Soft Computing Algorithms Accelerate and Improve the History-Matching Process: Elk Hills, California-29R Reservoir

摘要

This paper presents the application of soft computingrn(virtual intelligence) techniques1 to a reservoir simulationrnhistory matching problem. The objective of this work wasrnnot to automate the history matching process as hasrnbeen discussed by others2-7, but rather to provide thernengineers with the necessary information to improve thernspeed and quality of their results.rnThrough the use of virtual intelligence techniques historyrnmatch error (mismatch – the difference betweenrncalculated and observed flow and/or pressure) isrncorrelated to variations in individual history matchrnparameters such as porosity and permeability. Anrnobjective function that describes the criteria to minimizernthe mismatch is defined.rnThis technology always finds the global minimumrnassociated with the objective function. The technologyrnprovides multiple solutions that satisfy any error criteria,rnand produces related statistical information. Therntechnology can handle continuous or discrete historyrnmatch parameters.rnIn this paper we discuss the successful application ofrnthis technology to a simulation study of a complex,rnfractured, porcelanite oil reservoir (29R, Elk Hills field,rnCalifornia). This field has 28 years of history with 42rnproduction wells. A dual porosity formulation wasrnnecessary to properly model the fractured nature of thernreservoir. Successful history match results obtained in arnshort period of time for this field showed the accuracyrnand practicality of this unique history matchingrntechnology.