Factors Controlling Rate and Magnitudes of Injection-Induced Earthquakes

Abstract

Advancements central to the world’s transition to a greener economy such as geothermal energy are hampered by a lack of understanding and control of the maximum magnitude event. We develop a physical model for induced seismicity rate and identify the dominant factors on induced event magnitudes from the data. Induced seismicity observed during Enhanced Geothermal Stimulation (EGS) at Otaniemi, Finland and the Basel Deep Heat Mining project are reproduced by a physical model based on pore pressure diffusion and rate-and-state friction. The physical models produce simulations closest to the observations when assuming rate-and-state friction for shear failure with a diffusivity matching the pressure build-up at the well-head at onset of injections. At the same time, pressure drawdown immediately following shut-ins is best modeled with a lower diffusivity, indicating a significant change in diffusivity through fracture opening and closure. Seismicity rate following shut-in in Basel also indicates slower diffusion mechanisms. The potential of the model to serve as a forecasting tool is demonstrated by a pseudo-forecast that closely matches the entire seismicity rate history, using only the first injection stage as the training period. Next, we focus on the magnitude-frequency distribution of seismicity and how they may correlate to the injections. We observe a dominant, negative correlation of the b-value and maximum magnitude to depth in Otaniemi, although we do not observe the same relationship in Basel. Following a correction of the magnitudes to de-trend the dependence on depth in Otaniemi, both catalogs show a positive correlation between maximum magnitude and cumulative injected volume. We also observe a similar non-linear relationship between the magnitude-frequency distribution and distance from the injection source for the two catalogs. The identification of factors controlling seismicity rate and magnitudes provides basis for a control and optimization framework that may be utilized for sophisticated design of injection schedules.