基于组合赋权法与云模型的致密砂岩储集层有利区评价方法研究:以松辽盆地南部乾深区块青三段x小层为例OA

To reduce the subjectivity in indicator weighting and the fuzziness of grading criteria in favorable area evaluation of tight sandstone reservoirs,we develop an integrated method that couples combined weighting with a normal cloud model,and apply it to the X subunit of the Member 3 of the Qingshankou Formation(Q3)in the Qianshen Block,southern Songliao Basin.First,subjective weights are derived from expert judgment using the Analytic Hierarchy Process(AHP).Objective weights are then obtained from the dataset using the CRITIC method.The two sets of weights are finally integrated to produce combined weights,improving both geological rationality and data-driven objectivity.Meanwhile,a normal cloud model is introduced to characterize uncertainty during grade assignment.Standard cloud parameters-expectation(Ex),entropy(En),and hyper-entropy(H,)—are determined from the score intervals of evaluation grades under constraints of a boundary membership threshold,enabling a flexible mapping between evaluation grades and measured data.Results show that sedimentary facies belt,oil-bearing property,and effective thickness are the dominant controls on favorable areas,with combined weights of 0.253,0.182,and 0.124,respectively.Evaluation of 20 wells identifies 12 wells for ClassⅠ,five wells for Class Ⅱ,and three wells for Class Ⅲ wells;the outcomes show 95％agreement with previous interpretations,with a Cohen's kappa coefficient(κ)of 0.91.The method therefore provides a practical and reliable tool for favorable area evaluation of tight sandstone reservoirs.