Numerical modeling of fractured geothermal reservoir is conducted to describe coupled hydromechanical behavior at Pohang Enhanced Geothermal System (EGS) site. A hydraulic stimulation was conducted in the PX-1 well at the depth of 4,362m in Pohang EGS site from Dec 2016. Stress-induced permeability changes are inferred to have occurred from well head pressure and injection rate versus time curves during the stimulation. A numerical model of Pohang EGS reservoir is built to simulate hydromechanical behavior during the hydraulic stimulation at PX-1 well. The well head pressure and injection rate curves are reproduced considering corresponding permeability changes by effective stress changes and hydroshearing. History in hydromechanical property changes such as permeability during the stimulation is estimated from the modeling results. In addition, a relationship between effective stress and permeability is obtained through model calibration against the well head pressure and injection rate data. For the numerical modeling, TOUGH-FLAC, a simulator for coupled thermal-hydraulic-mechanical processes in geological media, is used.
Pohang Enhanced Geothermal System (EGS) project has been operated in Pohang, South Korea since 2010. The geology consists of sedimentary rock from ground surface to 2.4km deep, and of granodiorite below the depth of 2.4km. Two boreholes, PX-1 and PX-2, are drilled up to 4,217m and 4,348m respectively. Hydraulic stimulations were conducted in PX-2 from Jan. 29 to Feb. 10, 2016, and in PX-1 from Dec 15, 2016 to Jan 11, 2017. In PX-2, 1,970m3 of water was injected, and the maximum wellhead pressure of 89.2MPa was observed. The total amount of injected water to PX-1 was 2,689m3, and maximum wellhead pressure reached 27.7MPa. The main flow path is expected to be one or two major fault zones intersecting PX-1 and PX-2. Hydroshearing on the pre-existing faults is highly likely to have happened at wellhead pressure of around 15MPa during the hydraulic stimulation in PX-1 according to a result interpretation (Park, S. et al., 2017).
