Except the wells can cause earthquakes. In some cases, the quakes begin as far as 15 miles from the wells. In a new study in the journal Science, scientists describe for the first time how earthquakes can be triggered so far away from the wells. An efficient practice by the oil and gas industry is creating a ripple effect far beyond its drilling locations.
Geologists have linked injection wells to quakes, with findings based on years of observation. Human-made earthquakes, though most are moderate in size, put 1 in 50 people in the United States at risk, according to a recent U.S. Geological Survey analysis. Wastewater injection wells are concentrated in Oklahoma, Texas, California and Kansas, according to the Environmental Protection Agency.
“Induced earthquakes are becoming more and more of an issue in central and the eastern U.S.,” said University of California at Santa Cruz seismologist Thomas Goebel. In 2011, an injection well in Oklahoma was responsible for a magnitude-5.6 earthquake that damaged a highway, shook buildings and generated a dozen aftershocks.
To figure out how there could be such a distance between well and earthquake, Goebel, along with fellow UC-Santa Cruz earthquake expert Emily Brodsky, sifted through quakes triggered by dozens of waste injection sites in several states as well as in Australia and Europe. (There are so many wells in Oklahoma they could not link an individual well to the surrounding earthquakes.)
Industrial techniques such as hydraulic fracturing, or fracking, shove water underground to force oil and gas out of shale deposits. Most induced earthquakes are not a result of fracking itself but wastewater generated at the oil and gas wells. Some of that water can be reused or treated. The rest is buried in wells.
Earthquakes occur when a crack underground — a fault — pulls apart. A few decades ago, when scientists were beginning to understand that humans could generate earthquakes, the idea was “you put water directly into the fault,” Brodsky said. It was assumed water would pry apart the fault, like a hydraulic jack lifting a car, triggering a quake.
But that theory could not explain the quakes that happen miles from the wells.
The study authors were able to identify two types of earthquakes triggered by wastewater wells, having everything to do with what kind of rock the water is being injected into.
One kind of earthquake formed close to the injection well but stopped abruptly at about a half-mile from the site, Goebel said. If a well dumped its wastewater into rigid bedrock, earthquakes occurred within a close distance. There, pressure from water that spilled into a fault triggered the earthquake.
The other kind had a “very long-distance tail” — the quakes could appear far from the well, with the triggers petering out only after several miles. This occurred if a well dumped its wastewater into softer sedimentary rock. This was a result of what the researchers called "poro-elasticity."
Unlike solid bedrock, sedimentary rocks have lots of holes, like a sponge. Because sedimentary rock is more permeable than bedrock, it makes sense to dispose of fluid there — more holes mean more space for wastewater.
But the new study suggests energy companies are injecting waste into the wrong place to avoid earthquakes. Sedimentary rocks are not completely rigid. They’re squishy. They deform. Wastewater might not shove open a fault in the squishy rocks, Brodsky said, but as the ground fills with water “it also pushes on the surrounding rocks.”
Goebel likened it to stepping on a latex balloon sitting in a cardboard box. The balloon bulges outward, and as it does, it presses against the walls of the box. Likewise, as the rock bulges, it can nudge faults far from the injection well. The result: seismic action at a distance.
MIT earth scientist Bradford Hager, who described this research as a “really good empirical study,” said this report convincingly described the two ways humans trigger earthquakes, through pressure or poro-elasticity in rocks. But he was less convinced that the roles of sedimentary rock and bedrock could be so neatly divided. In some cases, pressure might trigger earthquakes in sedimentary rock, whereas poro-elasticity could play a role in bedrock.
Still, he said, this kind of work is begging to find its way into “regulatory behaviors.”
Brodsky anticipates that “there will be some resistance to this” research. “Multibillion-dollar industries are, you know, not rapid to change,” she said.
Meanwhile, Goebel said, the scientists are running small-scale laboratory experiments to further examine how the earthquake triggering mechanism works.