(50874091);陕西省教育厅科研计划项目(11JK1071)资助

摘　　要：目前,多点地质统计方法的应用越来越广泛。以江苏油田庄2断块阜一段E1f12-1小层为例,利用多点地质统计方法来进行岩相建模研究。首先依据E1f12-1小层的河道发育形态及砂体分布特征,利用不同的方法建立了3种训练图像。然后在这3种训练图像控制下,模拟实现了E1f12-1小层的岩相模型。结合钻井、测井信息综合分析后发现:依据沉积微相图所建的第2种训练图像控制下产生的岩相模型与已有的地质认识较为吻合。利用多点地质统计方法建立岩相模型有助于从地质的角度对模型进行约束,促进概念模型向定量模型的转化,最终建立反映地下真实地质情况的储层模型。Multi-point geostatistics has become an important method of reservoir description currently.Taking the E1f12-1 layer of the fault-block Z2 in Jiangsu Oilfield as an example,the lithofacies model is built up based-on multi-point geostatistics.Firstly,three kinds of training images are built up by different methods according to the characteristics of sand body of E1f12-1 layer.Then the lithofacies models of E1f12-1 layer are realized under the control of the three kinds of training images.And it is found that the lithofacies model that is under the control of the second training image is the most consistent with the existent geological knowledge.The application effect of the lithofacies model of E1f12-1 layer shows that the multi-point geostatistics is helpful to restrict the model from the geological point of view,promote the transform from conceptmodel to the quantitative model and build a reasonable 3D model reflecting the true subsurface situation.

关 键 词：多点地质统计方法 江苏油田 训练图像 岩相模型 

分 类 号：TE132.14[石油与天然气工程—油气勘探]