Efficient fracture network characterization in enhanced geothermal reservoirs by the integration of microseismic and borehole logs data

Abstract

Characterization of 3D fracture networks induced by hydraulic fracturing in enhanced geothermal reservoirs is crucial for safe and effective energy recovery from deep geothermal reservoirs. Microseismic monitoring is one of the most promising ways to infer the geometry of fracture network. However, the identification of fracture network from microseismic events is still a challenging task, attributed to high noises in microseismicity interpretation and fuzzy relationships between microseismic event locations and fracture planes. This study proposed a novel methodology to identify fracture network from the spatial distribution of microseismic events. The method was composed by (1) density-based spatial clustering algorithm for microseismic data denoising, (2) Monte Carlo optimization algorithm for fractures localization by minimizing the total distance between denoised microseismic events and fracture planes, and (3) Elbow method to decide the most likely number of fractures. Moreover, the method employed the prior knowledge of geo-stresses and initial fracture patterns inferred from borehole logs to constrain the geometry of fracture network. Following verifications with synthetic fractures and microseismic events, the method was implemented in a realistic enhanced geothermal system, which reliably identified the geometry of 3D fracture network induced by hydraulic fracturing.