Do I Think That Fracking Causes Earthquakes?

June 7, 2014

Alan Kafka
Weston Observatory
Department of Earth and Environmental Sciences
Boston College

Note: My analysis presented here is based on what was known about this issue at the time that I wrote this (June 2014). I don’t think the essence of this analysis has changed (yet?), but as more earthquakes occur near fracking sites the details of the issue are evolving.

These days, I can hardly go anywhere without being asked: “So, tell me, does fracking cause earthquakes?” Not wanting to get into a complex scientific and political argument on the spot (especially in social situations), I generally try to change the subject.

But, I guess there’s no avoiding it. Being a seismologist, I do have a responsibility to answer. So, here’s my take on it.

Short and simple answer: Yes.
Longer answer: But, it’s complicated…

“Hydraulic fracturing,” also known as “fracking”, is a method of injecting fluid into the ground to fracture rock for extracting natural gas. In the process, fracking creates many very small earthquakes. These small earthquakes are generally (so far?) too small to be felt or to cause structural damage or injury. However, part of a fracking operation involves the use of high volumes of water to release natural gas from dense rock, and disposing of the associated wastewater involves injecting it into deep rock formations. That wastewater injection can “induce” or “trigger” earthquakes in faults that have been dormant (and might have otherwise remained dormant) for a very long time. These wastewater-injection induced earthquakes are not necessarily so small, and can be damaging. And that is (so far?) where the fracking/earthquake problems and controversies lie.

Here’s my summary of the story:

  • There are well-documented cases of earthquakes clearly associated in time and space with fracking operations. But it’s not really the fracking itself that triggers the earthquakes that are likely to be damaging, it’s the disposal of fluids via injection of highly pressurized wastewater into faults that more likely tends to induce the larger, potentially damaging earthquakes.
  • Most earthquakes associated with fracking operations have been smaller than magnitude 3. But a few larger and damaging earthquakes are suspected of having been induced by the injection of wastewater from natural gas (and also oil production) operations. On the seismogram shown below, you can see the Weston Observatory/BC-ESP recording of a magnitude 5.6 earthquake that occurred in Oklahoma on November 6, 2011 and is suspected of having been induced by injection of wastewater from oil extraction. Although that earthquake has not been directly linked to fracking, the injection of wastewater from oil extraction is essentially the same process as what occurs in fracking operations.
  • There is a well-known theoretical explanation of why injection of highly pressurized wastewater into faults could induce earthquakes. The increased pore pressure along the fault effectively lubricates the fault, causing it to release stress that might have been building up for many years, but might not have slipped without the excess pore pressure.
  • Out of hundreds of thousands of fracking operations, and the associated disposal of wastewater, so far only a few have been clearly identified as being related to induced earthquakes of any significant size. The vast majority of operations have not, so the probability of a given fracking operation inducing damaging earthquakes is probably very low (much less than 0.1%).
  • Just how large and damaging a fracking/wastewater injection-related induced earthquake could be remains unknown. Although such earthquakes are generally smaller than magnitude 3, and the largest earthquakes claimed to be fracking-related are less than magnitude 6, even larger future earthquakes can not be ruled out. More research on this topic will be necessary to answer the question of just how big future wastewater injection-related induced earthquakes might ever be.
  • There are many other environmental issues related to fracking, such as heavy truck traffic and contamination of nearby well water used by local communities for drinking water. These are important issues to consider regardless of the question of whether or not fracking operations induce earthquakes.

Bottom line for me: These kinds of things are complicated. I think a more relevant question than “Does fracking cause earthquakes?” is: Given that wastewater injection procedures associated with fracking operations can induce earthquakes big enough to cause damage in some (small?) percentage of cases (and that there are other environmental hazards associated with fracking), what should we do about it? Fracking provides new sources of natural gas that could enhance our ability to generate electricity, heat homes, and provide fuel for transportation. Given the uncertainties, how do we to find the right balance to make informed decisions about the extent to which the risks associated with fracking, or with any other process for finding new sources of energy, are worth the benefits?

This is the challenge for all of us as citizens of planet Earth.


Weston Observatory/BC-ESP seismogram of magnitude 5.6 earthquake that occurred in Oklahoma on November 6, 2011. This earthquake is likely related to injection of highly pressurized wastewater (from oil extraction operations) into wells. It was big enough to cause injuries and damage more than a dozen homes. (Click on seismogram to enlarge.)

Further Reading:

For an excellent, more complete, analysis of the situation, see the USGS Science Feature article Man-Made Earthquakes Update, and follow the links in it.

