HUTCHINSON, Kan. (AP) — For at the last 15 months, Kansas geologist Rex Buchanan estimates, he's spent 90 percent of his time studying something once relatively rare in the state - earthquakes.
He has learned a lot, said the director of the Kansas Geological Survey, whose focus in the past was primarily on water and geologic formations, including just how little is known, The Hutchinson News reports.
One thing that has become clear, Buchanan said, is that there is a correlation between the majority of the 120-plus earthquakes recorded in Kansas last year and the injection of wastewater into deep disposal wells.
"What's very difficult is to point to a specific earthquake and say a specific disposal well caused that," he said. "So, how to mitigate the earthquakes is what we're struggling to resolve."
There has been consensus within the industry "since the '60s, if not before," Buchanan said, that pumping fluids into deep underground wells used to dispose of wastewater, also called waste injection wells, can trigger an earthquake.
While disposal wells have been used by oil producers for more than 50 years without producing earthquakes, there has been an explosion in the number of those wells - and the amount of fluid being pumped underground - with expansion of hydraulic fracturing, or "fracking," for oil and gas recovery.
The fracking boom followed the recent discovery that it's an effective method for tapping previously unrecoverable oil trapped in geologic shale formations found in certain U.S. locations, including northern Oklahoma and south central Kansas.
In the fracking process, each production well requires between 1 million and 4 million gallons of water to fracture the rock and remove the oil or gas. The process contaminates the water, so it must be disposed of.
"It's really easy to look at an earthquake swarm or even a single one, see a disposal well and say, 'That's the problem,' " Buchanan said. "But it may not be. It may be a low-pressure disposal well that's been putting water in the ground a long time. So you may go to the next higher volume well and say, 'That's it.' Proximity is a way to go about it, but it's not clearly the most efficacious way."
"It's not a simple thing," Buchanan said.
Movement along faults or fractures in the subsurface of the Earth is what causes earthquakes, whether natural or induced. A fault must be present for there to be an earthquake, Buchanan said.
Officials believe they know where some faults are in Kansas and Oklahoma, but not most.
"You can typically see faults from 3-D seismic data, which companies do as part of the exploration process in some cases," Buchanan said.
Seismic crews map the Earth with "thumper trucks" that shake the ground and record the reflected sound.
"There are places the seismic data is available and places it's not," Buchanan said. "It's proprietary information, held by the company. It's confidential, as part of the exploration process. To get data from just one place and interpret it is very difficult."
Making that data available to scientists studying earthquakes, or to regulators who approve injection well permits, Buchanan said, "is an area under conversation."
Under current Kansas Corporation Commission regulations, there is an attempt to identify geographic faults "and other phenomena" before an injection well is permitted, according to the KCC website. Officials admit, however, that little public data is available about faults.
The KCC has reached out to "several large acreage holders who may have (3D seismic data identifying potential faults)," said Ryan Hoffman, conservation division director with the KCC. "In some cases it's third-party information and they may not be able to provide it, based on contractual terms. In several instances where we've reached out they have been able to share after removing proprietary data that might be used by competitors, to be used strictly by the KGS to make a map. We've had positive discussions with operations in that regard."
Whether the KCC or KGS has legal authority to get seismic data any other way, Hoffman said, "gets into legal issues (he's) . not qualified to talk about."
Pressure from the water pumped underground is how a disposal well can trigger a quake.
The water doesn't "lubricate the fault," said Gail Atkinson, professor and Research Chair in Earthquake Hazards and Ground Motions at Western University in London, Ontario, but increases pressure on surrounding rock.
"The pressure from the water makes its way to the fault which is already sitting there, and enough stress builds up to make it slip," she said. "The more fluid you put down, the larger the area potentially that you can see a change in the state of stress on existing faults."
Atkinson compared it to hydraulics in a vehicle.
"If you put your foot on the brake pedal, the hydraulic pressure activates the brakes," she said.
Buchanan used the analogy of an air hockey table. When the air is off, the puck doesn't move, but turning the air on releases the friction between puck and table and it slides easily.
The proximity of a fault to an injection well isn't necessarily the determinant whether it will trigger an earthquake. The orientation of the fault line, how the well pressure engages it and properties of the surrounding subsurface rock also determine its potential for inducing a quake, Buchanan said.
Other issues that may come into play include the injection pressures, the duration of injections and the amount of water going in. Chemicals in the water that are related to hydraulic fracturing themselves don't appear to play a role in causing the quakes, Buchanan said.
