CarbFix at the Stanford Geothermal Workshop


Thomas Ratouis, a reservoir engineer at Reykjavík Energy presented his paper on “Modelling the Complex Structural Features Controlling Fluid Flow at the Carbfix Reinjection Site, Hellisheidi Geothermal Power Plant, SW-Iceland” at the 44th Annual Stanford Geothermal Workshop. His presentation attracted enthusiastic guests who ended up filling the conference hall.

His paper will be a part of the 44th Workshop on Geothermal Reservoir Engineering Proceedings and can be read here.


Injection of CO2 and H2S emissions from the Hellisheiði Geothermal Power Plant, SW-Iceland, as part of the CarbFix project, has been taking place in the Húsmúli reinjection zone since 2014. To study the fate of the dissolved CO2 and H2S in the geothermal reservoir, a large-scale three-dimensional model centered on the Húsmúli reinjection zone and the nearby Skardsmýrarfjall production zone is currently being developed. This model will be used to investigate the flow-paths of the injected fluid, the depth to which the injected fluid circulates, and the geochemical reaction between the injected acidic fluid and the basaltic host-rock along the flow path.
Injection of separated water from the power plant started at the Húsmúli injection site in 2011 and a tracer test was conducted at the site in 2013-2015. Simple flow path models using tracer recovery data from that test indicate that production wells exhibiting significant tracer recovery could be seriously affected by the injection of cooler fluids (cooling of up to 25-30°C). However, monitoring data collected since large-scale reinjection started at Húsmúli in late 2011 doesn’t indicate any significant cooling of the production wells monitored, which contradicts the cooling predictions of these simple models. This indicates that a more complex representation of the subsurface at Húsmúli is required to model the flow of the injected fluid.
The work here presents the preliminary efforts to constraint the pure water flow model against the 2013 tracer test data available using the TOUGH2 non-isothermal flow simulator. It was found that to match the tracer returns, strong anisotropy controlled by large extensional and strike-slip faults is necessary in the model, which act as preferential pathways for the fluid. In addition, it was found that a dual porosity approach was required to replicate the fast and strong recovery of tracers found in wells at Skardsmýrarfjall. The method of Multiple Interacting Continua (MINC), a generalization of the dual-porosity concept used here, allows for a better numerical approximation of the flow in a fractured continuum and includes transient fracture-matrix interactions.
This work will be used as the basis for a fully coupled reactive transport model to capture the mineralization processes of the injected, dissolved CO2 and H2S along the flow path, and to estimate the extent of the effective storage area of the basaltic host rock at Húsmúli. The impact of long-term reinjection of separated water on the Skardsmýrarfjall production area is also of interest here. This model will be used to estimate any potential adverse effects on the enthalpy of production at Skardsmýrarfjall as part of the resource management of the Hellisheiði geothermal resource.