Abstract
In this field-data-based simulation study, we apply novel rate-dependent, anisotropic saturation functions curve-fitted to match the flow behaviour of laminated sand- and siltstones from the CO2CRC's Otway International Test Centre (Australia), in well-spot simulations of plume spreading at the site. Scenario analysis is conducted by performing simulations on various stochastic model realisations under different conditions, investigating what controls plume migration. Results indicate that high-permeability streaks and intraformational baffles determine plume spreading in flat-lying stratigraphy. Displacement of CO2 is unstable and focussed by high-permeability streaks, leading to multi-layer plumes. In inclined stratigraphy, buoyancy forces control plume migration channelling it up dip. However, high-permeability streaks still influence plume shape. Comparisons for different injection rates show that heterogeneity-induced fingering dominates at high injection rates, whereas gravity override and vertical movement of the plume are prominent at low injection rates. Continuous build-up of fluid pressure in the storage horizon influences the properties of CO2 when no-flow (closed) lateral boundary conditions are applied. By contrast, a marked pressure drop occurs in the near-well region for hydrostatic (open) lateral boundary conditions. These also foster uneven heterogeneity-fingering of the plume. Since flow velocity in the highly heterogeneous Paaratte formation varies over many orders of magnitude, force balances range between the capillary and the viscous limit even at some distance to the injector. A comparison with simulations based on standard Brooks-Corey saturation functions shows that taking flow-rate dependence and anisotropy into account leads to significant differences in the shape of the evolving CO2 plume and the saturation within.
Users
Please
log in to take part in the discussion (add own reviews or comments).
