CO <sub>2</sub>-Brine Co-Injection in Carbonate Aquifers Using Geothermal Doublets

Abstract

In the aim to reduce GHG emissions, CO2 storage in deep and large saline aquifers has been recognized as a technology being able to abate drastically CO2 emissions and participate to the ‘two-degree scenario (2DS)” strategy. Current challenges are focused on increasing CO2 storage capacity from Mega-Tones to Giga-Tones in order to have a sensible impact on the reduction of CO2 atmospheric concentration and global warming. Numerous large-scale facilities will have to be deployed in coming years through regional or state hubs to capture, transport and store a CO2 coming from various industrial feeds: natural gas processing, fertilizer production, steel manufactures, cement plants, hydrogen production, wood or biomass boilers, BECCS, etc. When these CCS hubs are far from small to medium CO2 industrial emitters (+/- 100 kt/year), local CCS projects could also be considered through small size CO2 geological storages: depleted hydrocarbon reservoirs or aquifers. A CO2-brine co-injection in geothermal doublets could be for example an alternative process for low-intensity CO2 storage. A geothermal doublet consists in two wells drilled within a “hot” aquifer: a hot water producer and a cold-water injector separated around 1 000 to 2 000 m away. The main idea of this process is to inject a CO2 dissolved in water at surface, within normal geothermal operations for several years before breakthrough at the producer occurs (or economic re-circulation limit is attained). Injected CO2 concentration will be constrained by its solubility limit at the injection conditions (pressure, temperature and salinity). In this work, numerical modelling was done using CO2STORE option of ECLIPSE simulator. In order to evaluate the process feasibility and efficacy, a synthetic 3-D gridded geological model was build using various bibliographical data on typical geothermal installations. A sensitivity analysis in the synthetic model was performed in order to determine key parameters that maximize the CO2 storage capacity and breakthrough time at the producer. Initial reservoir temperature, injection temperature, injection rate, CO2 molar fraction at injection and gas relative permeability endpoint (Krg) were evaluated. Experimental design technique permitted to identify injection rate and CO2 molar fraction as the most relevant parameters in the process. Four operational scenarios were evaluated: (i) continuous CO2-brine co-injection: two injection rates were studied (200 and 300 m3/h); (ii) seasonal-varying brine-CO2 co-injection: 100 m3/h during spring/summer and 200 m3/h during fall/winter periods; (iii) alternated brine and brine- CO2 co-injection: normal geothermal operations during 6-months and brine- CO2 co-injection the others 6-months evaluated at two injection rates (200 and 300 m3/h); (iv) continuous CO2 injection: classical CO2 injection in supercritical state at equivalent CO2 injection rate that CO2-brine co-injection cases. As mentioned before, injection CO2 molar fraction is one of the key parameters in the process. Higher the injection CO2 molar fraction leads in higher the storage capacity. On the other hand, lower the temperature higher the solubility, so lower temperatures are preferred in order to maximize storage capacity. For all cases, injection temperature was fixed at 50°C and two molar fractions were evaluated (1.5 and 2 %). When solubility limit is reached at injection conditions a dense phase is formed in the aquifer (CO2 in super-critical state), that decrease storage capacity. It was also evidenced that reducing injection rate also prevents the development of free CO2 dense phase. Scenario (ii) resulted in the optimum storage capacity and the longest breakthrough time in the producer. The heterogeneity of the carbonate reservoirs is considered as a main challenge on the process performance, therefore a sensitivity analysis considering different geological scenarios was also performed considering different facies extension during the static model generation. It was shown that heterogeneities in the reservoir can affect negatively CO2 storage capacity when high permeability channels create preferential flow paths between injector and producer. Typical modelling in geothermal applications using homogeneous geological features was also evaluated resulting in very optimistic breakthrough time (ten-times longer than heterogeneous models). Spacing between wells was also addressed, by considering three spacing: 1000, 1500, 2000 m. Results showed that the longer the spacing the better the storage capacity due to breakthrough retardation. This study demonstrates that the combination of geothermal operation and CO2 co-injection can be suitable for small to medium GHG emitters while validated simulation software can be used to properly design the optimum operational strategy if a detailed geological model is used.