Rapid Large-scale Trapping of CO2 via Dissolution in US Natural CO2 Reservoirs

Naturally occurring CO₂ reservoirs across the USA are critical natural analogues of long-term CO₂ storage in the subsurface over geological timescales and provide valuable insights into the fate of CO₂ in the subsurface (Fig. 1). Previous measurements of CO₂ to ³He ratios within gas samples obtained from six natural CO₂ reservoirs, located in the United States, show that the CO₂ originated from magmatic degassing[1]. Variation in CO₂/³He across each reservoir suggests that significant amounts of CO₂, equivalent to hundreds of megatonnes, have been stored by solubility trapping[2]. However, the key question of whether CO₂ dissolution occurred during emplacement, or by diffusion and convection over geological time remains unanswered.

Here we present the results of integrating geochemical measurements with reservoir modelling to quantify both the mass of CO₂ emplaced and the proportion dissolved within each of the six natural CO₂ reservoirs. Given the magmatic origin of the CO₂, we use the known age dates of associated igneous rocks to estimate the timing of CO₂ emplacement in each reservoir. Using these emplacement ages, we show there is no relationship between the duration of CO₂ storage and the proportion of solubility trapping that has occurred. This shows that the proportion of dissolved CO₂ does not significantly increase over geological timescales. 

Further, we find that the original mass of CO₂ does not influence the proportion of CO₂ that is solubility trapped. We also find that rock properties, the present-day pressure, temperature and salinity of the reservoirs do not control the fraction of CO₂ dissolved, suggesting that the circumstances of CO₂ migration and filling are more critical. Our conclusion is supported by reservoir simulation at our exemplar site, Sheep Mountain in Colorado (Fig. 1), where we show that CO₂ dissolution after structural trapping is a minor contribution to amount of CO₂ residually trapped.

Our findings support a model where the majority of solubility trapping occurs on CO₂ injection and during the migration of the CO₂ plume. This indicates that the proportion of solubility trapping after the CO₂ has become structurally trapped is comparatively minor. Therefore, in engineered CO₂ stores, considerable amounts of injected CO₂ can be solubility trapped within CO₂ injection and post-injection monitoring timescales.

References

[1] Gilfillan et al., 2008, Geochimica et Cosmochimica Acta, 72 (4), 1174-1198
[2] Gilfillan et al., 2009, Nature, 458 (7238), 614-618

Fig. 1 (a) Location map of the west and central United States, showing the location of the natural CO₂ reservoirs investigated in this study. (b) Map of Four Corners area USA, showing the location of significant natural CO₂ reservoirs, topographic elevation and Cenozoic-age igneous rocks. The CO₂/³He ratio of samples from Bravo Dome, St Johns Dome, McElmo Dome, Sheep Mountain and McCallum Dome show a magmatic CO₂ source.