Prediction of sand production onset in petroleum reservoirs using a reliable classification approach  

作　　者：Farhad Gharagheizi Amir H.Mohammadi Milad Arabloo Amin Shokrollahi 

机构地区：[1]Department of Chemical Engineering,Texas Tech University,Lubbock,TX,United States [2]Thermodynamics Research Unit,School of Engineering,University of KwaZulu-Natal,Howard College Campus,King George V Avenue,Durban,4041,South Africa [3]Institut de Recherche en Genie Chimique et Petrolier(IRGCP),Paris Cedex,France [4]Departement de Genie des Mines,de la Metallurgie et des Materiaux,Faculte des Sciences et de Genie,Universite Laval,Qu
ebec(QC),G1V 0A6,Canada [5]Young Researchers and Elites Club,North Tehran Branch,Islamic Azad University,Tehran,Iran

出　　处：《Petroleum》2017年第2期280-285,共6页油气（英文）

摘　　要：Controlling sand production in the petroleum industry has been a long-standing problem for more than 70 years.To provide technical support for sand control strategy,it is necessary to predict the conditions at which sanding occurs.To this end,for the first time,least square support machine(LSSVM)classification approach,as a novel technique,is applied to identify the conditions under which sand production occurs.The model presented in this communication takes into account different parameters that may play a role in sanding.The performance of proposed LSSVM model is examined using field data reported in open literature.It is shown that the developed model can accurately predict the sand production in a real field.The results of this study indicates that implementation of LSSVM modeling can effectively help completion designers to make an on time sand control plan with least deterioration of production.

关 键 词：Sand production Least square SVM ROC graph Classification description Modeling Sanding onset 

分 类 号：F42[经济管理—产业经济]