Evaluating Liquefaction Triggering Potential from Induced Seismicity in Oklahoma, Texas, and Kansas

Deep wastewater injection-induced seismicity has led to over a thousand magnitude (Mw) > 3 earthquakes and four Mw>5 earthquakes in Oklahoma, Texas, and Kansas (OTK) over the last ten years. Liquefaction observed following the 3 September 2016, Mw5.8 Pawnee, OK, induced earthquake raises concerns regarding the liquefaction risk posed by future induced earthquakes. The stress-based simplified liquefaction evaluation procedure is widely used to evaluate liquefaction potential. However, empirical aspects of this procedure were primarily developed for tectonic earthquakes in active shallow-crustal tectonic regimes (e.g., California). Consequently, due to differences in ground motion characteristics and regional geology, the depth-stress reduction factor (rd) and Magnitude Scaling Factor (MSF) relationships used in these variants may be unsuitable for use with induced earthquakes in OTK. This is because both rd, which accounts for the non-rigid soil profile response, and MSF, which accounts for shaking duration, are affected by ground motion and soil profile characteristics.

The objective of this research is to develop and test a new liquefaction triggering model for use in assessing the regional liquefaction hazard in OTK from injection-induced earthquakes. This model incorporates induced seismicity-specific rd and MSF relationships. To assess model efficacy, the liquefaction potential is evaluated for several sites impacted by the 2016 Pawnee earthquake using the model developed herein, as well as several models commonly used to evaluate liquefaction potential for tectonic earthquakes. Estimates are then compared with field observations of liquefaction made following the Pawnee event. This analysis shows that, at most sites, the induced seismicity-specific model more accurately predicts liquefaction severity than do models developed for tectonic earthquakes, which tend to over-predict liquefaction severity.

The liquefaction triggering model developed herein is also used to assess the minimum magnitude (Mmin) of induced earthquakes capable of triggering liquefaction. For sites capable of supporting structures, it is shown that Mmin = 5.0 is sufficient to fully capture liquefaction hazard from induced events in OTK. However, for extremely liquefaction-susceptible soil profiles that are potentially relevant to other infrastructure (e.g., pipelines and levees), consideration of Mmin as low as 4.0 may be required.