Geological controls on the performance of CO2 injection and storage sites
Abstract
Using experience gained from several pioneering industrial-scale CO2 storage projects we identify the key factors controlling injection performance, both for the injection period (c. 10 years) and for the longer term (>10 years). The Sleipner, In Salah and Snøhvit storage sites, which together have succeeded in disposing of over 16 Mt of CO2, present highly variable reservoirs, covering a range of reservoir depth, permeability and geological architecture. Reservoir modelling studies show how CO2 plume development has been controlled by geology, geo-mechanics and fluid phase behaviour. Three general themes about CO2 in the subsurface emerge from these cases studies: a) Geology: Variations in geological architecture result in highly non-uniform plume development. Short-term and near-well injection performance is viscous dominated, while longer-term and far-field performance becomes gravity/capillary dominated. The case studies illustrate how thin-shale layers, top-reservoir topography, faults and fractures control actual CO2 plume development. Understanding how these controls affect field performance is best achieved by careful tuning of reservoir models to observations from reservoir monitoring. b) Rock mechanics: Geo-mechanical properties affect storage capacity, limit well injectivity and determine pressure development within the reservoir. Simple linear elastic models are generally insufficient, whereas by using detailed mechanical models we have been able to fine-tune the reservoir and overburden mechanical models to learn how the rock system responds to fluid pressure. c) Phase behaviour: While pure CO2 phase behaviour is well understood, the behaviour of gas mixtures, within the inherently uncertain ranges of in situ reservoir pressure and temperature, presents a significant challenge, leading to large uncertainties in gas density, viscosity, and solubility. Detailed models of near-wellbore multi-phase flow show how injectivity and plume growth around the well-bore are affected by the geological and fluid phase assumptions. Can all these factors be successfully merged into a single coupled modelling package - probably not? We prefer to use detailed models focussed on individual fluid-rock responses and processes. These include models of gas-phase thermodynamics in the well-bore and near well-bore volume, models of rock mechanical (elastic) response with detailed mechanical layering and failure zones, and models of gravity-dominated CO2-plume dynamics.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2011
- Bibcode:
- 2011AGUFM.H24B..03R
- Keywords:
-
- 1847 HYDROLOGY / Modeling;
- 1857 HYDROLOGY / Reservoirs;
- 5114 PHYSICAL PROPERTIES OF ROCKS / Permeability and porosity;
- 8020 STRUCTURAL GEOLOGY / Mechanics;
- theory;
- and modeling