A contribution to risk analysis for leakage through abandoned wells in geological CO2 storage

The selection and the subsequent design of a subsurface CO2 storage system are subject to considerable uncertainty. It is therefore important to assess the potential risks for health, safety and environment. This study contributes to the development of methods for quantitative risk assessment of CO2 leakage from subsurface reservoirs. The amounts of leaking CO2 are estimated by evaluating the extent of CO2 plumes after numerically simulating a large number of reservoir realizations with a radially symmetric, homogeneous model. To conduct the computationally very expensive simulations, the ‘CO2 Community Grid’ was used, which allows the execution of many parallel simulations simultaneously. The individual realizations are set up by randomly choosing reservoir properties from statistical distributions. The statistical characteristics of these distributions have been calculated from a large reservoir database, holding data from over 1200 reservoirs. An analytical risk equation is given, allowing the calculation of average risk due to multiple leaky wells with varying distance in the surrounding of the injection well. The reservoir parameters most affecting risk are identified. Using these results, the placement of an injection well can be optimized with respect to risk and uncertainty of leakage. The risk and uncertainty assessment can be used to determine whether a site, compared to others, should be considered for further investigations or rejected for CO2 storage.