News blog

On fracking and seismic hazards: better data, more quakes

US Geological Survey

A detailed analysis of seismic activity around hydraulic fracturing operations in Texas suggests that small earthquakes generated as a result of the reinjection of waste fluids could be more common than once believed.

Cliff Frohlich, a seismologist at the University of Texas at Austin, used highly sensitive data from the seismic component of the National Science Foundation’s EarthScope initiative to investigate the impact of hydraulic fracturing operations tapping into the Barnett Shale. Operating across a 70-kilometre grid, Frohlich identified 67 tremors with magnitudes as small as 1.5 from November 2009 and September 2011, according to the study, published today in the Proceedings of the National Academy of Sciences. All of the 24 earthquakes for which the epicentre could be reliably located were within 3.2 kilometres of at least one injection well.

The analysis seems to provide the most detailed view yet into an issue that has garnered increasing press and scientific scrutiny (for some of Nature‘s earlier coverage, check here and here). Reports of fracking-related earthquakes have surfaced in the both the United States and the United Kingdom over the past year; back in April the buzz was about a US Geological Survey study linking a broader increase in earthquake activity to fracking. Here’s the bottom line of that analysis:

“Beginning in 2001, the average number of earthquakes occurring per year of magnitude 3 or greater increased significantly, culminating in a six-fold increase in 2011 over 20th century levels.”

But although earthquakes seem to be associated with wells, Frohlich’s study did not find that all wells were associated with earthquakes. Of the 161 wells in the study area with significant and sustained injections (defined as 150,000 barrels of water per month), 90% had no nearby earthquakes. Among those that did, the injections had been going on for more than a year in most cases.

The danger of such relatively small tremors remains unclear, but such studies show that scientists are beginning to home in on the basic facts.


  1. Report this comment

    John Vidale said:

    I’m surprised that the expectation is not that MOST injections cause at least a few earthquakes, generally with the normal Gutenberg-Richter distribution of many more tiny earthquakes and far fewer big earthquakes, just with widely variable rates of earthquakes. We think the crust is often critically stressed, and know that the more carefully we look, the smaller are the earthquakes that are visible, and hence likely the larger fraction of wells that would be showing seismicity.

    It seems to me the question at each site is “At what RATE are tiny earthquakes happening?”, not whether any earthquakes at all are present, and then the big question is “At what rate are disturbing earthquakes likely to strike?”, not whether disturbing earthquakes are possible or impossible.

Comments are closed.