Methods for Estimating CO2 Storage in Saline Reservoirs

Methods for estimating subsurface volumes in porous and permeable geologic formations are routinely applied in oil and gas, ground water, underground natural gas storage, and UIC disposal evaluations. In general, these methods can be divided into two categories: static and dynamic. The static models require only rock and fluid properties, while the dynamic methods require information about active injection, e. g. injection volumes and reservoir pressures. The static models are volumetric and compressibility; the dynamic models are decline curve analyses, mass (or volumetric) balance, and reservoir simulation. The volumetric method requires a simple geometric description of the formation that includes formation height and area, porosity, and some type of factor that reflects the pore volume that CO(2) can occupy. The compressibility method relates the change in pore pressure to the change in water and pore volume. Pore pressure will increase as CO2 is injected. This causes a decrease in water volume and an increase in pore volume. The sum of these changes can accommodate space for CO2 storage. This is primarily an issue for closed reservoirs. For active CO(2) injection, in addition to the static methods, dynamic methods can be used. Some geologic formations will incur flat or constant injection rates with variable injection pressure, while other formations will have decreasing injection rate with constant injection pressure. This is primarily due to surface operations and the outer boundary condition of the reservoir (closed or open). A decreasing injection rate trend can be analyzed and extrapolated to an uneconomic injection rate to infer the ultimate storage capacity of an active injection well. Mass or volumetric balance can be used to estimate subsurface injected volumes using cumulative injection volume as a function of pore pressure. Based on the geologic unit, combinations of parameters are used to allow straight line projections of the observed trends to make forecasts of the ultimate storage capacity. Injection zone flow simulation models can be used pre- or post injection, require the most input data and can accommodate a very rigorously defined geologic model. Storage estimates can be made readily from simulation for various development scenarios that include completion interval, well-type, injection rates, and multiple injection well spacing. The data requirements, assumptions and limitations for each of these methods are reviewed. Advantages and disadvantages of applying each method to basin-scale and specific site-scale storage estimates are discussed. (C) 2008 Elsevier Ltd. All rights reserved