CCS Site Optimization by Applying a Multi-objective Evolutionary Algorithm to Semi-Analytical Leakage Models

Carbon capture and storage (CCS) has been proposed as a method of reducing global carbon dioxide (CO2) emissions. Although CCS has the potential to greatly retard greenhouse gas loading to the atmosphere while cleaner, more sustainable energy solutions are developed, there is a possibility that sequestered CO2 may leak and intrude into and adversely affect groundwater resources. It has been reported [1] that, while CO2 intrusion typically does not directly threaten underground drinking water resources, it may cause secondary effects, such as the mobilization of hazardous inorganic constituents present in aquifer minerals and changes in pH values. These risks must be fully understood and minimized before CCS project implementation. Combined management of project resources and leakage risk is crucial for the implementation of CCS. In this work, we present a method of: (a) minimizing the total CCS cost, the summation of major project costs with the cost associated with CO2 leakage; and (b) maximizing the mass of injected CO2, for a given proposed sequestration site. Optimization decision variables include the number of CO2 injection wells, injection rates, and injection well locations. The capital and operational costs of injection wells are directly related to injection well depth, location, injection flow rate, and injection duration. The cost of leakage is directly related to the mass of CO2 leaked through weak areas, such as abandoned oil wells, in the cap rock layers overlying the injected formation. Additional constraints on fluid overpressure caused by CO2 injection are imposed to maintain predefined effective stress levels that prevent cap rock fracturing. Here, both mass leakage and fluid overpressure are estimated using two semi-analytical models based upon work by [2,3]. A multi-objective evolutionary algorithm coupled with these semi-analytical leakage flow models is used to determine Pareto-optimal trade-off sets giving minimum total cost vs. maximum mass of CO2 sequestered. This heuristic optimization method is chosen because of its robustness in optimizing large-scale, highly non-linear problems. Trade-off curves are developed for multiple fictional sites with the intent of clarifying how variations in domain characteristics (aquifer thickness, aquifer and weak cap rock permeability, the number of weak cap rock areas, and the number of aquifer-cap rock layers) affect Pareto-optimal fronts. Computational benefits of using semi-analytical leakage models are explored and discussed. [1] Birkholzer, J. (2008) "Research Project on CO2 Geological Storage and Groundwater Resources: Water Quality Effects Caused by CO2 Intrusion into Shallow Groundwater" Berkeley (CA): Lawrence Berkeley National Laboratory (US); 2008 Oct. 473 p. Report No.: 510-486-7134. [2] Celia, M.A. and Nordbotten, J.M. (2011) "Field-scale application of a semi-analytical model for estimation of CO2 and brine leakage along old wells" International Journal of Greenhouse Gas Control, 5 (2011), 257-269. [3] Nordbotten, J.M. and Celia, M.A. (2009) "Model for CO2 leakage including multiple geological layers and multiple leaky wells" Environ. Sci. Technol., 43, 743-749.