Deep-Embedded-Clustering of Microseismicity Identifies Multiple Failure Mechanisms at The Geysers Geothermal Field
P. Lara, N.R. Ranasinghe, V.V. Vesselinov, C.W. Johnson
Geophysics (submitted)2026DOI 10.22541/essoar.177265440.06972526/v1
Summary
Enhanced Geothermal Systems regularly produce microseismic activity as fluid injection perturbs the thermo-hydromechanical state of the reservoir. This study applies Deep Embedded Clustering to waveform, spectrogram, and cepstrogram data from the 2011–2012 stimulation of the Prati-32 EGS well at The Geysers, identifying three distinct failure classes and helping relate waveform characteristics to underlying physical mechanisms.