Reservoir Simulation Studies for Planning Monitoring Schemes for CO2 Storage

In this paper, we describe an approach to designing monitoring schemes for carbon dioxide sequestration in saline aquifers. Changes in key parameters are investigated over timescales of up to a thousand years. The study addresses movement of the CO2 plume, possible locations for observation wells and the period for which a storage location should be monitored. For the initial sensitivity analysis, we use a simple homogeneous reservoir simulation model to understand how reservoir, operational and model parameters affect the amount of mobile CO2 remaining at different times over the storage period. The parameters with the greatest impact are taken forward to uncertainty studies, which are conducted on two reservoir models with more realistic geological characteristics: one with lateral extensive baffles and one with sand channels. For these cases, we investigate the movement of the CO2 plume and its arrival at possible locations for an observation well. Results from the sensitivity analysis indicate that the most influential parameters are horizontal permeability, dipping angle, critical gas saturation, salinity, the period of injection and the capillary pressure curve. The results from the uncertainty studies indicate that for the two heterogeneous models, a reasonable monitoring period is in the range of 60 to 150 years and that the movement of the plume probably stops after approximately 100 years. The arrival time of CO2 at the observation well can be predicted with greater confidence when the well is in close proximity to the injector and in the direction in which CO2 will preferably move. A correlation analysis on the uncertain parameters shows that the main contributor affecting the amount of mobile CO2 is critical gas saturation, followed by dipping angle and the period of injection. While previous studies focus on how different parameters affect immobilization of CO2, this study aims to develop a methodology to plan long-term monitoring of mobile CO2. Prediction of the expected plume movement can help to determine suitable observation well locations and reasonable timescales for the monitoring process.