| Title | A New Modelling Study of the Kamojang Geothermal Field |
|---|---|
| Authors | John O’SULLIVAN, Fathan ABDURACHMAN, Greg BIGNALL, Chris BROMLEY, Ken DEKKERS, Muhammad GHASSAN, Michael GRAVATT, Pudyo HASTUTI, Astri INDRA, Rony NUGRAHA, Fernando PASARIBU, Joris POPINEAU, Imam PRASETYO, Vicky RAI, Theo RENAUD, Jeremy RIFFAULT, Dhanie YUNIAR, Michael O’SULLIVAN |
| Year | 2023 |
| Conference | Stanford Geothermal Workshop |
| Keywords | Kamojang, Geothermal reservoir simulation, Integrated modelling, AUTOUGH2, Waiwera |
| Abstract | Indonesia has the second highest geothermal power installed capacity worldwide and expects to develop more geothermal energy in the future. Kamojang is a vapour-dominated geothermal field located 40 km to the southeast of Bandung in West Java, Indonesia. It was the first geothermal project to be developed in Indonesia. In this paper we discuss a multidisciplinary approach to a new modelling study of the Kamojang Geothermal Field (KGF), Indonesia. The project team included members from PT. Pertamina Geothermal Energy, the Geothermal Institute at the University of Auckland, Geoenergis Solusi Indonesia, and independent New Zealand geothermal experts. It successfully achieved its objectives through close collaboration and an inclusive, transparent approach. A review and update of the conceptual model of the KGF has been conducted. The updated conceptual model was set up as a digital conceptual model in the Leapfrog software and used to develop a new numerical model. In setting up the numerical model we use an integrated modelling framework developed at the University of Auckland that allows the direct creation of a numerical model from a Leapfrog-based digital conceptual model. The numerical model can be run in AUTOUGH2, a local variant of the industry-standard simulator TOUGH2, or in Waiwera, a highly parallelised, open-source simulator developed by the University of Auckland and GNS Science. The new numerical model was calibrated against the measured data using the industry-standard approach. The state of model calibration achieved is good and we are confident that the model provides a good representation of the KGF and is appropriate to use for forecasting. |