Quantifying Fracture Networks Inferred From Microseismic Point Clouds by a Gaussian Mixture Model With Physical Constraints

Abstract

Microseismicity is generated by slip on fractures and faults and can be used to infer natural or anthropogenic deformation processes in the subsurface. Yet identifying patterns and fractures from microseismic point clouds is a major challenge that typically relies on the skill and judgment of practitioners. Clustering has previously been applied to tackle this problem, but with limited success. Here, we introduce a probabilistic clustering method to identify fracture networks, based on a Gaussian mixture model algorithm with physical constraints. This method is applied to a rich microseismic data set recorded during the hydraulic fracturing of eight horizontal wells in western Canada. We show that the method is effective for distinguishing hydraulic‐fracture‐created events from induced seismicity. These fractures follow a log‐normal distribution and reflect the physical mechanisms of the hydraulic fracturing process. We conclude that this method has wide applicability for interpreting natural and anthropogenic processes in the subsurface.

The Effect of Correlated Permeability on Fluid‐Induced Seismicity

“…McKean et al. (2019) also found that fracture networks follow a log‐normal distribution. Therefore, log‐normality of the fracture network (McKean et al., 2019) and log‐normality of permeability field (Leary & Al‐Kindy, 2002; Malin et al., 2020) go hand in hand.…”

Machine Learning‐Based Analysis of Geological Susceptibility to Induced Seismicity in the Montney Formation, Canada

“…Extensive use of hydraulic fracturing (HF) for unconventional oil and gas development has been accompanied, in some parts of the world, by anomalous-induced seismicity including triggered earthquakes up to M W 5.2 (Lei et al, 2019). The use of spatiotemporal clustering (McKean et al, 2019;Zaliapin et al, 2008) makes it increasingly clear that induced seismicity hazard from HF exhibits a high degree of spatial concentration, including areas in western Canada (Atkinson et al, 2016;Bao & Eaton, 2016;Schultz et al, 2017), United States (Skoumal et al, 2019), and China (Dengfa et al, 2019). Seismicity rate has been correlated with industrial parameters such as total injected volume, but robust trends often emerge only after filtering event catalogs to account for intrinsic geological susceptibility (e.g., Kao et al, 2018;Schultz et al, 2018).…”

Injection Parameters That Promote Branching of Hydraulic Cracks

“…However, in heterogeneous systems, the reality is more likely to be complex fracture networks that include multiple strands and crack branching (Bažant et al., 2014). Such possibilities have been witnessed in lab experiments and field studies (Fischer et al., 2008; Frash et al., 2015, 2019; Haimson, 1981; Ishida et al., 2004; Jeffrey & Settari, 1995; Maxwell et al., 2015; McKean et al., 2019; Warpinski & Teufel, 1987). From the direct evidence alone, it is reasonable to deduce that hydraulic fracture patterns can be either planar structures or complex networks, perhaps transitioning from one to the other.…”