University of Oklahoma graduate students near Wellston, Okla., installing a seismometer to study central-Oklahoma's earthquake swarm
Joe Wertz / StateImpact Oklahoma
University of Oklahoma graduate students near Wellston, Okla., installing a seismometer to study central-Oklahoma's earthquake swarm
Joe Wertz / StateImpact Oklahoma
A series of peer-reviewed papers published June 2 in The Leading Edge detail new research on disposal well-triggered earthquakes in Oklahoma.
The June 2015 issue of the journal of the Society of Exploration Geophysicists includes a special section devoted to the manmade quakes — a phenomenon known as “induced seismicity.” The research adds to a growing chorus of scientific evidence linking wastewater injection wells to atypical earthquakes recorded around the world — most notably, and dramatically, here in Oklahoma.
The full issue contains papers on likely induced quakes recorded in California, Kansas, Ohio, and countries like Canada and the Netherlands, but my primary interest, as always, is on Oklahoma.
The new papers have been added to our Oklahoma Earthquake Research Reading List, but here are eight quick — and key — takeaways from the new research.
For clarity, I removed some of the parenthetical citations to make it easier to read some of the selected and quoted selections.
As of early-2015, Oklahoma’s earthquake rate was continuing at “unprecedented levels,” lead author William Ellsworth of the U.S. Geological Survey and six other scientists write in one paper, “Increasing Seismicity in the U. S. Midcontinent: Implications for Earthquake Hazard.”
In just the month of January 2015, residents of Oklahoma reported feeling more than 130 earthquakes, including 59 with magnitudes between 3.0 and 4.3.
The 5.7-magnitude earthquake that struck near Prague, Okla., in November 2011 injured two people, destroyed more than a dozen homes and caused damage that Ellsworth’s team estimates totaled around $10 million, but the scientists note:
… the main effect of most of the hundreds of felt earthquakes has been the nuisance of unexpected shaking, fraying nerves, and anxiety over the unknown potential for stronger shaking in the future.
And, using data obtained from the USGS’s “Did You Feel It Program,” Susan Hough suggests that ground shaking from artificial earthquakes could be less intense than movement triggered by naturally occurring quakes. From Hough’s paper, “Shaking Intensity From Injection-Induced Versus Tectonic Earthquakes in the Central-eastern United States”:
Although moderate injection-induced earthquakes in the central and eastern United States will be felt widely because of low regional attenuation, the damage from earthquakes induced by injection will be more localized to the event epicenters than damage from tectonic earthquake because of lower stress drops of induced events. Regardless of the interpretation, a growing body of well-constrained DYFI data provides prima facie evidence that shaking from injection- induced earthquakes is significantly lower at regional distances than shaking from tectonic earthquakes in the same region.
About 3,200 disposal wells are in active use throughout Oklahoma, but a very small number of those wells are likely triggering earthquakes, Ellsworth writes in the “Implications” paper:
If disposal of wastewater by injection is the principal cause of the excess seismicity, as now appears almost certain, it nonetheless needs to be stated clearly that disposal of wastewater by injection in UIC Class II wells more often than not results in no detectable seismic response. Consequently, the existence of a well has low predictive power for seismicity by itself.
From Ellsworth’s “Implications” paper:
The magnitude of the earthquake rate change in the western area after 2008 precludes a random fluctuation about the historic mean as a viable explanation.
How do Ellsworth and the other scientists know that Oklahoma isn’t experiencing natural earthquakes? Because natural “tectonic” earthquake tend to change the shape of the Earth’s crust. In Oklahoma, the Earth’s crust doesn’t seem to be changing:
To test this possibility, we examined the crustal-deformation rate in Oklahoma by using continuous GPS data from 2002 to the end of 2014 … Across Oklahoma, GPS data contain no evidence for contemporary crustal deformation.
And other factors aren’t the main reason for the earthquakes either, Ellsworth and Co. write:
Several alternatives to the fluid-injection hypothesis, including drought, recharge of reservoirs as drought eases, and natural variations in earthquake activity, have been put forward as explanations for the increased seismicity in Oklahoma and Texas, although none that we are aware of appears in the published literature …
… We found no evidence for deformation anywhere in the state that would increase the stress driving the contemporary seismicity and concluded that there is no evidence for a natural tectonic origin of the increased seismicity.
In a paper by lead author Daniel McNamara of the USGS and 13 other scientists called “Efforts to Monitor and Characterize the Recent Increasing Seismicity in Central Oklahoma,” researchers conclude that most of Oklahoma’s recent earthquakes occurred on particularly risky “optimally oriented” faults. Those faults are …
… near-vertical, optimally oriented (northeast-southwest and northwest-southeast) strike-slip faults in the shallow crystalline basement. In 2014, 17 earthquakes occurred with magnitudes of 4 or larger. It is suggested that these recently reactivated fault systems pose the greatest potential hazard to the region.
“Shallow crystalline basement,” also called “crystalline granite” or simply “basement rock” is important because it directly involves disposal wells in Oklahoma. Wells that allow waste fluid to come into contact with the crystalline basement are considered especially risky, which is why state oil and gas regulators are cracking down on them.
The USGS assembles 50-year earthquake forecasts for people like building designers, public safety planners and insurance companies. But induced earthquakes don’t fit that long-term model, Ellsworth and Co write in the “Implications” paper.
But, scientists might be able to build shorter-term forecasts for artificial earthquakes, and that could be helpful:
We might be able to make viable forecasts of tomorrow’s earthquakes based on the past week, month, or year …
… more than 150 participants representing petroleum producers; geophysical service providers; university, private-industry, and government researchers; state geological surveys; state and federal regulators; the reinsurance industry; and users of hazard models, including local government, state departments of transportation, and state architects. There was broad agreement that a one-year forecast would have value
While many energy companies and their industry representatives now acknowledge there is at least some link between disposal wells and earthquakes, historically, some in the industry have denied the connection.
Some have suggested other factors, like drought, could be the culprit, but there aren’t any peer-reviewed studies to back that up. And the energy industry didn’t submit any research to The Living Edge issue on the topic, Robert Habiger and Gregory Beroza write in the section’s intro:
Although interest among operators in the topic is high, we did not receive submissions from oil and gas operators to this special section. Nevertheless, cooperation among operators, regulators, and affected local government and communities will be essential to manage the range of issues surrounding induced seismicity.