Three years after hydraulic fracturing for natural gas – also known as fracking – began around the central Alberta community of Fox Creek, the number of earthquakes in the region suddenly began to rise, including a magnitude-4.8 tremor that rattled the town in January, 2016, and forced the shutdown of a nearby gas well.
Now a scientific study of what happened at Fox Creek is shedding light on the connection between fracking and earthquakes with results that could help industry better control the seismic activity that fracking creates.
In the study, published Thursday in the journal Science, a Canadian team of researchers compared publicly available data on gas wells operating in the vicinity of Fox Creek with the earthquake frequency recorded by seismometers in the area.
Among the factors they examined, only one – the amount of fluid injected into the ground – seemed to make a difference to the number of earthquakes that were recorded around a given well site.
"Wells that have a higher volume [of fluid] tend to have earthquakes associated with them," said Ryan Schultz, a seismologist with the Alberta Geological Survey and lead author of the study.
While not all gas plays are seismically active, the findings suggest that well operators can reduce the risk of triggering quakes by restricting how much fluid they use in places where geological factors make earthquakes more likely. Those factors include the presence and orientation of fault lines where stress has built up and has the potential to be released by the injection of fluid at high pressure into the surrounding rock.
The new study adds to a growing body of evidence that hints at the cause-and-effect relationship between fracking and earthquakes. Internationally, the issue has led to rising concerns over the seismic risks associated with fracking, a process of drilling horizontally into rock and then injecting a gritty fluid under pressure to force open small fissures and release gas reserves that were previously inaccessible to extraction.
If a fracking operation is one that triggers earthquakes, "the rate of earthquakes is critical," said Gail Atkinson, a professor of Earth sciences at Western University in London, Ontario, who also contributed to the study.
In general, most earthquakes induced by fracking operations are too weak to be felt. But when earthquakes are more frequent, there is also a greater chance of producing a seismic event that can cause damage on the surface. (In general earthquakes start to be felt around magnitude-four but the possibility of damage depends on local conditions.)
"Roughly for every 100 magnitude-three earthquakes, you will get 10 magnitude-four earthquakes and one magnitude-five earthquake, and so on," Dr. Atkinson said.
The 4.8 earthquake that occurred near Fox Creek two years ago "got a lot of attention," said William Ellsworth, an expert in so-called induced earthquakes at Stanford University who was not involved in the study.
Dr. Ellsworth said the new study has practical value because it supports "traffic light" measures already in place in most states and provinces where fracking is permitted. Under such measures, a well goes from a green light to a yellow light if an earthquake occurs within a certain magnitude range (between 2.0 and 4.0 in Alberta). If the higher number is exceeded, the well hits a red light and a stop order is issued.
The challenge is that it is still not clear which wells will produce earthquakes ahead of time, Dr. Ellsworth added. "This has to be calibrated on a case-by-case basis."
Once earthquakes begin occurring around a well, however, the study offers some ability to predict how many more earthquakes might be reasonably expected if injected fluid volumes are increased.
"What all this means is that industry can control how much seismicity they trigger by controlling how much volume of fluid they inject," Dr. Atkinson said.
Mr. Schultz added that the relationship between volume and earthquake frequency explains the mystery of why it took three years of fracking before earthquakes began spiking around Fox Creek. Apparently the injected fluid had to reach a threshold volume before it was activating fault lines.
Another co-author, Honn Kao, a seismologist with Natural Resources Canada, based in Victoria, B.C., said the team's conclusion is in agreement with earlier findings from a gas play in northern British Columbia, and shows the link holds up with the much more extensive data that were available from the Alberta site.
"The result certainly becomes more convincing," he said.