Quality control of porosity changes predicted by geochemical modelling in CCS case studies

Abstract

Abstract Assessment of long-term stability of injected CO 2 for geological storage requires a certain level of dynamic modelling, including geochemical modelling of fluid-rock interactions triggered by the injected CO 2 in geological formations. One of the key outcomes of a geochemical simulation is the predicted porosity changes. Decreasing porosity implies an overall plugging of the pore system and therefore probably a reduction in permeability, while increasing porosity implies the possibility of the opening up of flow paths. Therefore the predicted porosity change is an important indicator for self-healing versus self-accelerating behaviour of potential CO 2 leak paths. Based on project studies and natural analogue comparisons, several easy-to-use metrics and cross-correlations are developed in this paper. This is done by starting from a mass balance analysis, and re-writing it as a volume balance analysis plus correction terms. In this way porosity changes can be rigorously split up in distinct contributions from salinity changes, density changes and (de)mineralisation of CO 2 and H2O. By cross-correlating the porosity changes to these contributions, using 11 CCS case studies and natural analogues examples (covering a wide range of settings such as aquifer/gas field, geographic location, geochemical reaction paths, salinity, temperature), distinct patterns and rules of thumb emerge. The methodology and resulting cross-correlations can be applied as a quick and powerful sanity check on geochemical modelling results. The key conclusions that can be drawn from this work are: • The amount of mineralised CO 2 is a reasonable predictor for the porosity change • The amount of mineralised H2O and the density change can be significant contributors to the porosity change, but (without taking into account mineralised CO 2 ) are poor predictors for the porosity change The salinity change is only a minor contributor to the total porosity change. The main conclusion is that the porosity is likely to be reduced if CO 2 is mineralised. This should be considered a rule of thumb. Exceptions are possible (in the dataset used in this paper, 1 out of 11 cases, and the exception was caused by large density differences between dissolved and precipitated minerals). For new geochemical/Reactive Transport Modelling studies it is advisable to check how the calculated porosity change fits on the general trends presented in this study (especially the correlation to mineralised CO 2 ). It is highly advisable to scrutinise exceptions and to understand their root causes. The root cause might be a physical process (like for the 1 exception out of 11 cases we studied), but it could also point to an error in the geochemical model setup or in the model post-processing.