API B I L D M Second Edition April ail. 19S-1
STATISTICAL ANALYSIS OF CRUDE OIL RECOVERY AND RECOVERY EFFICIENCY
U.S. PATENT OFFICE
AMERICAN PETROLEUM INSTITUTE Washington, D.C.
Issued by AMERICAN PETROLEUM INSTITUTE Production Department 211 North Ervay, Suite 1700 Dallas TX 75201 IIUMUMMUIUIIUMMIIIIIIUIIUUIUIII uiiiiimituu^miiiiHiiiimiiiinHtiiiiiiitiiiiiiiiimiiiiiiiiiiiiituitiim
Copyright « 1984 by American 1'etroieiim Institute Users of th s publication should become familiar with its scope, content, and inherent limitations. This publication is intended to supplement rather than replace individual engineering judgment.
American lYtruleurn Institute
T A B L E 1-2 SECONDARY RECOVERY E F F I C I E N C I E S Secondary Recovery Method Pattern Waterflood
Primary Plus Secondary Recovery Efficiency at Average OOIP*, percent
Ratio of Secondary to Primary Recovery Efficiency at Average OOII'*, percent
California Louisiana Oklahoma Texas Wyoming
35 51 28 38 45
0.33 0.4(1 (U12 0.50 0.89
Edfre Water Injection
Gas Cap Injection
California 44 Texas 43 *OOIP = Origina oil-in-place, refer to Appendix D, Nomenclature.
The Subcomm ttee. with completion of this study, concludes that a reliable statistical correlation cannot be achieved for the prediction of recovery and/or efficiency for individual reservoirs based on the readily definable and available reservoir parameters. An important factor, reservoir heterogeneity, cannot be readily defined, certainly not by the assignment of a mere numerical value. It is believed that not being able to include heterogeneity and possibly other factors as independent parameters contributes to the inability to develop valid statistical correlations. The accuracy ( f definable parameters such as porosity and initial water saturation is limited by the quality of measu*ement techniques available at the time of discovery and development. Such measurements have been subject to constant appraisal and improvement, but there is no simple way to correct older measurements. The Subcommittee did not attempt such corrections nor did they request respondents to do so. The impact of older, less accurate measurement techniques on such a study as this can be appreciated when it is recognized that approximately 75% of anticipated ultimate oil production from presently known U.S. fields will be derived from fields discovered prior to 1951. The recoverable oil to be expected from a given reservoir can be estimated with reasonable accuracy only after a thorough engineering study of that reservoir is undertaken. Such estimates of recoverable oil are improved by Hatching the performance of a reservoir model to the reservoir's production history and then using the matched model to predict ultimate recovery. Although statistical correlations of recovery efficiency based on readily definable reservoir and fluid parameters fall snort of being able to predict accurately the recovery of an individual reservoir, the calculated average recoveries in a single geological trend are significant. Comparison of these calculated averages shows s lbstantial recovery efficiency differences between reservoirs with different indigenous drive :nechanism:;, with the application of different secondary recovery techniques, and in different producing areas. The Subcommittee on Recocery Efficiency cautions against continued use of the correlations from AF'I Bulletin DlJ,: A Statistical Study <>f Recocery Efficiency. 'October 1!)C>7, to predict recocery or recocery (fficiency for any om resen-oir. Further, to avoid any undue significance being attached to the correlations developed in the current study, only those results required to substantiate the expressed conclusions are cited in this repor ..