Alternative Modelling and Simulation of the Greenshank CO2-Geo-Sequestration Prospect, Gippsland Basin, Victoria

ALTERNATIVE MODELLING AND SIMULATION OF THE GREENSHANK CO2-GEO-SEQUESTRATION PROSPECT, GIPPSLAND BASIN, VICTORIA
Hossein Agheshlui, Caroline Milliotte (NFR Studies GmbH, Austria), Stephan Matthai

Previous CO2 -geo-sequestration simulations of the stratigraphic/aquifer trap referred to as Greenshank, indicate multi-pathing of the plume. Injection of 5Mtpa over 25 years (125Mt = desired storage capacity) creates a plume that reaches the boundary of the static model after 500 years. There are considerable uncertainties in these forecasts, and the use of regular (structured) corner-point grids (CPG) has imposed serious limitations on the physical realism of the representation and the accuracy of geologic structures in reservoir simulation models. The faults in the Greenshank area have complex shapes. Arguably these cannot be captured realistically by CPGs, but have a major impact on fluid flow.

Unstable displacement of the CO2 that manifests itself in viscous- and heterogeneity-induced fingering is another uncertainty factor. The CO2 displacement patterns are also strongly affected by buoyancy leading to gravity override and other phenomena. Such rate-dependent behaviour is not captured by standard saturation functions that perform poorly for injection processes in the presence of heterogeneity.

This project applies a more flexible alternative model building, (unstructured) finite-element meshing, and simulation approach. A suite of highly realistic geo- and reservoir simulation models of CarbonNet’s Greenshank CO2 -geo-sequestration target are being built for a sensitivity analysis of the effects of depth-conversion uncertainty (as will be captured by a suite of different horizon position models), connectivity between highly permeable geobodies, and permeability anisotropy. The analysis will be conducted with the CSMP++ CO2-geo-sequestration simulator, establishing a thorough understanding of how petro-physical variations and coupling/uncoupling of the component physics at work are likely to give rise to a plume behaviour where even small changes in the early path can lead through a “butterfly effect” to quite large changes in outcomes at later times. Understanding the impact of the faults on the plume spreading and how rate dependence and hysteresis of the saturation functions affect plume sweep and residual trapping will form part of this analysis and will lend more credibility to the forecasts. The central question that will be addressed concerns the storage capacity of the site: can 125 Mt be stored? – When will the boundary or spill-points be reached? How does the maximum amount vary with injection strategy?