State of the art review of CO2 Storage Site Selection and Characterisation Methods

 
1. INTRODUCTION
 
2. GEOLOGICAL CHARACTERISATION OF THE SITE
 
3. FLOW MODELLING
 
4 REACTIVE FLOW MODELLING
 
5. COUPLED GEOMECHANICAL AND FLOW MODELLING
 
6. ENVIRONMENTAL IMPACT AND RISK ASSESSMENT
 
7. ECONOMIC ANALYSIS
 
8. PUBLIC PERCEPTION AND ACCEPTANCE
 
9. CONCLUSIONS AND RECOMMENDATIONS
 
 
3.1 Flow and transport mechanisms
 
3.2 Overview of numerical models used in flow modelling for CO2 stora...
 
3.3 Flow Modelling in CO2 Injection Operations
 
3.4 Conclusions
 
 
3.3.1 K12-B Gas Field
 
3.3.2 Weyburn and Midale Oil Fields
 
3.3.3 The Frio Brine Pilot Experiments
 
3.3.4 Nagaoka Pilot Site
 
» 3.3.5 Sleipner CO2 Storage Project
 
3.3.6 In Salah CO2 Storage Project
 
 

3.3.5 Sleipner CO2 Storage Project

The Sleipner CO2 storage project is the world's largest and longest running Carbon Capture and Storage (CCS) project. The Sleipner fields are situated in the Norwegian North Sea. They produce gas with a high CO2 content from Jurassic and Tertiary reservoirs. The CO2 is separated from the hydrocarbons at the Sleipner T platform and, since 1996, 13 million tonnes of CO2 have been injected into the Utsira Formation of Miocene age. This formation consists of up to 300 m thick sandstones with 90 - 98 % sand content, with average porosity of 35 - 40 %, net/gross ratio of 0.90 - 0.97, and permeabilities in the 1-8 Darcies range. The subsurface CO2 plume has been monitored from the surface by six time-lapse seismic surveys. Chadwick et al., 2009 presented seismic images of the CO2 plume at Sleipner showing its development up to 2006 (Fig. 3-4). The first repeat seismic survey (1999) revealed that migrating CO2 had spread to nine distinct layers - one of these lying above a 5 - 6.5 m thick shale. The migrating CO2 cumulating under the low-permeability layers appears to have been fed from a central vertical feeder, which is referred as a seismic chimney in the seismic data. (Alnes et al., 2011; Hermanrud et al., 2009).

E. Fig . 3-4Fig. 3-4: Seismic images of the Sleipner plume showing its development to 2006. Top) N-S seismic section through the plume. Topmost CO2 layer arrowed. Bottom) plan views of the plume showing total integrated reflection amplitude (Chadwick et al., 2009).

Instead of a complex numerical approach, Bickle et al., 2007 studied CO2 migration and accumulation at the Sleipner storage site in the North Sea using modified well-known solutions for (axisymmetric) gravity flows within a permeable medium. The results indicate that CO2 accumulation under the shallower low-permeability layers within the reservoir started sometime after the start of injection of CO2 in the field. Modelling of the time variation of thickness in two of these layers indicated that their CO2 input increased with time. Conversely, under the deeper layers net CO2 inputs appear to decrease with time. It seems most probable that the deeper layers progressively leaked more CO2 through their thin caprock mudstones with time and the growth of the overlying layers reflects this increasing supply of CO2 from below.

The reactive transport modelling study conducted by BRGM (Picot-Colbeaux et. al., 2009) using TOUGHREACT showed that after three years of injection, the primary trapping process is the geological structure of the system. The upward migration of the free CO2 due to buoyancy forces is prevented by the low-permeability shale layers but mainly by the caprock formation overlying the reservoir aquifer. The dissolution mechanism is also modelled and it was seen that although the solubilisation had started, it did not represent more than 10% of the injected CO2 amount (Audigane et al., 2011).

One of the studies to understand plume evolution and storage performance is performed by Chadwick and Noy, 2010. The need for the adjustments on the capillary entry pressure and the permeability of the intra-reservoir mudstone in a 2D THOUGH2 flow model (Fig. 3-5) in order to match the observed arrival time of CO2 at top of the reservoir, showed that CO2 migration through the mudstones is actually not by Darcy flow but via some form of pathway flow, possibly associated with networks of small faults or perhaps holes. According to the 3D THOUGH2 flow model, the increased CO2 fluxes to the topmost layer suggest that the feeder pathways are evolving, becoming either more transmissive with time and/or increasing in number. The history match of the lateral spreading rates could be achieved by using very high anisotropic permeabilities as can be seen in Fig. 3-6.

E. Fig . 3-5Fig. 3-5: Simulated growth of the CO2 plume from 1999 to 2006 from the THOUGH2 axisymmetrical flow model (Chadwick and Noy, 2010).

Migration of CO2 through permeable pathways, formed by overlapping turbidite lobes or channels, in otherwise sealing cap-rocks has been modelled by Grimstad et al., 2009. The simulation results are then fitted using an analytical expression. It was concluded that the longer it takes the CO2 to migrate to the surface, the larger the lateral spread of the CO2 plume would be.

E. Fig . 3-6

Fig. 3-6: The topmost layer in 2006. a) observed extents; b) THOUGH2 simulation with k=6 Darcy; c) THOUGH2 simulation with k=3 Darcy east-west and 10 Darcy north-south; d) THOUGH2 simulation with k=3 Darcy east-west, 10 Darcy north-south and higher reservoir temperature (Chadwick and Noy, 2010).