Seismicity Induced by Wastewater Injection in Washington County, Ohio: Influence of Preexisting Structure, Regional Stress Regime, and Well Operations

Abstract

Recent seismicity in Washington County, Ohio, has been suggested to be induced by wastewater disposal operations despite injection ~2 km above the Precambrian basement. We investigated the relationships between disposal well locations and operational histories, spatiotemporal patterns of seismicity enhanced by waveform correlation, and mapped subsurface structures. We also analyzed proxies for in situ stress conditions to evaluate existing principal stress orientations and magnitudes. Double‐difference relocations of small (ML ≤ 2.1) seismicity revealed linear clusters defining a NE‐SW orientation that parallels the trend of small‐amplitude folds in Upper Paleozoic rocks within 1–2 km of the injection well. Hypocenter depths of ~3.1 to 4.3 km indicate seismicity occurred along faults below the Silurian injection interval in lower Paleozoic strata and Precambrian crystalline basement. Similar folds/faults imaged in nearby seismic reflection lines were associated with basement‐involved fault systems that cut the injection interval targeted by the injection well, providing permeability pathways for fluid pressure increases that initiated slip. Interpreted faults were also optimally oriented in the regional stress field for reactivation due to injection‐related decreases in effective stress. Well completion records suggest that during the recorded seismicity, the Long Run‐1 well was injecting disposal fluids close to or above the fracture pressure of the targeted Silurian injection interval. A comparison of observed seismicity rates and monthly injection records from the disposal well also indicate a close correspondence between earthquake frequency and injection rate. Taken together, these results provide a physical framework for understanding why wastewater disposal operations induced the Washington County seismic sequence.

Seismicity induced by hydraulic fracturing and wastewater disposal in the Appalachian Basin, USA: a review

“…The relocated hypocentral depths correspond to lower Cambrian and upper Precambrian strata, reported by Baranoski (2013). The epicenters formed elongated cluster with NE-SW trend that Currie et al (2018) also found parallels the trend of small-amplitude folds in Upper Paleozoic rocks and is optimally oriented within the regional stress field. Hypocentral patterns and similar structures imaged in nearby seismic reflection lines indicate the seismicity occurred on a positive flower structure initiating in the basement and separating into braided segments in the sedimentary layers.…”

“…As expected, seismicity rates followed the monthly injection rates during the studied time frame. Taken together, Currie et al (2018) concluded the seismicity was the result of pore-pressure change caused by WD, similar to the Youngstown sequence.…”

“…The HypoDD double-difference algorithm (Waldhauser and Ellsworth 2000) was applied to relocate most of the induced sequences discussed in "Documented episodes of seismicity induced by WD" and "Documented episodes of seismicity induced by HF" sections (Skoumal et al 2014(Skoumal et al , 2015a(Skoumal et al , b, 2018Kozlowska et al 2018). Each of these studies has used a 1D velocity model originally derived for eastern Ohio and adjusted to local conditions.…”

“…In 2013, ODNR deployed a local network of five stations in the area of observed seismicity. Currie et al (2018) analyzed their data between May 2013 and July 2015 and performed multi-station template matching identifying over 300 events down to magnitude − 0.7. The double-difference relocation of these events with the local network (Fig.…”

Seismicity induced by hydraulic fracturing and wastewater disposal in the Appalachian Basin, USA - a review

“…Despite the local variability in the prevalence of induced seismicity, there is remarkable coherence in the orientation of faults that have induced seismicity. Many of these sequences have created extensive sequences of similar events with nearby recordings to allow for double difference relocations that reveal linear fault segments (Friberg et al 2013;Skoumal et al 2014;2015a;2015b;Kozlowska et al 2018;Currie et al 2018). In addition, many of these sequences have produced a large enough event to determine focal mechanism (Kim, 2013;Friberg et al 2013;Skoumal et al 2015a;2015b;Kozlowska et al, 2018).…”