Exploring CO2 geological storage and geothermal productivity benefits from EGS application in the Paris Basin saline aquifer: A numerical approach

Recent numerical studies have shown good prospect of the use of CO2 as a working fluid in Enhanced Geothermal Systems (EGS) with the additional benefit of achieving geological sequestration of CO2 (Pruess, 2006). At present, no practical application of this methodology has been carried out yet, and a numerical approach is required to preliminary assess the benefits related to the use of CO2 as heat transmission fluid in geothermal settings. In this paper we present a case study in the Paris basin geothermal field, France, and the numerical simulations we performed to investigate both CO2 storage capacity and enhanced geothermal productivity aspects related to the injection of supercritical CO2 at depth.
The application of EGS techniques in the Paris Basin is appealing because water injectivity tests already performed in the field have revealed a rather low productivity of reservoir rocks. Due to the favorable thermodynamic and transport properties of supercritical CO2 with respect to liquid water (i.e. larger expansivity, lower viscosity, capability as solvent to reduce or inhibit overall geochemical reactivity and scaling effects; Brown, 2000), carbon dioxide injection is expected to be a possible way to locally enhance permeability or, at least, to reduce the transmissivity decline observed with water as working fluid.
The target of our study is a permeable sandstone formation of Triassic age which has already been envisaged for geothermal exploitation. The reservoir is located at a depth higher than 2000 m b.g.l. and is characterised by a temperature of about 150 °C and a pressure of 200 bars and fully liquid conditions (i.e. Sl, liquid saturation, is 1.0). Simple idealized geometries have been applied to represent the Triassic formation. In our model fluid injection and production is performed according to the classical geothermal doublet scheme. The sensitivity of the model has been tested for different system parameters, such as injection rate, well spacing and spatial permeability distribution. Simulation results under CO2 injection conditions are directly compared to water injection system.
In a first phase, non isothermal multiphase flow modeling has been conducted in order to predict heat and mass flow production rate. Based on the results from the first phase, reactive transport modeling has been then carried out to evaluate the feedback of geochemical reactions likely induced by the injection of water and supercritical CO2 on reservoir porosity and permeability. The amount of CO2 permanently trapped as dissolved and mineralized phase has been also evaluated.
All simulations have been performed with the TOUGH2 family codes from the Lawrence Berkeley National Laboratory.