| Summary: | 碩士 === 國立成功大學 === 資源工程學系 === 102 === Duong’s method of forecasting production and estimated ultimate recovery (EUR) in low permeability reservoirs with a long-term linear flow has been verified by several authors. However, Duong’s method overestimates future production during boundary-dominated flow. It is reasonable to combine Duong’s method and the Arps decline relations for fractured wells exhibited linear flow followed by boundary-dominated flow, because the Arps decline relation has better prediction for boundary-dominated flow. For wells that have not reached boundary-dominated flow, Wattenbarger et al.’s linear flow theory is frequently used to determine the duration of linear flow, and the end of linear flow time (telf) can be used to estimate when to switch from Duong’s method to the Arps decline relation.
This paper focuses on the availability of the end of linear flow time determined by linear flow theory with a hybrid forecasting method, which combines Duong’s method and the Arps exponential relation for tight sand/shale gas reservoirs. To obtain the proper end of linear flow time, modified Wattenbarger et al.’s analytical solution and modified empirical solutions (Duong’s method combines the Arps exponential relation) was derived. The golden section search method is used to find the modified end of linear flow time that minimizes the difference between modified analytical and empirical solutions.
Then, the hybrid forecasting method was used to forecast production of synthetic production data generated from an analytical solution. A number of cases were tested to verify the efficacy of this method for forecasting production. To account for tight sand/shale gas reservoirs, reservoir permeability ranging from 10-1 md to 10-4 md was considered. The results indicated that the hybrid forecasting method used with the modified end of linear flow time is theoretically accurate for production forecast and EUR estimation. In low-permeability reservoirs with a permeability ranging from 10-1 md to 10-3 md, this method provides a significant improvement in EUR estimation. Finally, application of this method to field examples from Barnett shale gas and Bossier sand gas were also presented.
|
|---|
