CO2 storage well rate optimisation in the Forties sandstone of the Forties and Nelson reservoirs using evolutionary algorithms and upscaled geological models
Babaei, Masoud; Pan, Indranil; Korre, Anna; Shi, Ji-Quan; Govindan, Rajesh; Durucan, Sevket; Quinn, Martyn. 2016 CO2 storage well rate optimisation in the Forties sandstone of the Forties and Nelson reservoirs using evolutionary algorithms and upscaled geological models. International Journal of Greenhouse Gas Control, 50. 1-13. 10.1016/j.ijggc.2016.04.011
Before downloading, please read NORA policies.Preview |
Text (Open Access Paper)
1-s2.0-S1750583616301396-main.pdf - Published Version Available under License Creative Commons Attribution 4.0. Download (4MB) | Preview |
Abstract/Summary
Optimisation is particularly important in the case of CO2 storage in saline aquifers, where there are various operational objectives to be achieved. The storage operation design process must also take various uncertainties into account, which result in adding computational overheads to the optimisation calculations. To circumvent this problem upscaled models with which computations are orders of magnitude less time-consuming can be used. Nevertheless, a grid resolution, which does not compromise the accuracy, reliability and robustness of the optimisation in an upscaled model must be carefully determined. In this study, a 3D geological model based on the Forties and Nelson hydrocarbon fields and the adjacent saline aquifer, is built to examine the use of coarse grid resolutions to design an optimal CO2 storage solution. The optimisation problem is to find optimal allocation of total CO2 injection rate between existing wells. A simulation template of an area encompassing proximal-type reservoirs of the Forties-Montrose High is considered. The detailed geological model construction leads to computationally intensive simulations for CO2 storage design, so that upscaling is rendered unavoidable. Therefore, an optimal grid resolution that successfully trades accuracy against computational run-time is sought after through a thorough analysis of the optimisation results for different resolution grids. The analysis is based on a back-substitution of the optimisation solutions obtained from coarse-scale models into the fine-scale model, and comparison between these back-substitution models and direct use of fine-scale model to conduct optimisation.
Item Type: | Publication - Article |
---|---|
Digital Object Identifier (DOI): | 10.1016/j.ijggc.2016.04.011 |
ISSN: | 17505836 |
Date made live: | 09 Aug 2016 08:18 +0 (UTC) |
URI: | https://nora.nerc.ac.uk/id/eprint/514204 |
Actions (login required)
View Item |
Document Downloads
Downloads for past 30 days
Downloads per month over past year