| Energy | Dallas News
The problem with Oklahoma earthquakes is what scientists don't know, the director of the Oklahoma Geological Survey told a roomful of geoscientists in Dallas.
Jeremy Boak was one of the key presenters at a special session of the annual conference of the Society of Exploration Geophysicists meeting at the Kay Bailey Hutchison Convention Center. Others in the program offered new ways to predict significant human-induced quakes that make people nervous and new ways to analyze tiny micro-tremors that could help the oil and gas industry figure out where to drill next.
Boak's talk depended a lot less on advanced math than most of the others in that session.
"This will be a little more newsy and a bit less theoretical," he said.
Mostly, Boak offered a quick tour about what's known regarding induced quakes in Oklahoma.
They're closely related to deep injection wells that dispose of salty, toxic water that comes out of gas and oil wells. The quakes affect about 18 percent of the state. The rate of quakes has dropped significantly from the peak last June.
Only 5 percent of the injected water is backflow from fracks; Oklahoma has lots of oil and gas fields that produce much more wastewater than petro-products. (Although many of the newer wells that produce a high volume of water would never have been drilled except for fracking's success.)
Most of Oklahoma's quakes are small and centered in identified zones near the injection wells and particularly vulnerable faults far below the surface of the earth. But the biggest quakes that have hit the state, including the Oklahoma-record temblor last month, have been at the far edges of the zone of highest concern, Boak said. And that gives him pause. Somehow, the pressure created by the injection wells is pushing faults that are relatively far away.
"Where is the fault we've not seen that's not yet been hit by this pressure pulse?" he said.
The September quake, measured at 5.8, triggered a rapid state and federal response, shutting down specific injection wells temporarily. But as researchers studied the data, they realized they'd been focused in the wrong place, Boak said. It looks like the September quake was caused by a fault in a place they'd not known about.
"It's an interesting testament to the fact that you can get into trouble by responding too quickly, as well as by responding too slowly," he said.
Carson Dinske from Freie Universitat Berlin was one of the other presenters. He offered a new way that mashes together elements from other models to predict quakes caused by injection. He said it works well taking old data and predicting quakes that happened in various parts of the world. But only recently has it been applied to a couple of areas in Germany to predict future events in real time. So far, he said, it's doing well.
Azra Tutuncu, a professor at the Colorado School of Mines who organized and chaired the session, was skeptical about how good the predictions would be. That's because she's seen the limited success of predictive models new and old.
"Everyone has lots of assumptions," she said.
This week's conference is the 86th annual meeting of the society. More than 5,500 attendees plus vendors from about 250 companies are spending four days discussing hundreds of highly technical aspects of the search for oil, gas and precious minerals and metals.
This is the third year Tutuncu has organized an "injection induced seismicity" session at the annual conference.
"It's somewhat of a new topic for most geophysicists," she said.
And while the science is still reaching for precision in its ability to predict how human activity will create earthquakes, it's moving in the right direction, she said.
"I think we are doing better," she said.
Image: National Ground Water Assoc