Read our COVID-19 research and news.

In 2011, a magnitude 5.7 earthquake near Prague, Oklahoma, damaged buildings—and was linked to wastewater injection from oil and gas operations.

In 2011, a magnitude 5.7 earthquake near Prague, Oklahoma, damaged buildings—and was linked to wastewater injection from oil and gas operations.


Are earthquakes triggered by oil and gas production becoming deadlier?

SAN JOSE, CALIFORNIA—Over the past several years, a torrent of small earthquakes has accompanied the glut of oil and gas produced by industrial operations across the central United States. In 2014, Oklahoma saw three times as many earthquakes magnitude 3.0 or greater than California. Hydraulic fracturing, or fracking, is not the main culprit. Rather, most of the small earthquakes have been linked to injection wells, which dispose of huge quantities of water used to flush out oil and gas in extraction operations. Here today at the annual meeting of AAAS (which publishes Science), Science had a chance to catch up with three experts working at the forefront of this field of induced seismicity: William Ellsworth, a geophysicist at the U.S. Geological Survey (USGS) in Menlo Park, California; Mark Zoback, a geophysicist at Stanford University in Palo Alto, California; and John Parrish, the California state geologist in Sacramento.

So far, most of the earthquakes associated with injection have been small—things that might rattle the china cabinet. What is the potential for larger, life threatening earthquakes?

Ellsworth: That’s really one of the hardest problems we have. It requires knowing where the faults are, and what their dimensions are. And knowing if they’re oriented properly, and if there is sufficient stress to drive them. We’re really stuck using what we’ve seen happen in the past. In Oklahoma, we know that there was an earthquake about magnitude 7, 1300 years ago. We also saw the magnitude 5.7 just in 2011. So things that have happened in the past should be considered in the future.

Zoback: We know the risk is posed by large faults which are potentially active in the current stress field. The challenge is, can we find those faults and avoid them. The thing the public has to remember is that not all faults are dangerous.

USGS is trying to incorporate induced earthquakes in its hazard maps. You’re saying that places like northern and central Oklahoma are becoming potentially as hazardous as places like the New Madrid region near Memphis, Tennessee, and Charleston, South Carolina, which have a history of highly damaging earthquakes?

Ellsworth: Over the short term—what’s likely tomorrow—those would be areas of high hazard. Whether they’re high over the timescales we think about for building codes is a whole other question. That I don’t think we have an answer to.

This could eventually force stricter building codes in places like Oklahoma. Is it fair for those additional costs to be born by builders when the earthquake risk is essentially a dial controlled by oil and gas industry?

Zoback: That’s a policy question. If you go back to building bridges and roads, trucks pay fees, because they’re hard on bridges and roads. How these things get paid for is a policy decision.

Parrish: It’s really going to be a local decision.

Sometimes you say there’s a strong link between certain wells and earthquakes. Other times you say there’s a lot of uncertainty in making that connection. What allows you to make a definitive diagnosis?

Ellsworth: If there’s one well, and there are earthquakes directly beneath that well … then you can make a pretty good connection as to what’s going on. If you have multiple wells, we know that the pressure field is going to be distributed, and it can be very far reaching.

Why does California see so many fewer induced earthquakes than places like Oklahoma?

Parrish: Our geology is quite a bit different from the geology in the midcontinent and the eastern states. For one thing, their production formations are very old. California, we have very poorly consolidated sediments that are much younger.

Ellsworth: The change has been when water begins to be injected in virgin formations, places where no perturbations have happened before. That seems to be to me at least one of the signatures for situations where can say these earthquakes have been induced.

Aside from identifying problematic faults, what other techniques are there that industry can adopt to minimize induced earthquakes?

Zoback: Let’s take Youngstown, Ohio. It’s right on the border with Pennsylvania. What was happening in Pennsylvania is that the flowback water after hydraulic fracturing was being trucked to Ohio so that it could be injected into … permitted injection wells where the geology was better. They’ve migrated from trucking water out, to recycling the water.

Are these good times to be an earthquake scientist?

Zoback: It’s always a good time.

Ellsworth: We have new challenges, which is good.

Parrish: I like the saying: Love a geologist, and feel the earthquake.

Check out our full coverage of the AAAS annual meeting.

What message would you send into space? Tell us on Twitter and Vine with #msgtospace!