In recent years a consensus in the scientific community has emerged that, while hydraulic fracturing (“fracking”) rarely causes earthquakes large enough to be felt, the disposal of waste fluids produced in fracking operations in deep injection wells can, in some cases, induce earthquakes on nearby faults.
Although natural earthquakes occur in many of these same areas – e.g., Oklahoma, Texas, Arkansas – the rate of events has climbed precipitously since 2009. For example a recent study by the US Geological Survey and Oklahoma Geological Survey found that 145 earthquakes of magnitude 3.0 or greater occurred in Oklahoma in the first 4 months of 2014, compared to an average of about 2 per year before 2009. These earthquakes, most of which were at least locally felt, have raised concerns about possible future damaging earthquakes.
Researchers have worked for more than 50 years to understand injection-induced earthquakes. What causes earthquakes to happen when fluids are injected? Why are earthquakes induced at only a small percentage of deep wells? How can one know if any particular earthquake is natural or induced? There are many important but unresolved questions.
Where earthquake hazard is concerned, the study of earthquake ground motions is where the rubber meets the road. That is, in general, the damage potential of any earthquake is expected to depend on its magnitude and location relative to population centers, but the actual level of hazard depends on the severity of the shaking—what seismologists call the ground motions.
Seismologists have done a lot of work over the years to characterize the level of shaking that will be generated by an earthquake of a given magnitude in a given region. We know that in the central and eastern U.S., seismic waves travel especially efficiently through the earth’s crust, extending the impact of an earthquake relative to a temblor of the same size in California. We have developed relationships that predict expected shaking from, say, a magnitude 5.5 earthquake in the central U.S.
Relatively little work has been done to-date on the shaking severity from induced earthquakes. In general, we assume an earthquake is an earthquake is an earthquake. Until recently, there was no reason to expect that an induced earthquake of a given magnitude will generate shaking that is different on average from a natural earthquake of the same magnitude.
And yet, it turns out, they do. Since 1999 the USGS “Did You Feel It?” (DYFI) Web page has produced maps of shaking severity for significant earthquakes, not using data from seismometers, but from information provided by citizens via detailed Internet questionnaires. These intensity data turn out to provide reliable and invaluable information, in particular in places like the central US, where seismometers are sparse. Considering DYFI data from the largest earthquakes in the central and eastern U.S. that are believed to have been induced, my research indicates that, on average, an induced magnitude 5.0 earthquake generates shaking comparable to that from a natural earthquake with magnitude 4.2. The difference holds at all distances except within 10 km of the epicenter, where shaking is about the same as that from natural earthquakes.
How could shaking from injection-induced earthquakes be so different from natural earthquakes with the same magnitude? It turns out that the magnitude scale that seismologists normally use in modern practice is not a direct measure of shaking severity. Rather, the magnitude reflects the size of a fault that moves, and the distance that it moves, or slips, during an earthquake. Shaking, in contrast, depends on the details of the movement, for example whether the motion was relatively fast or slow. Natural earthquakes differ from one another, but on average we assume that all earthquakes of a given magnitude generate similar shaking.
“Did You Feel It?” data reveals that induced earthquakes are different beasts, generating less shaking than natural events, at least at distances greater than 10 km from the earthquake. This effect is offset at close distances by the especially shallow locations of induced earthquakes. Earthquake depth matters because, even if an earthquake generates relatively low shaking, the shaking will tend to be stronger if it is closer (in a vertical direction from the ground surface) to where people live. Overall the results of my study suggest that damage from induced earthquakes will be especially concentrated near the epicenter compared to their natural earthquake counterparts. Much work remains to be done to characterize fully the potential hazard posed by such events.