Evaluation of the Uncertainty and the Risk of CO₂ Storage by the Stochastic Modeling With Aids of Seismic and Well Data

Multi-phase flow simulation is generally employed to evaluate CO₂ storage capacity in geological sequestration projects. However, due to a scarcity of data, heterogeneity of a reservoir or a seal is sometimes neglected and a homogeneous layer model is accepted in the simulation at the early stage of the projects. In general, the simulation results of the homogeneous case and the heterogeneous case are not comparable. Also, the output of the homogeneous model with the average properties does not give us the average result of the multiple heterogeneous models in non-linear systems. Therefore, we need to conduct the simulation with multiple heterogeneous models to assess the uncertainty. We populate multiple heterogeneous models for the reservoir and the seal with aids of seismic, well logs and geological prior information in a stochastic manner. One of the stochastic modeling, COSISIM (Sequential Indicator co-simulation), is employed in this study. In COSISIM, probabilities of sand and shale, which are derived from well logs and seismic, are used as soft data. We create 25 heterogeneous models for the inputs of the flow simulation. The results indicate the large variations or uncertainty in the CO₂ mass stored in the reservoir and the seal. Some models show a leakage of CO₂ from the top of the seal since we include the heterogeneity in the seal as well as in the reservoir. The leakage mass ratio can be considered as a risk profile that will be served as a crucial factor in the decision making. This kind of information is obtainable only by doing the simulation with multiple realizations; therefore the stochastic modeling approach is necessary for the CO₂ project to evaluate the risk associated with CO₂ storage.