Examining the Capability of Statistical Models to Mitigate Induced Seismicity during Hydraulic Fracturing of Shale Gas Reservoirs

Abstract

Injection into the subsurface is carried out by industry for a variety of reasons: storage of waste-water; enhanced oil recovery; and for hydraulic fracture stimulation. By increasing subsurface pressures, injection can trigger felt seismicity (i.e. of sufficient magnitude to be felt at the surface) on pre-existing faults. As the number of cases of felt seismicity associated with hydraulic fracturing has increased, strategies for mitigating induced seismicity are required. However, most hydraulic stimulation activities do not induce felt seismicity. Therefore a mitigation strategy is required that is capable of differentiating the "normal" case from "abnormal" cases that trigger larger events. In this paper we test the ability of statistical methods to estimate the largest event size during stimulation, applying these approaches to two datasets collected during hydraulic stimulation in the Horn River Shale, British Columbia, where hydraulic fracturing was observed to reactivate faults. We apply these methods in a prospective manner: using the microseismicity recorded during the early phases of a stimulation stage to make forecasts about what will happen as the stage continues. We do so to put ourselves in the shoes of an operator or regulator, where decisions must be taken based on data as it is acquired, rather than a post hoc analysis once a stimulation stage has been completed. We find that the proposed methods can provide a reasonable forecast of the largest event to occur during each stage. This means that these methods can be used as the basis of a mitigation strategy for induced seismicity.

Real‐Time Imaging, Forecasting, and Management of Human‐Induced Seismicity at Preston New Road, Lancashire, England

“…are similar to values found by Verdon and Budge (2018) for hydraulic fracturing in the Horn River Basin, British Columbia, Canada, where -4 < SI < -1, and towards the lower end of the range found by Schultz et al (2018) for hydraulic fracturing sites in Alberta, Canada, where -2.5 < SI < -0.5. The most notable past case of injection-induced seismicity in the U.K. for…”

“…We note that, as found by Verdon and Budge (2018), SI follows a similar trend to log 89 %&& , which is not surprising because the total moment release will depend on the number of events that occur. We also find that the +01 values derived from the SI measurements are larger than those derived from the %&& measurements, as also found by Verdon and Budge (2018). Dinske and Shapiro (2013) catalogue SI values for a range of injection sites, finding values ranging from -9 < SI < 1.…”

“…If a fault is intersected events may begin to line up along the structure, allowing it to be identified and mapped (e.g., Maxwell et al, 2008;Wessels et al, 2011;Hammack et al, 2014;Kettlety et al, 2019Igonin et al, 2019Eyre et al, 2019). In many cases fault reactivation can also be identified by a decrease in Gutenberg and Richter (1944) b values (e.g., Maxwell et al, 2009;Verdon and Budge, 2018;Kettlety et al, 2019), or by an increase in the rate of microseismicity relative to the injection rate (e.g., Maxwell et al, 2008;Verdon and Budge, 2018).…”

“…As such, expected event magnitudes can be forecast by characterising these relationships for the site in question, and then extrapolating them to the planned injection volume. This approach has shown significant promise when applied in a pseudo-prospective manner (e.g., Verdon and Budge, 2018).…”

“…In practice it is more useful to define a value for +01 that is unlikely to be exceeded. Using synthetic event distributions, Verdon and Budge (2018) showed that adding a value of 0.5 to Equation 2 is sufficient to capture 95% of the variance between true and re-constructed model populations.…”