Expert opinions divided on human fault for 2011 earthquakes
Two OU seismologists have different ideas about what will benefit the public when it comes to deciding if earthquakes are caused by activity in an oilfield.
A debate has sprung from a recently released report that concluded the earthquakes that shook Oklahoma and surrounding states in 2011 likely were caused by oil wells injecting fluids into the earth.
Katie Keranen, lead author of the report and OU geology professor, said it’s important to know if oil wells are causing earthquakes so the oil well operators can change their practices.
However, Austin Holland, seismologist for the Oklahoma Geological Survey, said assuming an earthquake didn’t result from natural causes could prevent seismologists from using it to predict future earthquakes.
“One of the missions, and the reason why I’m working as a seismologist here in Oklahoma, is we want to provide for public safety,” Holland said.
The U.S. Geological Survey’s Earthquake Hazard Program includes a catalog with data on naturally occurring earthquakes seismologists use to calculate when and where another earthquake might occur in the future, Holland said.
However, if an earthquake is identified as induced and not natural, then it’s removed from that catalog and not included in the calculations, he said.
“So misidentifying an earthquake can actually make people less safe than they were before, if it was considered a natural earthquake,” he said.
Four days before Keranen’s study was published, the Oklahoma Geological Survey published a statement online saying the earthquakes examined in the study were most likely the result of natural causes.
“We knew [Keranen’s report] was coming out, so we had the statement ready and posted on our website,” Holland said.
Three earthquakes and a series of aftershocks occurred on Nov. 5, 2011, near Prague, Okla., about 44 miles northeast of Oklahoma City, according to the U.S. Geological Survey.
The largest of the three earthquakes was magnitude 5.7 and was the largest that has been reported in Oklahoma, according to the survey.
The earthquakes were felt in 17 states and injured two people, destroyed 14 homes and buckled part of U.S. Highway 62, according to the survey.
When the first earthquake hit, Keranen and other seismologists rushed to the scene to measure the aftershocks and try to figure out where the ground was moving, she said.
From the data they collected, they were able to identify where the earthquakes had occurred and the depth of the faults, or sites where the rocks had slipped and caused the earthquakes, Keranen said.
Once they figured out the location and depth of the faults, they found that the first earthquake occurred right next to where oil wells were injecting wastewater into the earth, she said.
When oil and gas are produced at an oil field, mostly water is produced with a little bit of oil and gas, Keranen said. The oil and gas are separated from the water, which is too salty to be drinkable, so it’s re-injected back into the rocks in the subsurface of the earth.
Keranen picked up a small rock in her office to demonstrate where the water goes when it’s injected back into the earth.
“You can’t really see all the holes [in the rock], but there are a lot of holes where water can go in,” she said. “So I can put this in water, and it would soak up a lot of the water.”
The water is supposed to diffuse away after it’s injected into the rocks, but if that water can’t get away, the volume and pressure in the rocks will start to rise, Keranen said.
“It’s basically like filling a balloon too full,” she said.
Keranen and the other seismologists working with her ended up finding barriers in the subsurface making it difficult for water to flow past and increasing pressure deep in the ground, she said.
“So we interpret that that increase in water and that increase in pressure triggered this earthquake,” she said.
“We see a really clear spatial correlation to the wells,” Keranen said. “They’re injecting very close to faults.”
However, a correlation isn’t always enough evidence to prove a cause, Holland said.
“There’s a perfect correlation between sale of ice cream in the United States — as sales go up, the number of drowning deaths go up,” Holland said. “And no one would assume that people are drowning because they’re consuming ice cream.”
There isn’t enough evidence in this case that pressure was increasing in the earth as a result of the injected wastewater, he said.
“Pressures may have risen within the earth, but they also may not have,” he said. “And there are ways that we can get that data and begin to look at that, and that’s what we’re doing here at the geological survey.”
But the pressure data the survey cites in its statement released a few days before Keranen's report was published are from pressure tests conducted after the earthquakes occurred, which wouldn’t be an accurate measurement because earthquakes themselves release pressure, she said.
“So these tests really aren’t relevant,” she said.
Keranen expected there to be some debate and discussion over the two interpretations, she said, but it’s something she welcomed.
“It’s really a healthy thing,” Keranen said.
However, seismologists from the Oklahoma Geological Survey have yet to show Keranen data backing up their statement, she said.
“To me, it’s really important that they back it up with the data so others can evaluate both sides,” she said.
Holland has data behind all the points in his statement, but a lot of questions remain, and they’re trying to figure out a way to bring it all together to make a coherent paper or maybe several papers, he said.
While Keranen and Holland disagree on whether Oklahoma’s 2011 earthquakes were caused by wastewater injection, both agree that oil and gas companies should provide more data so this issue can be examined adequately.
Despite the debate, both seismologists agree they need more data to provide accurate assessments in the future.
It’s difficult to be certain if these earthquakes were induced because we can’t look inside the earth, Holland said.
“How do you observe something that you can’t see?” Holland said.
Because of this, they have to make interpretations, and those are based on a limited set of data, he said. However, there is data that could be collected ahead of time, so if anything occurred, they would be able to answer these questions better.
The problem is that much of this data isn’t currently available to seismologists, Keranen said.
Seismologists could know if an earthquake had been induced if they knew how much pressure was increasing from the injected fluid, and they could understand this if they knew the rate at which the fluid was being injected, she said.
Operators at oil wells in Oklahoma give seismologists records of how much water they’re injecting over a month, but these records aren’t helpful because some days they inject a lot of fluid at one time, and other days they don’t inject any at all.
“It’s like if I ate 20 jellybeans in an hour, but I told you I ate 20 jellybeans in a month - it really isn’t the rate, because I ate them really quickly,” she said.
It also would be helpful if operators set up instruments to monitor the subsurface where they’re injecting fluids, so if earthquakes started to occur, they would know to start backing off, she said.
“It’s either going to be with new rules and regulations, new laws or just oil companies taking that responsibility amongst themselves,” Holland said about getting more data from well operators.
Keranen thinks that while the best interpretation based on data available is that these earthquakes were induced, she said; even if she’s wrong, she hopes that when people look at her study they will recognize there is a very strong spatial correlation between the earthquakes and the wells, so these companies should be more careful in the future.
“It can be done safely, so it’s not a matter of saying this process inherently is unsafe,” she said. “It’s a matter of saying there are a few cases where it is unsafe, and let’s see if we can find ways to catch those ones in advance and avoid building up pressure to cause an earthquake.”