Assessing the impact of relative permeability and capillary heterogeneity on Darcy flow modelling of CO₂ storage in Utsira Formation

Predicting CO2 plume migration is an important aspect for the geological sequestration of CO2. In the absence of experimental data, the storage performance of CO2 geo‐storage can be assessed through the dynamic modelling of the fluid flow and transport properties of the rock‐fluid system using empirical formulations. Using the van Genuchten empirical model, this study documents a Darcy flow modelling approach to investigate different aspects of CO2 drainage in a sandstone formation with interbedded argillaceous (i.e. mudstone) units. The numerical simulation is based on the Sleipner gas field storage unit where several thin argillite layers occur within the sandstone of the Utsira Formation. With respect to forward modelling simulations that have used Sleipner Formation as a case study, it is noted that previous attempts to numerically calibrate the CO2 plume migration to time‐lapse seismic dataset using software governed by Darcy flow physics achieved poor results. In this study, CO2‐brine buoyant displacement pattern is simulated using the ECLIPSE ‘black oil’ simulator within a two‐dimensional axisymmetric geometry and a three‐dimensional Cartesian coordinate system. This investigation focussed on two key parameters affecting CO2 migration mobility, namely relative permeability and capillary forces. Examination of these parameters indicate that for the gravity current of CO2 transiting through a heterogeneous siliciclastic formation, the local capillary forces in geologic units, such as mudstone and sandstones, and the relative permeability to the invading fluid control the mass of CO2 that breaches and percolates through each unit, respectively. In numerical analysis, these processes influence the evaluation of structural and residual trapping mechanisms. Consequently, the inclusion of heterogeneities in capillary pressure and relative permeability functions, where and when applicable, advances a Darcy modelling approach to history matching and forecasting of reservoir performance. Results indicate that there is a scope for a revision of the basic premise for modelling flow properties in the interbedded mudstones and the top sand wedge at the Sleipner Field when using Darcy flow simulators.