5.3.5 Reactive transport

The intense flushing of the reservoir rock around CO2 injectors with large quantities of dried super critical CO2 can cause desiccation of the remaining brine in the pore space, leading to substantial precipitation of salts and sulphate minerals, blocking the pores and diminished injectivity. On the other hand, carbonate dissolution by the acidified brine could cause porosity and permeability to increase.

Both precipitation and dissolution can cause geomechanical effects, given that large pressure variations can occur close to the injector. Although Thermal-Hydraulic-Chemical (THC) codes present many advantages in forecasting injection flow rates or chemical processes, they only consider interactions between minerals and aqueous phases. THC codes do not integrate the mechanical deformation involved in CO2 injection. For the time being, fully coupled thermal-hydraulic-chemical-mechanical codes are still in the development stage. The first simulations, with external coupling between the reactive transport model and geomechanical model, give encouraging results. Johnson et al. (2005) simulated long-term caprock integrity as a function of geochemical and geomechanical contributions to permeability evolution using the reactive transport simulator NUFT and distinct-element geomechanical model LDEC (Gaus et al., 2008).

Li et al., 2006 built a model that uses a sequential coupling approach to investigate the thermo-hydro-mechanical behaviour of CO2 injection around a fault environment. The effects of temperature, initial geological stress, injection pressure and CO2 buoyancy on the mechanical behaviour of the fault were studied. The injection pressure has a larger influence on the relative slip change of the fault than the buoyancy induced by the CO2 plume. Although at the initial stage of the injection the pore pressure of the storage formations is affected by the injection pressure, as time passes, the CO2 plume-induced buoyancy plays a key role, influencing the pore pressure of the storage system.

Heffer et al., 2007 suggested that statistical modelling using the principal component analysis of well rate fluctuations can be used to identify the faults that are mechanically active during project development. Coupled geomechanical-flow models were used to demonstrate the observed correlations between rate and fault-related characteristics.

Chang and Bryant, 2009 studied the effects of declined and inclined faults on the behaviour of CO2 plumes in 2D and 3D Formations. Several fault properties (conductive vs. sealing, angle relative to dip, distance from initial plume location) were examined to understand the dynamics of CO2 behaviour such as residual phase trapping and direction of the plume. They stated that a large amount of CO2 leaks into the fault below the top seal. However, the fault also creates a virtual source for up-dip migration into the permeable bed. This attenuates the leakage and results in significant additional residual saturation trapping, as can be seen in Fig. 5-7.

E. Fig . 5-7

Fig. 5-7: Inclined and conductive fault's effect on CO2 plume behavior (Chang and Bryant, 2009).