层次分析法在致密油有利区优选中的应用——以川中侏罗系大安寨段为例

Application of AHP Method to Favorable Area Optimization for Tight Oil: A Case Study in Daanzhai Formation,Jurassic,Central of the Sichuan Basin

层次分析法常用于解决复杂系统分析与决策中目标优选、排序等问题,也适用于油气勘探领域。在前人致密油地质特征研究的基础上,总结了致密油有利区优选指标。使用层次分析法优选有利区的核心内容是建立层次结构模型和构建2类矩阵。以四川盆地中部侏罗系大安寨段为例,介绍了使用层次分析法优选有利区5个主要步骤:1根据研究区地质特征,确定目标及方案,构建目标层和方案层;2根据影响致密油富集的要素,确定准则层;3构建准则层中各要素的成对比较矩阵,确定各准则要素的相对权重;4构建各准则要素对不同方案的判断矩阵,确定不同方案中各准则要素的权重;5将不同方案各准则要素的权重矩阵与准则要素的相对权重矩阵相乘,得到各方案的层次排序,进而选出最佳方案。优选的川中侏罗系大安寨段致密油最有利勘探区为金华—莲池区。操作简单、不需考虑下限值、适用范围广是层次分析法优选有利区的优点,但由于层次分析中的标度根据人为经验赋值,不同评价人员给出的评价结果有所不同。将层次分析法与GIS、特尔菲法、模糊数学法相结合,可以提高准则要素的优选合理性及赋值准确性。

The Analytic Hierarchy Process（ AHP） is a decision-making method which permits the relative assessment and prioritization of alternatives under complex contexts. This method is also applied to petroleum exploration. Based on the previous geological characteristics study,we outlined the favorable area optimization criteria of tight oil. The hierarchical model and two matrices are key contents of favorable area optimization in AHP. A case study in Daanzhai Formation,Jurassic,central Sichuan Basin was presented to show the five main steps of AHP optimization. First of all,a goal and alternatives should be determined according to the geological characteristics of the studied area. The second step is to select the favorable area optimization criteria of tight oil. According to qualitative judgment of relative importance of these criteria,then a pairwise comparison matrix can be constructed. The last but one step is to construct judgment matrices of each alternative to determine the weight of each alternative in different criteria. Finally,the multiplication of eigenvectors of the pairwise comparison matrix and judgment matrices will show the rank of each alternatives. In the Daanzhai tight oil case,the favorable exploration target is Jinhua-Lianchi Block. The advantages of AHP optimization are simple operation steps,unrestricted limit values and wide application range. However,the appraisal result varies by evaluator for the 9-point scale is typical subjective or linguistic equivalents. AHP can be combined with GIS,Delphi process or fuzzy method to increase the rationality of optimization and the accuracy of scale calibration.