In geothermal energy production from hot aquifers, the cooled and condensed geothermal brine may be reinjected into the water-bearing strata to maintain reservoir pressure and sequester toxic minerals. The solubility of amorphous silica reduces due to loss of heat, and the concentration may increase due to loss of water. Deposition (scaling) of silica may result, narrowing the fluid pathways in the reservoir and reducing the injectivity and the economic lifetime of the well.

In the present work, a comprehensive model is developed, to predict the injectivity as a function of time under the effects of reactive transport of relevant chemical species (e.g. silica, calcite, anhydrite, etc.), using a finite volume Eulerian approach. The fluid pathways in the geothermal reservoir are modelled as parallel flat plates. 2D unsteady transport partial differential equations are solved to model the heat and mass transfer. The consumption of the reactants and the formation of the products are implemented using sink and source terms respectively. The surface chemical reactions are modelled using semiempirical formulas developed from experimental results reported by other workers, with care taken to select data taken under conditions close to those in geothermal reservoirs where possible. The decrease in reservoir porosity due to accumulated scale is modelled by applying the porosity-permeability correlation proposed by Verma and Pruess (1988). The increase in porosity due to reservoir stimulation processes is also modelled.

Since the reinjected fluids (i.e. injectate) are supersaturated with respect with amorphous silica, silica can deposit directly (i.e. molecular deposition) but also can polymerise, nucleate, and form nanoparticles. A validated semiempirical method is proposed to model the growth and Ostwald Ripening of silica nanoparticles. The interactions between colloidal silica are found to differ from those predicted by the classic DLVO theory. These interactions are found to be better described by the soft particle model (Ohshima 2015). Therefore, a semi- theoretical model of silica polymerisation developed by Weres et al. (1981) and the soft particle model proposed by Ohshima (2015) are integrated to quantitively model the amorphous silica deposition where the dissolved silica is actively polymerising. The parameter describing the thickness of the soft shell of silicate chains, which was not quantified in previous studies of the soft-shell model, is estimated by fitting to colloidal growth data to complete the model. The model outputs are validated by comparing to experimental results and real-world experience at laboratory and full scale (Huminicki and Rimstidt 2007; Tamura et al. 2018; Carroll et al. 1998; Mroczek et al. 2017; Van den Heuvel et al. 2018; Xu et al. 2004): the thermal and chemical sub-models are validated independently, as well as validating the code as a whole against injectivity measurements. The validation results suggest that the model can be useful in predicting the lifetime of geothermal injection under varying conditions.

A sensitivity study is carried out to rank processes which may be effective in extending the injection lifetime. The results suggest that., of the six analysed inputs (injection temperature, mass flow rate, pH, critical porosity and power exponent (used to estimated permeability based on porosity), feedzone thickness, and fracture aperture), injection temperature and pH appear to be more important than others.

Injection of chemical species that are expected to improve the geothermal injection (injection stimulation) is modelled. The model may be of use to estimate potential results of proposed stimulation plans. Several conceptual ideas to extend the injection lifetime are discussed, such as reservoir stimulation by acidizing, ageing and injection at low temperature, acidizing using weak acid, silica removal before injection, and reservoir recovery using hydrofluoric acid.

The developed holistic model is named as GEOREPR (GEOthermal Reinjection lifetime PRediction), which is offered as a free open-source code under the GNU (General Public License) since 2019, and may be developed further in the future. The potential users can use, modify, and redistribute the code for both commercial and non-commercial proposes. One can refer to https://github.com/MITHRILTech/GEOREPR for the latest version of GEOREPR.