The fluid flow in a reservoir largely depends on faults, as faults can act like conduits or barriers. The influence of three fault parameters (permeability, thickness and angle) on the temperature and pressure behaviour of a geothermal reservoir, has been investigated by modelling a faulted reservoir in 2D in MATLAB, based on the Finite Element Method. The reservoir is assumed to be sandstone with a permeability of 40*10-14 m2. The fault has been modelled with a permeability ranging from 10-11 m2 to 10-17 m2, a thickness ranging from 10 cm to 150 m and an angle ranging from 0 to 90 degrees. While varying one parameter, the other two stayed fixed. The base values for the fault permeability, thickness and angle were 10-14 m2, 20 m and 0 degrees respectively. Both a finite and infinite fault have been modelled in the reservoir. Decreasing the permeability led to a higher production temperature, especially at a finite fault, where the fluid is able to bypass the fault and thus sweep the warmth of a larger part of the reservoir. At a fault permeability of 10-14 m2 the fluid has seemed to have extracted the maximum amount of heat in the reservoir, further decrease in permeability did not lead to a higher production temperature. At an infinite fault there is only a small increase in temperature at a fault permeability of 10-16 m2. The impedance at a finite fault is not much affected, as the fluid is able to flow around the fault. The impedance at an infinite fault, however, rapidly builds up with decreasing permeability. A thickness of 20 m, at the base permeability value of 10-14 m2, results in the highest production temperature. Larger thicknesses cause larger amounts of unrecoverable area and consequently less heat can be extracted. At a finite fault the impedance climbs only to a value of 4 MPa at a thickness of 150 m, but at an infinite fault the impedance already reaches 10 MPa at a thickness of 80 m. The fault has its largest influence at a perpendicular position to the flow. At a parallel position, the results are the same as a reservoir without fault, as the fluid will not have to dodge a lot to reach the production well. All three fault parameters have a significant influence on the temperature and pressure behaviour of the geothermal reservoir. They are all able to enlarge and eliminate the influence of the fault. A finite fault can cause a favourable higher production temperature without a large impedance, as the fluid is able to bypass the fault. An infinite fault has less effect on the temperature and causes the impedance to build up rapidly, which requires an unrealistic high pumping power and consequently an unprofitable geothermal system.

This study investigates the thermal recharge of deep geothermal reservoirs by free convection. A coupled numerical model that is built in Comsol Multiphysics is presented in order to estimate the thermal recharge during the first 30 years of production from a typical high enthalpy geothermal reservoir. A sensitivity analysis was carried out to find which parameters have most significant impact on the thermal recharge by free convection and reservoir lifetime. The sensitivity analysis is mainly focussed on fault properties. The developed model consists of a reservoir with two confining layers and vertical faults. This model is used to calculate the produced energy, extracted energy from the reservoir and the recharge by conduction based on which the thermal recharge by free convection can be calculated. Results show that free convection is a significant part of the produced energy. Therefore, it is important to take free convection into account by estimating the recoverable energy. From the results can also be concluded that fault aperture, fault height, fault permeability and thermal conductivity of the confining layers have a significant impact on the thermal recharge by free convection of a faulted geothermal system. Fault density also has impact on the thermal recharge, but the relation between fault density and thermal recharge by free convection is less clear. Thermal conductivity of the reservoir rock has no impact on the thermal recharge by free convection according to the results of this model.