Injection-Induced Seismicity: End of the Beginning?

For the last several years there have been numerous technical workshops on injection-induced seismicity which, looking back, serve as snapshots documenting the rapid evolution in our understanding of this important topic. With time, the tone of these workshops has changed in step with advancement by industry, regulators and academics managing this critical issue. Early discussions focused on education and awareness of the causes and cases of induced seismicity. This was followed by reactive implementation of regulated traffic light systems, and now development of more proactive operational strategies to mitigate induced seismicity. Within a cooperative environment including various stakeholders, industry best practices have emerged. Now mitigation success stories can potentially be seen, most notably in both Western Canada and in the state of Oklahoma, where seismicity rates appear to have declined over the past few years. Improvements coincide with both operations slowing down during the financial downturn and impacts of operational improvements and regulations. Such improvements led to Professor Mark Zoback (Stanford) to recently proclaim “the end of the beginning”, recounting Sir Winston Churchill’s famous quote at the turning point in WWII towards allied victory.

Over the past few years, there have been an increasing number of induced seismicity workshops and focused sessions at annual meetings of professional societies. For example, the SEG and SPE have co-sponsored a series of workshops starting in Broomfield in 2012, followed by events in Banff in 2014, Fort Worth in 2016 and Dallas in November 2017. In Calgary, an initial workshop was sponsored by CSEG Chief Geophysicists Forum and MUG in November 2013, targeting industry awareness of the issue. Among other local topical workshops and technical events, the Canadian Society of Unconventional Resources (CSUR) has been hosting a workshop series since 2015, targeting the issue from a Western Canada perspective. In all cases, these meetings have been a gathering between operating companies, researchers, regulators and vendors, which has demonstrated the cooperative exchange of information, experiences and learnings dealing with injection-induced seismicity.

The resurgence of induced seismicity has been surprising and therefore most of the early meetings were largely educational in nature, helping to provide access to existing scientific information about injection-induced seismicity observations and causal mechanisms. Discussions of case histories from the 70’s and 80’s at Rocky Mountain Arsenal and Rangely oilfield, the Eagle Field near Fort St. John BC and Strachan gas plant near Rocky Mountain House served as reminders that that this is not a new issue, although recent experiences have led to a revival of interest in the topic of injection-induced seismicity. “Traffic light protocols”, initially developed for geothermal operations using local seismic data, were introduced. As experiences and occurrences of induced seismicity expanded, these traffic light protocols have become the foundation of industry responses and regulations. The need for sensitive seismic recording to inform these protocols and to understand the seismic response of specific operations has led to the rapid expansion of local seismic monitoring arrays. A rich database of seismic observations is being acquired throughout North America, although much of this data is proprietary and supplements public arrays with enhanced detection. The seismic information is typically used to just trigger the traffic lights but more importantly, the seismic data is often closely scrutinized to understand fault activation and the corresponding seismic characteristics to better inform mitigation strategy decisions.

Considering the CSUR workshop last December in comparison to the SEG/SPE workshop a month earlier in Dallas, an interesting contrast between Canadian and US approaches is evident. Differences should perhaps be expected, since the types of injections inducing seismicity tend to be different. In Oklahoma, seismicity is mostly associated with widespread salt-water disposal. The outcome of this is the development of regional patterns of seismicity, apparently induced by continuous injection into multiple wells, resulting in complex causal relationships to specific injection wells.

Oklahoma focused research typically involves integrated reservoir flow, geomechanical and seismological models to understand the combined effect of regional injections. Characterization investigations have focused on mapping and categorizing faults in terms of propensity to slip as a logical proxy for seismic hazard. Traffic light protocols have resulted in regulated injection reductions in wells on a regional basis. Notably, the USGS has issued an updated US seismic hazard that includes increased risk in seismogenic regions associated with injection-induced seismicity.

