An Intercomparison Study of Simulation Models for Geologic Sequestration of CO2

Mathematical models and numerical simulation tools will play an important role in evaluating the feasibility of CO2 storage in subsurface reservoirs, such as brine aquifers, producing or depleted oil and gas reservoirs, and coalbeds. We have proposed and initiated a code intercomparison study that aims to explore the capabilities of numerical simulators to accurately and reliably model the important physical and chemical processes that would be taking place in CO2 disposal systems. A first set of simulation problems has been specified with input from several other organizations. These are considered prototypical for different potential storage systems and have been posted on the Internet at http://esd.lbl.gov/GEOSEQ/. Issues being addressed include mixing of carbon dioxide and methane (for CO2 sequestration in natural gas fields with enhanced gas recovery), displacement of water by CO2 (for CO2 disposal in saline aquifers), potential loss of CO2 through leaky faults, hydro-mechanically coupled processes such as stress-induced caprock deformation and permeability enhancement, mineral alteration as a consequence of CO2 injection, and CO2 injection into an oil reservoir. The intercomparison study is open to interested technical groups worldwide who are invited to model and study the test problems using their own codes and funding. Simulation results will be collected through the Internet and will be compared, discussed and documented through a series of workshops. It is anticipated that through interactions with other interested groups additional problem sets will be developed, proceeding from simple to complex, and from hypothetical test cases to actual laboratory and field problems. This paper presents the approach used for the code intercomparison study and briefly summarizes specifications of the first problem set. INTRODUCTION Geologic sequestration of CO2 can be accomplished by separating CO2 from flue gases and subsequently injecting it into a variety of storage reservoirs, including brine aquifers, producing or depleted oil and gas reservoirs, and coalbeds. Mathematical models and numerical simulation tools will play an important role in evaluating the feasibility of CO2 storage in subsurface reservoirs, in designing and analyzing field tests, and in designing and operating geologic CO2 disposal systems. In order to establish credibility for numerical simulators as practical engineering tools, it is necessary to demonstrate that they can model accurately and reliably the important physical and chemical processes that are taking place in the system of interest. The purpose of the code intercomparison study outlined here is to evaluate key processes in CO2 geologic sequestration, and to contribute to the acceptance of numerical simulators as viable tools for modeling CO2 disposal. To initiate the study we propose a preliminary set of simulation problems that are intended to cover some of the important phenomena and mechanisms that would arise in geologic sequestration of CO2. We envision an interactive process through which different technical groups with interests and capabilities relevant to geologic disposal of CO2 will participate in defining, solving, refining, and augmenting simulation problems. Code intercomparison studies have been successfully used as a means for establishing confidence in simulation tools in related technical fields such as petroleum engineering (Firoozabadi and Thomas, 1989), geothermal reservoir engineering (Stanford, 1980), and in nuclear waste management (Larsson, 1992; Chapman et al., 1994; Jing et al., 1995; Stephansson et al., 1996). Depending on the storage reservoir of interest and the composition of the waste gas stream (pure CO2 vs. mixtures of CO2 with other gases), injection of CO2 in geologic formations may give rise to a number of physical and chemical phenomena, such as miscible or immiscible displacement of native fluids, dissolution of injected fluids into reservoir fluids, changes in effective stress with associated porosity and permeability change and the possibility of inducing seismic activity, chemical interactions between fluids and solids, and nonisothermal effects. Key issues arising in process simulation include (1) thermodynamics of suband supercritical CO2, and PVT properties of mixtures of CO2 with other fluids, including (saline) water, oil, and natural gas; (2) fluid mechanics of single and multi-phase flow when CO2 is injected into aquifers, oil reservoirs, and natural gas reservoirs; (3) coupled hydro-chemical effects due to interactions between CO2, reservoir fluids, and primary mineral assemblages; and (4) coupled hydromechanical effects, such as porosity and permeability change due to increased fluid pressures from CO2 injection. These issues can be tracked through a matrix of property and process issues for different CO2 storage reservoirs as shown in Table 1. Additional topics that need to be addressed include space and time discretization and their impacts on the solution of the underlying mathematical model, and the dependence of processes and parameters on space and time scale. The code intercomparison study should progress from relatively simple, uncoupled problems that address specific issues to increasingly complex problems in which several effects would occur simultaneously. Ultimately it would be desirable to achieve a comprehensive coverage of all process aspects and couplings. The coverage of issues achieved through the eight simulation problems proposed here is shown by entering the problem