Investigations of heat extraction for water and CO2 flow based on the rough-walled discrete fracture network

The Enhanced Geothermal Systems (EGS) has been developed to enhance geothermal energy extraction efficiency from geothermal reservoirs by generating effective fracture network. The fracture effective network that are comprised by rough-walled fractures can provide a more realistic reflection of natural reservoir conditions. This study focuses on the construction of a rough-walled discrete fracture model to simulate mass and heat transfer in a geothermal reservoir by using water and CO2 as the working fluid. The thermal-hydraulic-mechanical (THM) coupling process was integrated into the rough-walled discrete fracture network model. The heat mining processes with two working fluids (water and CO2) were simulated with the parallel-plate and rough-walled discrete facture network models. The influences of the parallel and rough-walled discrete fracture network models on flow behaviours, heat transfer and corresponding mechanical responses for water and CO2 were presented and analysed. What's more, the heat extraction efficiency based on the combination of rough-walled discrete fracture network and the THM method were calculated and analysed in details. The influences of pressure, temperature, fluids and reservoir rock properties on heat extraction rates were also considered. It is found that CO2 leads to faster pressure changes compared with water in both parallel-plate and rough-walled discrete fracture network models. The influences of the pressure and temperature distributions on the permeability and normal stress were also evaluated, which contribute to the investigation of the deforming mechanisms of the rock matrix and fractures. In addition, there are certain differences in the heat production rate between the parallel-plate and rough-walled discrete fracture networks, which reflects that the rough-walled discrete fracture network used in this study has a better conductivity. It is also found that water can be a more efficient working fluid for heat extraction from geothermal reservoirs compared with CO2 during certain production durations in this study. The results of this study present that the rough-walled discrete network model has a great significance in the simulation of the real conditions and processes in geothermal reservoirs during the heat mining process.