There is "clear evidence" that gas drilling activities triggered the magnitude-5.3 earthquake that shook Colorado in August 2011, a U.S. Geological Survey study has found.
That rupture was the largest quake in a 13-year pattern of shaking along the New Mexico border, which the study links to disposal of waste water from coalbed methane production.
"The earthquakes are clustered around wells that have been quite active since about a year before the earthquakes started," said Art McGarr, one of four USGS scientists who worked on the study.
The peer-reviewed study, published today in the online version of the Bulletin of the Seismological Society of America, says that there's been a big increase in earthquakes in the Raton Basin around Trinidad, Colo., and the only other thing that has changed has been the arrival of new disposal wells.
But one of two companies producing gas in the area sharply disagrees.
"We would categorically disagree," said Jennifer Webster, spokeswoman for Irving, Texas-based Pioneer Natural Resources Co. "We're not seeing any connection with disposal activity in the area."
The area has a history of natural earthquakes, she noted. And for the past 18 months, Pioneer has been monitoring a sensitive array of 25 instruments in the area. Webster said the seismicity they detected is far from the injection wells and 2 miles deeper than the injection zone.
The other operator in the area, Atlas Resource Partners of Philadelphia, did not return a phone message seeking comment.
The two companies are not engaged in high-volume hydraulic fracturing of shale formations. Instead, they are tapping into the natural gas found in coal formations of the Raton Basin. Coal beds are found much shallower than shale, but producing from them also creates significant amounts of wastewater. The Raton field, though, is in decline.
As with other studies from Arkansas, Oklahoma and Texas, the USGS study links the quakes to disposal of waste fluid.
From 1972 through July 2001, there was one quake in the area larger than magnitude 4. Then 12 occurred between August 2001 and 2013, mostly within 3 miles of active disposal wells. The study says the statistical likelihood that such a rate change would occur if earthquakes behaved randomly in time is 3 percent. In addition, earthquake activity remains low outside the drilling zone.
As soon as the shaking started in 2001, seismologists were suspicious that the convulsions in the Raton Basin were linked to drilling. But for years, USGS scientists were equivocal. McGarr said by the time of the magnitude-5.3 earthquake, the relationship between injection and earthquakes was a lot more conclusive.
Another look at Colo. quakes
Prompted by the magnitude-5.3 quake, which occurred the same day as a better-known magnitude-5.8 quake that shook Virginia and the East Coast, USGS re-examined the Colorado earthquakes going back to the "swarm" of 2001. Seismologists at the agency put out new instruments, went back into the data they had gathered in the past 10 years and began to point the finger at drilling activity with increasing certainty.
The series of foreshocks and aftershocks in 2011 was centered within 6 miles of five injection wells in the Raton Basin, the study says, two owned by Atlas Resource Partners on the same site and three owned by Pioneer. All but one, the study says, are "high-injection-rate, high-volume wells." The ARP wells are within 1.7 miles of where the 2011 sequence began.
"The proximity of the [ARP] wells to the 2011 earthquake sequence also suggests that they are the wells most likely to have induced the earthquake sequence," the study says.
Colorado officials, though, have long rejected the USGS conclusions as premature. In 2012, then-Colorado State Geologist Vince Matthews said, "These cowboys from USGS are sure these are induced. They're jumping to conclusions" (EnergyWire, Dec. 3, 2012).
State officials, under fire from suburbanites and environmentalists charging lax regulation of drilling, have been taking a less dismissive tone on the issue in recent months. After a smaller, magnitude-3.2 quake near Greeley in May, the Colorado Oil and Gas Conservation Commission asked the operator of a nearby deep-injection well to temporarily shut down for 20 days. In July, the state allowed the well to reopen at a lower pressure and less injection than before. The state said the well was "potentially" related to earthquakes in the area.
Colorado has a long history with man-made quakes. In the 1960s, disposal wells drilled at the Rocky Mountain Arsenal near Denver, where the Army manufactured chemical weapons, were the first to be linked to earthquakes. The largest Rocky Mountain Arsenal earthquake was magnitude 4.85.
After that, the Bureau of Reclamation began tracking man-made quakes in a river desalination project in the Paradox Valley of western Colorado.
There was drilling and disposal in the Trinidad area from 1994 through July 2001, with no uptick in earthquakes. But in early 2001, the study says, injection rates in the Colorado portion of the field dramatically increased, rising from a median rate of 500,000 barrels a month to 1.2 million barrels a month. The earliest earthquakes were located in the eastern portion of the gas field, shortly after six wastewater injection wells were put into operation.
"Total injection volumes and the number of earthquakes roughly track each other," the study says.