In the research of oil and gas exploration and development,reservoir classifica-tion is an important link.Currently,with the continuous in-depth study of unconventional oil and gas reservoir development,the development technology of tight sandstone gas reservoirs is becoming increasingly mature.This type of gas reservoir is usually a low porosity and low permeability reservoir,so studying low porosity and low permeability reservoirs is crucial for understanding tight sandstone gas reservoirs.This article combines the characteristics of tight sandstone reservoirs in the Shihezi formation of the Qingshimao area in Ningxia,and selects porosity,permeability,and flow unit factors as reservoir characteristic parameters.In order to make the classified reservoirs more representative and accurate,before conducting reservoir classification,the selected characteristic parameter table is first used to identify anomalies in the data.Then,the K-means clustering analysis method is used to obtain the reservoir clas-sification standards for the Qingshimao area,and finally,the decision tree method is used to identify reservoir types.The conclusion drawn from this is in line with the characteristics of low porosity and low permeability of tight sandstone,and combined with its gas testing data,it can greatly reduce production costs and improve production capacity.
关键词
致密砂岩储层/异常识别/K-means均值算法/决策树
Key words
tight sandstone reservoir/abnormal identification/K-means clustering algorit-hm/decision tree
维普
万方数据