Title :
New method for predicating reservoir sensitivity rapidly
Author :
Sun, Yu-xue ; Xiao, Chang ; Zhao, Jing-yuan
Author_Institution :
Coll. of Pet. Eng., Northeast Pet. Univ., Daqing, China
Abstract :
In resent years, many methods have been developed to forecast reservoir sensitivity, and we have found that it´s an ideal, fast, new method to use single-correlation analysis and multiple regressions to predicate reservoir sensitivity. On the basis of conventional core analysis and sensitive mineral analysis, we extract information relevant to every sensitivity, and use the new method above mentioned to predict reservoir-sensitivity, the accuracy of forecasting results can reach 85%, basically meet the needs of reservoir sensitivity predicted. Compared with single methods, the method combined single-correlation analysis and multiple regression is obviously improved when predicate reservoir sensitivity, and it is simple, widely applicable and with explicit physical significance.
Keywords :
correlation methods; hydrocarbon reservoirs; minerals; regression analysis; core analysis; information extraction; multiple regressions; predicate reservoir sensitivity; reservoir sensitivity forecast; sensitive mineral analysis; single-correlation analysis; Correlation; Minerals; Permeability; Reservoirs; Sensitivity; Water; core analysis; multiple regression; sensitive mineral; sensitivity analysis; single-correlation analysis;
Conference_Titel :
Machine Learning and Cybernetics (ICMLC), 2011 International Conference on
Print_ISBN :
978-1-4577-0305-8
DOI :
10.1109/ICMLC.2011.6016859