Quantification of solubility trapping in natural CO₂ reservoirs using gas geochemistry and reservoir modelling

Carbon capture and storage (CCS) is an important technology for mitigating industrial CO₂ emissions and is considered essential by many organisations for achieving net-zero targets. Secure long-term geological storage of CO₂ is critical for CCS to be successful. In this thesis, I use naturally occurring geological CO₂ reservoirs as analogues for the long-term behaviour of engineered CO₂ storage reservoirs. Solubility trapping is a key CO₂ trapping mechanism in CO₂ storage reservoirs. Solubility trapping involves the dissolution of gaseous or supercritical CO₂ into the formation water. The dissolution of CO₂ into formation water increases storage security and capacity by reducing the mass and volume of mobile CO₂. In natural CO₂ reservoirs, the masses of CO₂ that have dissolved, and the rate of dissolution over time are uncertain. Prior to the research conducted in this thesis, dissolved CO₂ mass and dissolution rates had only been calculated within one natural CO₂ reservoir, Bravo Dome, in New Mexico USA. In Chapter 3 of this thesis, I present a meta-analysis of published numerical modelling, natural analogue and pilot site studies that quantified solubility trapping at a full-reservoir scale. The studies showed that solubility trapping was the principal form of chemical trapping in saline aquifer CO₂ storage and varied between 10–50%. Analogue and pilot site studies did not show the high fractions of solubility trapping (>60%) observed in some previously published modelling studies. In Chapter 4 using gas geochemistry and reservoir modelling techniques, I quantified the mass and the fraction of CO₂ dissolved at five additional natural CO₂ reservoirs in the USA. These reservoirs were; Sheep Mountain (Colorado), McCallum Dome (Colorado), McElmo Dome (Colorado), St Johns Dome (Arizona and New Mexico) and Jackson Dome (Mississippi). Interpreting CO₂/³He and δ¹³C(CO₂) data from reservoir CO₂ samples, I showed that 17–56% of CO₂ in naturally occurring reservoirs was removed by solubility trapping. These solubility trapping fractions equated to hundreds of millions tonnes (Mt) of CO₂ trapping. There was no relationship between the mass of CO₂ emplaced and the fraction dissolved in the natural CO₂ reservoirs, nor did the geological the properties of the reservoirs appear to influence solubility trapping. Using the known age of associated magmatic rocks, I constrained the timing of CO₂ emplacement in each reservoir. There was no relationship between the duration of CO₂ storage and the fraction of solubility trapping. This result suggests that the reservoirs are in dissolution equilibrium over geological timescales and that most dissolution occurs during the initial emplacement and migration of the CO₂ plumes. At my primary study site of Sheep Mountain in Colorado, the proximity of CO₂ samples to the gas water contact (GWC) did not influence the fraction of CO₂ dissolved. This result also supports a model of rapid and early dissolution, with minimal dissolution occurring over geological timescales across a stable GWC. In Chapter 5, I present the first ever numerical flow simulation of the filling of a natural CO₂ reservoir. I used the geochemically determined solubility trapping fraction of 29% (based on CO₂/³He and δ¹³C(CO₂)), as the ‘ground truth’; against which, I compared different CO₂ filling scenarios. These scenarios simulated the emplacement of magmatic-origin CO2 into the reservoir by injecting a CO₂ mass of 131 million tonnes into the primary reservoir unit at Sheep Mountain. The simulation results showed that most solubility trapping occurred during injection and migration, with limited dissolution after structural trapping of the CO₂ plume. CO2 plume migration distance was found to have the greatest influence on solubility trapping. A scenario of downdip CO₂ injection, with a 5.1 km plume migration distance, resulted in 17% of CO₂ dissolving. An updip CO₂ injection scenario, with a plume migration distance of 1.2 km, resulted in only 11% of CO₂ dissolving. The simulation results also showed that different CO₂ filling rates did not influence long-term solubility trapping, as the CO₂ plumes followed the same migration pathway and contacted a similar volume of water. Across all simulated scenarios, solubility trapping varied between 11–19%, significantly below the 29% expected from geochemical analysis. We attributed the high fraction of dissolved CO₂ at Sheep Mountain to either meteoric aquifer recharge or longer CO₂ plume migration distances of ~10 km. My results show that natural CO₂ reservoirs can securely retain hundreds of millions of tonnes of CO₂ over geological timescales. Solubility trapping is an important trapping mechanism that can remove significant fractions of originally emplaced CO₂. The results strongly support a model of most dissolution occurring during the migration of the CO₂ plumes. The results of this thesis are promising for engineered CO₂ storage in sedimentary reservoirs. I show that solubility trapping has the potential to significantly enhance storage security and capacity over a timescale of years to decades. I show that CO₂ migration distance is a key parameter for enhancing solubility trapping and suggest that engineered CO₂ storage sites could be designed to enhance solubility trapping via extended CO₂ migration distances.