| Title | Quantifying geothermal resource potential and uncertainty analysis using a natural state model of Kotamobagu Geothermal Field in North Sulawesi, Indonesia |
|---|---|
| Authors | B.B.R. Nagoro, J. OSullivan |
| Year | 2023 |
| Conference | New Zealand Geothermal Workshop |
| Keywords | Geothermal resource potential, uncertainty quantification, 3D natural state model, geothermal reservoir modelling |
| Abstract | Kotamobagu Geothermal field, located in North Sulawesi, Indonesia is expected to hold significant potential as a geothermal energy resource. This study presents a calculation of its resource potential using an uncertainty analysis and a natural state model. Through the utilization of publicly available reports and data, a comprehensive 3D geological model of the field has been constructed. This process was followed by setting up a numerical model that was used to simulate the steady state of the geothermal system. Many sample models were run and conditioned against the limited data and this process was used to determine the reservoir potential. The construction of the 3D geological model involved the integration of geology, geophysics, geochemical and spatial data using LEAPFROG Geothermal. Based on this, a numerical model was developed that is compatible with both AUTOUGH2 and Waiwera. The model was calibrated by adjusting the permeability and upflow to simulate the natural state of the geothermal reservoir. Next, 1000 samples of possible alternative models were generated based on the distribution of upflow and the rock permeabilities and simulated using high-performance computing facility. These samples were then filtered using the estimated temperature below the clay cap. Finally, the filtered samples were used to estimate the resource potential by applying a production algorithm. The results show a range of reservoir temperature from 200-250°C with applying 109 kg/s upflow. The model estimates that the upflow of heat arises underneath Mt. Ambang and along the northeast-southwest trending faults. Additionally, the same family of faults at the east side of the volcano acts as a barrier preventing the fluid from seeping out. The filtering process resulted in 213 sample models which were used in the Approximate Bayesian Computation (ABC) to show that the P90, P50, and P10 estimates of power outputs are 70 MWe, 160 MWe, and 270 MWe, respectively, for the 25 years of simulated production. |