Carbon sequestration in saline aquifers connected to EOR fields: Modelling implications

Geological carbon sequestration (GS) adjacent to CO2-EOR projects (i.e. up or down-dip) is attractive due the economic advantages and availability of the necessary infrastructure. CO2 injection into an aquifer connected to an active EOR site provides the flexibility to deal with variable flow rates provided by capture units. Another advantage for operating companies is that down dip injection can help to avoid pressure losses to the aquifer, which helps achieve oil- CO2 miscibility pressure faster. A complication for simulating adjacent GS is that CO2-EOR activities will have significant boundary effects for nearby GS simulations. Another challenge is that most of the time due to confidentiality or lack of data, modeling of oil production in EOR sites is not possible. What is needed is a simple approach for handling EOR activities that have significant impact on GS simulations. The method we have used to simplify the neighboring (up-dip) multi-phase region for applying reasonable boundary conditions for down-dip GS involves replacing oil production with cumulative liquid produced (hydrocarbons + brine) and modifying production rates and the permeability field only in the EOR region. An obvious advantage of removing the oil zone and replacing it with brine is reduced simulation time due to reduced number of components and flow complexity which will allow computation time to be focused on simulating the GS domain without considering the oil phase and production specifics up dip. In this study we derive the required modification based on the pressure propagation speed in different fluids. The first step consists in modifying the permeability field to ensure the same pressure propagation observations in reservoir. The second step adjusts the production rates in EOR site to match the actual bottom hole pressure at producers. Justification of our analytical derivations with numerical simulations from CMG-IMEX and CMG-GEM shows at distances of our interest, estimated pressure values have 2% or less error, and suggest the method has broad applicability for other GS projects with associated but distinct EOR projects.