Fracture network organization with continuation of injections in the Habanero field (Cooper Basin, Australia) disclosed by coda wave fluctuations

Abstract

Hydraulic stimulation of a geothermal reservoir aims to increase the permeability of the rock formations by creating/enhancing a fracture network, to take advantage of the geothermal energy hosted by deep hot formations. A common side effect of the fracturing process is the small-in-magnitude but abundant induced seismicity. In this work, we study the seismicity induced during the hydraulic stimulation of the Habanero 4 well, in the Cooper Basin (Australia), where an extensive network of sensors has been deployed to monitor the stimulation process. The detailed monitoring of the induced seismic events enabled the study of the dynamic characteristics of the enhanced fracture network by making use of the seismic coda waves scattering properties. Since the Earth’s crust has inhomogeneities with fractal properties, manifested in the seismic coda waves, we address the fracture network as a complex, non-equilibrium system. In this line, the framework of Non-Extensive Statistical Physics (NESP) and the maximum entropy principle were used to study the fluctuations of the seismic coda waves and their possible variations during the hydraulic stimulation. The analysis shows that the probability density functions of the normalized increments of S-coda waves present heavy tails with asymptotic power-law behavior, enhancing the probabilities of large fluctuations. The results further indicate that as the stimulation progresses, this enhancement becomes even more pronounced. The observed scaling behavior can exactly be approximated with the q-Gaussian distribution, derived in the framework of NESP, showing a modest uprising trend of the entropic index (q-value) with continuation of injections. These results suggest that the system of fractures progressively becomes more organized, with stronger memory effects and long-range interactions, as the hydraulic stimulation progresses.