Question: If the science of predicting the risks and mitigation of induced seismicity, figuratively speaking, was a glass of hydraulic fracturing injection fluid, would it be half full or half empty?
The answer is . . . yes.
One on hand, geoscientists are getting better about predicting the when and where of seismicity; on the other, there is a lot more seismicity in a lot more places that have to be predicted.
How much more?
Worldwide, in 2021, according to Statista, a German company that specializes in market and consumer data, there were three earthquakes with a magnitude over 8.0, which is the highest number since 2007. The number of magnitude 5-5.9 earthquakes also increased from 2021 with more than 2,047 recorded.
In the United States, according to a study by the independent energy research and business intelligence company Rystad Energy, tremors above the magnitude of 2 on the Richter scale have quadrupled since 2020 and are expected to increase if oil and gas activity sticks to its current drilling methods at the same pace.
“Our understanding of earthquakes in general and induced earthquakes in particular has advanced immensely,” said Jens-Erik Lund Snee, research geologist at the U.S. Geological Survey.
And this understanding includes the many behavioral similarities between induced and natural earthquakes.
Which is half the battle.
“Although subsurface datasets have improved markedly in the past decade or two, thanks in many cases to large research projects or contributions from industry partners, they still lack sufficient detail nearly everywhere,” he added.
The challenge, it seems, has always been the challenge.
“While we are often able to forecast which faults are more likely to experience earthquakes associated with industrial activities, we do not understand the real-time stresses, rock properties and details of fault systems well enough to predict when and where a specific earthquake will occur,” said Snee.
As the Rystad research indicates, this will become more of an issue as subsurface industrial activities increase in the coming decades, due in large part to oil, gas and geothermal energy development, energy storage and carbon sequestration.
“It is important that we better understand and reduce the hazards and risks associated with induced earthquakes,” Snee said.
In some places, scientists are doing just that due to better seismic monitoring methods and networks, more detailed fault, stress and rock properties datasets, and better understanding of subsurface fluid flow and earthquake physics. These have all improved operational practices, meaning new research tools that enable decision-makers to estimate earthquake hazard, risk and risk tolerance.
As an example, Snee pointed to the seismicity rates in Oklahoma and the Dallas–Fort Worth Metroplex, which have declined.
Unfortunately, in the Permian Basin, West Texas and southeast New Mexico, they have not.
As the technology has improved over the years, so, too, has the understanding – and the conversation.
“Whereas less than a decade ago, it was controversial whether much of the increase in earthquake rates could be attributed to saltwater disposal and especially hydraulic fracturing at all, we are now at the stage where we have a good general understanding of the mechanisms for induced seismicity,” Snee said.
At the USGS, a current topic of research is in how changes in fluid pressure may cause seismic deformation in the subsurface, how that deformation can load stresses on faults that can potentially cause later earthquakes, and ways these effects can be measured from the surface. Snee said that regulators and industry are increasingly able to anticipate hazards and modify operational practices to avoid problems.
“We have seen that induced seismic hazards can be drastically reduced in operational areas by means of detailed site characterization to identify potentially hazardous faults, real-time seismic monitoring to flag incipient earthquake sequences, and clear protocols to modify operations when a problem is first detected,” he said.
That success, though, has not come easily or without cost.
“The marked decline of seismicity rates in Oklahoma in recent years could be called a success story, although it came only after residents endured years of earthquakes. It has certainly been valuable to understand how changes in operational practices can cause earthquake rates to decrease so much,” he said.
Further, work done at the Illinois Industrial Carbon Capture and Storage Project near Decatur has shown how collecting and integrating key subsurface datasets can be used to identify and avoid seismic hazards.
“From my perspective, one of the greatest successes has been the impressive levels of collaboration that we’ve seen between industry and researchers in the government and academia,” Snee noted.
All of that, he believes, has led to large increases in public databases of subsurface stresses, faults and rock properties in areas of interest for preventing induced earthquakes.
This flow of information – the cooperation between researchers, regulators and operators – was among the central talking points of the “Risk and Mitigation of Induced Seismicity” panel session at the recent Unconventional Resources Technology Conference, of which Snee was a part.
“It is important that we better understand and reduce the hazards and risks associated with induced earthquakes,” he said.
There is not a one-size-fits-all approach to mitigating such risks and exposures – and with good reason.
“Every community has a different level of risk exposure and risk tolerance,” said Snee.
Some have little or no tolerance of felt earthquakes, whereas others, particularly those who are invested in local resource development, have been more willing to tolerate felt earthquakes. Such exposure depends heavily on the population density, building construction standards and the presence of critical infrastructure.
“In most areas currently experiencing induced seismicity, operators and regulators are focused on preventing earthquakes large enough to be felt, with the goal of limiting nuisance earthquakes, and modifying or stopping operations well before a potentially damaging earthquake occurs,” Snee said.