A recently published study of Oklahoma earthquake activity contradicts some popular beliefs about man-made seismic events.
The paper, authored by AAPG members F. Rall Walsh III and award-winning geoscientist Mark Zoback, appeared in June in “Science Advances,” an open-access journal published by the American Association for the Advancement of Science.
Walsh is a fifth-year doctoral student at Stanford University. Zoback is professor of geophysics in Stanford’s School of Earth, Energy and Environmental Science, and the 2015 recipient of AAPG’s Robert R. Berg Outstanding Research Award.
Walsh will discuss their study in the presentation “Oklahoma’s Recent Earthquakes and Saltwater Disposal” at the AAPG Mid-Continent Section Meeting, Oct. 4-6 in Tulsa.
Their research findings counter some widely held beliefs about Oklahoma’s recent history of quakes.
Belief: Earthquakes originate at such great depth, they can’t be caused by wastewater injection.
Finding: Saltwater disposal in Oklahoma, primarily injection into the Arbuckle Group, increases pore pressure, spreads away from injection wells over time and eventually triggers slip on critically stressed faults in basement.
“Faults that are mechanically active today are hydrologically conductive,” Walsh noted.
According to the study, there’s no doubt that wastewater from oil and gas activities is directly tied to Oklahoma’s upsurge in earthquakes.
Belief: Oklahoma’s induced earthquakes are mainly caused by the results of hydraulic fracturing.
Finding: Flow-back water from hydrofracturing contributes only a fraction of Oklahoma’s injected wastewater. Almost all of the injected water in the earthquake study areas was produced water.
This is a relatively new idea - that the great majority of injected saline water in Oklahoma is produced water.
Belief: The small amount of energy used in hydraulic fracturing and other completion and production operations can’t account for the huge amount of energy released by an earthquake.
Finding: Walsh and Zoback make it clear that increased pore pressure is simply a trigger for the release of already-existing forces.
“In the context of induced seismicity, the largest earthquake that might be triggered is determined by pre-existing geologic conditions, not the magnitude of the perturbation of pore pressure,” they wrote.
Belief: Induced earthquakes can be mitigated or stopped by reducing injected volumes in the nearby injection well or wells.
Finding: “It is likely that even if injection from many wells were to stop immediately, seismicity would continue as pressure continues to spread out from past injection,” Walsh and Zoback wrote.
Oklahoma has a very large number of wastewater injection wells, often closely clustered. Because increased pore pressure spreads away from injection wells over time, it can be difficult to determine which well or wells have triggered events. The effects could be cumulative.
“One of the features of Oklahoma is that often we can’t identify the ‘one well.’ You might have 100 or 200 wells within an hour or an hour-and-a-half’s drive of each other,” Walsh said.
Where the Action Is
The Stanford study of induced seismicity in Oklahoma took more than two years to complete, according to Walsh. He said it began with an examination of microseismicity - intentionally caused small quakes like those resulting from hydraulic fracturing.
“That was sort of our jumping off point. We started looking at Oklahoma simply because that’s where the most earthquakes are,” he explained.
Oklahoma averaged fewer than two earthquakes of magnitude 3 (M3) or greater per year between the beginning of modern seismic recording in 1974 and 2008. By 2013, the number of M3 or larger quakes had increased to more than 100, and by 2014 to more than 500.
Six areas of the state were selected for the Stanford study, each 5,000 square kilometers and each seismically active. Just three of those areas included 71 percent of the greater than M3 earthquakes ever recorded in Oklahoma.
Getting good data for the study proved challenging, Walsh noted. Some injection wells were duplicated in state reports, some weren’t reported and some had obvious entry errors. Lags in the reporting system made data timeliness less than ideal.
“There’s a delay between when the injection happens and when it’s aggregated and reported,” Walsh said. “That was one of the bigger issues we brought up.”
Fortunately, the percentage of errors was small and the study report said fewer than 100 monthly injected volumes, out of 1.5 million, had to be adjusted.
Walsh said the data corrections might not have been perfect, “but we convinced ourselves that we did it well enough to clearly see the overall picture.”
And when they analyzed the injection well data and locations in conjunction with the earthquake data, the connection between the two was obvious.
“As soon as we plotted the right data,” Walsh said, “it jumped out at us.”
Water, Water Everywhere
Under the assumptions in the study, water used in hydrofracturing in the six areas accounted for a small portion of the injected saltwater. Total injected flowback approached 20 percent in only one of the study areas.
Oklahoma includes several plays that produce significant water, including an extensive dewatering project in the Hunton Reservoir, Walsh noted.
“There are plays like the Mississippi Lime where there’s high water cut from the beginning because you’re producing from a paleo reef,” he said. “In the Hunton play, dewatering is an entirely different process from hydraulic fracturing.”
While just a subset of faults in crystalline basement are potentially active, the Oklahoma study areas include stressed faults that have a history of non-negligible seismicity, Walsh and Zoback found.
“In the context of critically stressed crust, small perturbations of fluid pressure have the potential to initiate slip on pre-existing faults that are already close to frictional failure. The stresses on the faults are the result of natural geologic processes – the same process that results in naturally occurring seismicity in other intraplate areas,” they wrote.
“We understand that the size of the earthquake is going to depend on the size of the fault and the stresses on the fault, which depend on its slip history,” Walsh said.
Further studies could help pinpoint the specifics of induced seismicity in Oklahoma, according to Walsh. He said geoscientists and petroleum engineers have reached a high degree of expertise in generating relevant models.
“The limiting factor isn’t their ability and knowledge,” he said. “Really, the limiting factor is having the right kind of data.”
Shared Responsibility
Researchers at Stanford are working to identify the critically stressed active faults in Oklahoma, “to empower the operators,” Walsh said.
He believes future earthquake activity could be minimized without halting injection operations - for instance, by injecting wastewater back into producing formations instead of the Arbuckle.
“One of the things we have said is that injecting into the producing formation is one of the biggest no-brainer ways to continue the injections,” Walsh said.
But the study findings aren’t comforting to those who believe Oklahoma’s earthquake activity can be stopped simply by reducing injections at a few wells. Surrounding or nearby wells also might be contributing to the stress perturbations.
“Now,” Walsh said, “you need to worry about what the people around you are doing.”