According to several studies, it appears higher-pressure injection - putting more water down faster - is more likely to trigger an earthquake than longer-term, low-pressure injection.
U.S. Geological Survey scientist Art McGarr, of the Earthquake Science Center in California, recently reported that the amount of fluid may affect an earthquake's strength, while its injection rates appear to control the frequency of occurrence.
More fluids can result in stronger quakes.
Over time, Western University's Atkinson said, the "pressure front" will move further and further out.
"There are places in Colorado, in Paradise Valley, where we've seen (earthquakes) 16 kilometers away from where they injected fluid. As time goes on and they put more and more water in, it pushes the pressure out further and further. We know it can push it as many as 20 kilometers, and nobody knows if that's the limit."
Katie Keranen, a former University of Oklahoma professor now with Cornell University, reported at a Seismological Society of America meeting, that seismic and hydrology data showed "a strong link between a small number of wells and earthquakes migrating up to 50 kilometers (31 miles) away."
One suggested method to respond to induced earthquakes is to slow injection if a well is suspected of causing seismicity, and to stop pumping if quakes don't significantly slow or stop.
Atkinson notes, however, that a pressure front can continue to diffuse for a long period of time, even after injection stops.
Officials believe injection that began 18 years earlier triggered a 5.7 magnitude earthquake in Prague, Oklahoma, Atkinson said.
"Over a period of time, the pressure front moved out and out and eventually it got to a place where it happened to trigger a fault."
Induced earthquakes generally have smaller magnitudes than natural quakes, which occur deep in the Earth's crust, Atkinson said, but they can still damage structures.
And it doesn't take a big fault to cause a magnitude 4 or 5 quake, Atkinson said.
"You're only talking about a kilometer" in length, she said.
"To date, the largest (documented) from fluid injection was the magnitude 5.7," she said. "We don't usually think you get a lot of damage from a magnitude 4 quake. But these events are very close to the surface, only a few kilometers deep. Other quakes are deeper, 10 kilometers or so."
Quakes that occur relatively close to the surface, while usually shorter-lived, can result in more intense ground motion.
"If you look at a typical earthquake in California, it's eight kilometers (about five miles) underground," she said. "If you induce an earthquake at two kilometers (1.2 miles) below the surface, ground motion is going to be a lot stronger."
There is more intense shaking right over the event, Buchanan said, but it tends "to die out more quickly."
"In theory it could cause more damage locally," he said, than from a deep quake.
"What protects us from earthquake damage is building codes," Atkinson said. "Building codes in most parts of the country have some level of seismic resistance in them. The key to the event is shaking. If you're building in California, you have to build to withstand significant earthquake forces. Building in Kansas or Oklahoma, the amount of seismic design is minimal. You don't expect a lot of ground shaking."
Most buildings in the rural areas of Kansas where quakes are occurring are older, making them more susceptible to damage. An older brick building, Atkinson said, such as a warehouse, is at much more risk than wood frame structures, "which do pretty well, though you could still get a lot of damage to a chimney."
One reason it has been difficult to tie seismic activity to specific injection wells is lack of precise information about the quakes.
If there are lots of monitors - and they're closer to the quake - the epicenter of the quake can be located within meters, Atkinson said.
But Kansas has few seismic monitoring stations.
"The earthquake info you guys see is from the USGS National Earthquake Information Center in Golden," Buchanan said. "It's based on a national network. The three stations in Kansas don't allow us to get real good location accurately. The epicenters it puts out you may be able to move 4 to 5 miles. There's a lot of uncertainty."
Over the past month, the state has purchased and located four temporary monitoring stations in south central Kansas, and plans are to set up two more.
"The more monitoring is a good start," Buchanan said. "Really, it's the key. Ultimately we have to figure this out and we can't figure it out without monitoring. It's a good first step. What to do with that information is the next question."
"If you look at it, there is a low chance of any individual well inducing seismicity," Atkinson said. "There are hundreds of thousands of them, and only a small percentage is inducing. We don't know why some are and others are not. But if there are 100,000 of these wells, it only takes a small percentage to generate a lot of earthquakes. If you get 100, that's a small percentage."
The rising number of manmade earthquakes, however, could pose a risk to critical infrastructure, such as dams and nuclear power plants, Atkinson suggested. The danger is still really unknown.
"It's like buying a lottery ticket," Atkinson said. "If you buy a lot, your odds of winning go up."