Application of smart proxy model for multi-optimization of CO2-EOR design in farnsworth unit with production-injection well conversionopen access
- Authors
- Na, Yoonsu; Song, Youngsoo; Ampomah, William; Wang, Jihoon
- Issue Date
- Dec-2024
- Publisher
- Elsevier Ltd
- Keywords
- CO2-EOR; Machine learning; Multi-optimization; Smart proxy model
- Citation
- Energy Reports, v.12, pp 3196 - 3206
- Pages
- 11
- Indexed
- SCIE
SCOPUS
- Journal Title
- Energy Reports
- Volume
- 12
- Start Page
- 3196
- End Page
- 3206
- URI
- https://scholarworks.bwise.kr/hanyang/handle/2021.sw.hanyang/212004
- DOI
- 10.1016/j.egyr.2024.09.006
- ISSN
- 2352-4847
2352-4847
- Abstract
- CO2-Enhanced Oil Recovery (CO2-EOR) is an effective development strategy and an economically viable CO2 storage method for a mature oil field. In this study, a smart proxy model (SPM) was constructed to establish the optimal development strategies in Farnsworth Unit (FWU) considering CO2 storage and oil production with the production-injection well conversion. An SPM was constructed with the results of 16 reservoir simulations and was validated with an R2 values of 96.3 % and 97.5 % for the CO2 storage and oil production, respectively. The obtained Pareto front indicates a range of the optimized CO2 storage and oil production. In addition, the maximum CO2 storage case shows 260 % and 61.1 % of the CO2 storage and oil production of the base case, respectively. When the maximum oil production is targeted, the CO2 storage and oil production are respectively 66.5 % and 111.3 % of the base case. It was concluded that the SPM is effectively applicable for multi-objective optimization with fewer simulation runs and shorter time. In addition, the well conversion is a viable method to improve both the CO2 storage and oil production in the FWU during CO2-EOR process.
- Files in This Item
-
Go to Link
- Appears in
Collections - 서울 공과대학 > 서울 자원환경공학과 > 1. Journal Articles

Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.