Indonesia Deep learning method in micro-earthquake phase determination in geothermal field

I Putu Raditya Ambara Putra, Muhamad Firdaus Al Hakim, Mohammad Rachmat Sule, Fanata Yudha Nugraha, Hadyan Pratama, Dina Kholish Annisa, Haris MuhlisinNon-Scopus1 April 2025Afiliasi: ID

Abstrak

Abstract Renewable energy has become a major focus in efforts to reduce greenhouse gas emissions and reduce dependence on fossil fuels as a result of increasing electricity demand worldwide and the challenges of climate change. One of the renewable energy potentials that Indonesia has is the utilisation of geothermal energy, where we know Indonesia is located in the ring of fire which of course has many sources of heat energy underneath. The utilisation of geothermal energy can usually change the stress conditions below the surface, which can trigger micro-earthquakes. Monitoring of micro-earthquakes is very important to know the stress conditions and even the structural conditions in the geothermal field. This micro-earthquake monitoring is very dependent on the accuracy in determining the arrival phase of the wave, especially for phase P and S. Automatic phase detection methods have been developed both conventionally and with a machine learning approach. In this study, we will compare the automatic detection results of one machine learning model with some conventional methods that are very commonly used. When compared to other conventional models, the results of automatic detection utilizing PhaseNet (a machine learning model) indicate less time difference than manually pick catalog. It is also reasonably straightforward to use, with minimal parametrization.

Identifier

Scroll to Top