APPLICATION OF ARTIFICIAL INTELLIGENCE IN ESTIMATION AND OPTIMIZATION OF GEOTHERMAL RESERVOIR PRODUCTION

• Main Author: Kusumo Wardoyo, Galih
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/48341
• Institution: Institut Teknologi Bandung

Reservoir is the main asset of geothermal exploitation business. The decline in reservoir performance affects the sustainability of geothermal fluid production. Reservoir has a heterogeneous nature with high degree of uncertainty, so the development and estimation of reservoir performance using a numerical model will require a relatively long time. This study aims to create alternative models for optimization and estimation of geothermal production with artificial intelligence (AI). The AI model was arranged, trained and validated using measurement data of production parameters (wellhead pressure, wellhead temperature, trim valve open, fluid mass flow) in the Patuha steam domination geothermal field, Indonesia with variations scenario use of the data and implemented on a single well for prediction, multi wells for prediction and optimization. The model created in the study produced a mean absolute percentage error <9% with the scenario of using 80% data for the training process and validation using 100% (of the total data). The AI model can be used as an alternative method for optimizing and forecasting Patuha geothermal reservoir production with a relatively shorter time than the current method.