|
Carbon dioxide sequestration in deep aquifers and depleted oil-fields is one of the technical solutions proposed to reduce greenhouse gas release in the atmosphere. The gas containment relies on several trapping mechanisms and a low permeability cap-rock to prevent CO2 from leaking upwards. During the several phases of the sequestration, the host-rock, the cap-rock and the borehole seals will be in contact with a supercritical CO2 phase, containing various amounts of dissolved water. However, the reactivity of the rock minerals in CO2 fluids is thoroughly unknown.
Preliminary tests have demonstrated the effective reactivity of several pure mineral phases, including in water-free experiments (regnault et al. (2005)). In particular, the results have shown that the portlandite, chosen as an analogue of the cement phases, was entirely recrystallized into calcite in a short amount of time.
A series of experiments has been devised to observe and quantify the reactivity of portlandite with supercritical CO2. Experiments have been carried out in a batch reactor under different conditions: pressure 160 bar, temperatures 80, 120, and 200° C, presence or absence of liquid water. Initial reaction rates are proposed, based on two independent techniques: measure of the advancement of reaction by X-ray diffractometry, and monitoring of CO2 injected in the batch to compensate for pressure drop during the experiment.
SEM observations show that in presence of liquid water, the carbonation leads to the complete dissolution of portlandite and precipitation of large and well crystallized rhomboidal calcite. On the other hand, in absence of a liquid phase, carbonation leads to the passivation of the surfaces by precipitation of a coating of amorphous calcite, and an incomplete reaction of the portlandite. This discrepancy of behaviours has been linked to the dielectric constant of the reacting fluid, which permits solvatation and export of ions only in presence of liquid water.
A specific geochemical model has been developed in order to account for the particular conditions of our experiments. Indeed, the very small amount of water in the chemical system requires to solve the chemical eaquations using the constituent masses instead of their concentration. Also the portlandite reactive surface evolution (especially the surface passivation in presence of vapor phase) has been accounted for.
The results of the portlandite carbonation in presence of supercritical CO2, and the water-poor geochemical system modelling techniques should be useful to simulate wellbore cement reactivity experiments, and to understand the durability of these materials in the context of CO2 sequestration.
|
