中坝气田须二段气藏裂缝分级评价方法  

Fracture grading and evaluating methods for Reservoir Xu-2 in Zhongba Gas Field

作　　者：程超[1] 何亮 张本健[2] 曹建[2] 张亮 朱雷 CHENG Chao;HE Liang;ZHANG Benjian;CAO Jian;ZHANG Liang;ZHU Lei(School of Geoscience and Technology,Southwest Petroleum University,Chengdu 610500,China;Northwest Sichuan Gas Mine,PetroChina Southwest Oil and Gas Field Company,Jiangyou 621700,China;Southwest Branch of CNPC Logging Co Ltd,Chongqing 400021,China)

机构地区：[1]西南石油大学地球科学与技术学院,四川成都610500 [2]中国石油西南油气田分公司川西北气矿,四川江油621700 [3]中国石油集团测井有限公司西南分公司,重庆400021

出　　处：《大庆石油地质与开发》2020年第2期56-64,共9页Petroleum Geology & Oilfield Development in Daqing

基　　金：中国石油天然气股份有限公司科研项目“中坝气田须二气藏精细描述”(2014B-0608)。

摘　　要：中坝气田须二段气藏为裂缝—孔隙型致密砂岩老气藏,经历了50多年的勘探开发,截止到目前累计产油气量已接近当初的探明储量,但生产动态资料证实气藏仍有潜力可挖。裂缝作为其主要的流通通道,一方面对储层的储集性能和油气藏的产能贡献巨大,但同时裂缝也是水淹或水驱的重要通道。因此对该气藏裂缝发育级别的重新评价对于开发方案的调整及剩余储量挖潜具有十分重要的意义。充分利用表征裂缝发育的测井信息、岩石力学参数及构造曲率等多维数据集,运用AdaBoost算法模型对裂缝进行分级评价实验。结果表明:该算法与KNN算法、LR算法、支持向量机和RF等传统算法相比精度更高;基于该分类结果的气藏裂缝发育级别平面分布与气藏生产动态资料吻合。研究成果为老气藏裂缝发育级别的分级评价提供了一种新思路。Reservoir Xu-2 in Zhongba Gas Field is the old fracture-porous tight sandstone gas reservoir,which has experienced more than 50 years of exploration and development.Up to now,the accumulative productions of the oil and gas are close to the original proved reserves,but the dynamic production data prove that the gas reservoir still has potentials to be tapped.As the main flow channel,the fractures contribute greatly to the reservoir performances and hydrocarbon reservoir productivity,but they are also important channels for the watering out or water flooding.Therefore,the re-evaluation of the fracture developed levels in this gas reservoir is of great significance to the adjustment of the developing plan and the potential tapping of the remained reserves.With the help of the full use of the multidimensional data set including the logging information,mechanics parameters of the rocks,constructional curvature and so on that can represent the development of the fractures,and moreover by means of the model by AdaBoost algorithm,the grading and evaluating experiments on the fractures were conducted.The results show that this algorithm is more accurate than the traditional algorithms such as KNN algorithm,LR algorithm,support vector machine,RF algorithm and so forth;and furthermore the planar distribution of the fracture development levels based on the graded results is consistent with that of the dynamic production data of the gas reservoirs.The research results provide a new idea for the grading and evaluation of the fracture development levels in the old gas reservoir.

关 键 词：致密砂岩老气藏 裂缝—孔隙型 多维数据集 ADABOOST算法 裂缝分级评价 

分 类 号：TE343[石油与天然气工程—油气田开发工程]