In contrast, Western Canada has experienced seismicity mostly attributed to hydraulic fracturing operations in the Horn River Basin, Montney and Duvernay Shale plays. Hydraulic fracturing-induced seismicity tends to be localized to the treatment wells and occurs either during the active operations or shortly thereafter, making the causation between specific operations and induced seismicity sequences much more straight forward. Regulatory responses to traffic light protocols target specific causal wells. In Western Canada, there has been much more focus placed on understanding ground motions. An early workshop hosted by the University of Calgary in the fall of 2015 critically examined common traffic light protocols based on observed magnitudes and reviewed potential benefits of a ground motion-based protocol. Despite challenges intrinsic to relying on observed ground motions, including local site amplification effects, the characteristic is more closely associated with undesired consequences of damage or being felt by the local population. Indeed, one of the main challenges identified during various topical workshops is the need for data to calibrate the near-source ground motion of induced seismicity. The BCOGC was the first regulatory body to dictate ground motion measurements, in part to quantify damage potential and understand felt earthquake reports from local residents.

Ground motion was a recurring theme throughout the December 2017 CSUR workshop, serving as the focus of several talks. Presentations highlighted new ground motion prediction models for various seismogenic regions, calibrated by compilations of near source ground motions measured for small magnitude events. Characterization is also being performed to understand shallow, near surface properties to define potential site amplifications factors. However, operating company presentations illustrated that ground motion models have become an important screening tools for future operations. The CSUR workshop also included a background seismicity investigation prior to the onset of operations and a critical review of some methods to quantify the maximum expected magnitude. Jeremy Boak, Director of the Oklahoma Geological Survey gave a keynote describing recent induced seismicity activity in Oklahoma, including recent reduction in the earthquake rate attributed to a combination of operational slowdown during the downturn and regulated reductions in injection rates.

Canadian efforts also have evolved a unique and effective tone which is apparent at all technical gatherings. Indeed, there appears to be a widely held international view of an effective stakeholder effort in Western Canada, likely fueled by the broad awareness of the larger hydraulic fracturing-induced events. Much of the cooperative Canadian efforts have been centered around Calgary, where the highly concentrated petroleum activities proves to be an ideal venue to tackle a multitude of technical issues. Many factors likely contribute to a successful stakeholder engagement, not the least of which is a cooperative effort between industry, researchers and regulators. Notably, regional focused operator’s groups have been sharing observations and experiences for the benefit of all, a sign of the community importance of mitigating induced seismicity, overriding the typical secretive tendency for operators to maintain competitive advantage. However, the industry has also been quick to recognize the problem. This single factor makes for a positive environment for proactive technical development and has enabled the stakeholders to get in front of the issue. Go Canada!

Regardless of the region, north or south of the border, seismicity rates are clearly declining. While declines may be partially associated with slowing of operations during the recent downturn, regulations and operator mitigation protocols are also becoming increasingly effective. Four key aspects have emerged providing a foundation for these improvements:

Collaboration: An effective relationship has been developed between stakeholders, enabling efficient knowledge transfer between industry, regulators, academics and government groups.

Integration: Improved cross-disciplinary efforts have emerged between seismology, geophysics, geology, reservoir and geomechanical engineering.

Communication: Open dialogues have been created between various stakeholders for improved understanding of factors controlling induced seismicity and associations mitigation options.

Transparency and Trust: Exchanges of data and information have accelerated, although data accessibility remains an ongoing challenge.

Nevertheless, significant challenges remain. Characterizing the seismic potential of an area remains elusive, resulting in an initially reactive approach once regional induced seismicity is identified. The Duvernay is a good case in point, when induced seismicity unexpectedly began in December 2013. Potential induced seismicity is now also being closely scrutinized south of the border in the Permian basin, where most current industry activity is concentrated. Characterization and identification of potentially seismogenic faults also remain elusive. Effective seismic forecasting techniques are required, including quantifying the largest magnitude that could be expected. In addition to these technical challenges, the biggest upcoming challenge is effective public outreach.

Looking back over the evolution of induced seismicity, the issue does appear to be entering a new phase. General awareness of induced seismicity, the mechanisms and mitigation tools and techniques are now commonplace. Best practices have emerged and operators have been dealing with regulations for a few years. Particularly in Western Canada, a cooperative, integrated approach has emerged across stakeholder groups, with evidence of successful mitigation outcomes beginning to emerge. More importantly, technical challenges have become clear, a starting point for developing the solutions. Churchill’s famous quote does seem to be particularly relevant: “Now this is not the end. It is not even the beginning of the end. But it is, perhaps, the end of the beginning.” It will be interesting to see what comes after the end of the beginning, which will likely soon be evident as preparations are just getting underway for an international induced seismicity workshop scheduled for Banff in October 2018.