CarbFix Pilot Site. The original CarbFix site is located 3 km SW of the Hellisheiði geothermal plant (Fig.
1). During this pilot phase, two slug gas injection tests took place. The first one involved the injection of 175 tons of pure commercial CO2 from January to March 2012; the second one 73 tons of a CO2-H2S mixture captured and purified at the plant between June and August 2012 (Matter et al., 2016). Forced flow was induced by the continuous injection of brines in HN-2 and production at well HN-4 from 2011 to 2014. Water pumped from well HN-1 was co-injected with the gas in well HN-2 to a depth of 340 m, where it was released in the downflowing water through a xxxxxxx. The mixture was then carried down to 540 m through a mixing pipe, where the gases fully dissolved in the water as the hydrostatic pressure exceeded 40 bar (Sigfússon et al., 2015). The aquifer targeted flows through a permeable formation composed of altered tholeiitic lava flows (400-800 m) with a porosity of 8.5% (Aradóttir et al., 2012), which are overlain by a glassy hyaloclastite formation of low permeability (Xxxxxxxxxx et al., 2013). Secondary minerals such as smectite, calcite, and Na-zeolite are the most abundant alteration phases at these depths. The hydrology of the system was investigated prior to the gas injections with tracer tests and reported by Khalilabad et al. (2008) and Xxxxxxxxx et al. (2012). Tracer tests indicated two breakthroughs; while the water mainly travels by homogeneous porous flow, there is also a fast-flow path that channels 3% of the injected fluid volume through the formation. The composition of the HN-4 samples between 2012 and 2014 are published in four articles:
1) Major elements: Matter et al., 2016; 2) Xxxxxxxxxxxxxxx et al., 2017; 3) Mg isotopes: Xxxxxxx et al., 2019; 4) Ca isotopes: Xxxxx xxx Xxxxxxxxxx et al., 2019. Xxxxxxxxxxxxxxx et al. (2018) also reported reaction path modelling of in-situ CO2 mineralization. These studies estimated the quantity of CO2 and H2S captured between the injection well and the monitoring well. Xxxxxx et al. (2016) interpreted the evolution and arrival of the carbon- enriched plume from the analysis and evolution of several inert tracer (14C, SF6, and SF5CF3) concentrations and the associated variation of pH and dissolved inorganic carbon (DIC) content. The DIC measured is much lower than that expected from simple mixing of the injection waters in the aquifer (non-reactive transport). From this observation, Matter et al. (2016) estimate...
CarbFix Pilot Site. The activies reported in this Deliverable benefit from a large amount of field-site monitoring and sampling, as well as modelling analysis which were completed when the S4CE project started. This material allowed S4CE to focus on better constraining the mineralization of CO2 by investigating additional isotope systems. The Fe isotope results reported here indicate that they could be useful tracers to determine Fe-carbonate precipitation when only pure CO2 is injected. On the contrary, it appears that Zn isotopes are not controlled by mineralization processes. Further confirmation is needed, which can be obtained by analyzing pre-injection groundwaters. In this case, Zn isotopes cannot be used to quantify the fixation of CO2 and H2S, which had been injected at the CarbFix site. We have met the goal of quantifying of CO2 fixation for the original CarbFix site. Indeed, the articles of Matter et al. (2016) and Xxxxxxxxxxxxxxx et al. (2017) constrain the amount of CO2 captured, while the experiment of Xxxxx et al. (2019) confirms the lower boundary of the “sweet spot” of mineralization modelled by Xxxxxxxxxxxxxxx et al. (2018). Isotope work was carried out to further determine the fate of CO2 in specific precipitating minerals.