Clear and Simple Illustration of Mechanism to Explain Earthquakes in the Central and Eastern United States

June 6, 2014

This is a clear and simple illustration (adapted from the South Carolina Earthquake Education and Preparedness Program) of the commonly accepted “ancient zones of weakness” model for the cause of earthquakes in the Central and Eastern United States (and other “intraplate” regions far from plate boundaries). According to this model, preexisting faults and/or other geological features formed during ancient geological episodes persist in the intraplate crust, and, by way of analogy with plate boundary seismicity, earthquakes occur when the present-day stress is released along these zones of weakness.

IP_Quakes_Logs_AdaptedAlso see:

Why Does the Earth Quake in New England? and South Carolina Earthquake Education and Preparedness Program

Please Donate to Support Our Work

April 27, 2014

If you benefit from these blog posts and feel inspired to contribute to our work, please visit this page and make a donation.

At Weston Observatory, we’ve been busy this year envisioning and building an earthquake observatory for the 21st century, and we are now embarking on a fundraising effort to support our vision.

This year’s fundraising theme is education and public outreach. Funds contributed this year will primarily go towards bringing earthquake and related science into schools and public libraries, supporting our monthly public colloquium for adult learners, and of course towards maintaining this blog.

If you feel inspired to make a contribution, please click here to go to our donations page.

Thank you for any amount, small or large.


Broken Garage Door Springs, Earthquake Prediction, and Earthquake Triggering

March 6, 2014

Alan Kafka
Weston Observatory
Department of Earth and Environmental Sciences
Boston College

(Originally Posted: April 27, 2012)

A few days ago, one of the huge springs on our garage door opener broke, with quite a big KABOOM!

It seems to me that this is a perfect example of the problem of earthquake prediction. Here’s what I see as the connection:

In the morning, the garage door repair guy (who happened to be there fixing something else) looked up at the spring and said, “That deformed spot in the spring is ‘metal fatigue’ – it might break in 5 minutes, or it might break in 10 or 20 years.” Then, two hours later, it broke! (Perhaps he disturbed the internal stress field while he was fixing something else?)

This seems to be a perfect analog of the Elastic Rebound Theory of earthquakes, and the phenomenon of Earthquake Triggering!

It also explains why we knew that an earthquake was lurking in Haiti, but we couldn’t tell when it would occur: The tectonic stresses were ready to be released in an earthquake, but nobody could tell if the earthquake would occur in minutes, in years, in decades, or in centuries…

New Updated Seismicity Maps for Northeast U.S.

February 12, 2014

Click on the maps below to see the Weston Observatory seismicity maps of the Northeast United States, updated as of early February 2014.

The first map shows historical seismicity (from June 1638 to December 1974), and the second map shows network seismicity (from January 1975 to early February 2014).



The Detective Work of Seismologists: Earthquake or Blast?

January 10, 2014

Justin Starr
Weston Observatory
Department of Earth and Environmental Sciences
Boston College

Residents of the city of New Bedford, MA and the surrounding area felt something very strange at 10:52 am on January 9th, 2014. They heard a loud blast and felt a distinct rumble.

A seismic event had occurred.

Weston Observatory scientists quickly determined that the magnitude of this seismic disturbance was 1.9 and that it was located very close to New Bedford. Not very big, but shallow enough to be heard by many people in the area.

While speaking to the Massachusetts Emergency Management Agency (MEMA), Weston Observatory scientists were informed that there may have been blasting in the New Bedford harbor around the same time as the earthquake and very close to the possible epicenter. So was this a real earthquake? It was up to the scientists to find out.

Several phone calls went out and eventually, Weston Observatory scientists reached the New Bedford harbormaster who put them in touch with the Captain of… the Kraken! The Kraken is a drilling and blasting barge located in New Bedford and is tasked with widening the shipping lanes. As it turns out, the Kraken did not blast until 12:09 pm, over an hour after the earthquake occurred. In fact, the Captain received many phone calls asking if it was they who blasted and caused the shaking… but it was the earthquake!


The two seismograms shown below are from a USArray seismic station in Tiverton, RI. On those seismograms, you can see the Kraken Blast and the earthquake. Notice the big difference in magnitude (the earthquake is an order of magnitude larger).


And notice (below) the well-defined, high amplitude wave labelled “Rg” on the seismogram of the blast. The presence of these “Rg” waves on a seismogram are one of the ways that seismic “detectives” use to identify blasts, and distinguish them from earthquakes.


These “Rg” waves have been studied by seismologists, and are one of the ways that seismologists can distinguish between earthquakes and explosions. See, for example:

Kafka (1990): Rg as a Depth Discriminant for Earthquakes and Explosions

Yes, There Really is a “Sandwich Plate”

December 24, 2013

Alan Kafka
Weston Observatory
Department of Earth and Environmental Sciences
Boston College

An interesting pair of earthquakes recently occurred in the Scotia Sea region, south of Chile and Argentina: Magnitude 7.7 (November 17, 2013) and Magnitude 7.0 (November 25, 2013).

Click to enlarge.

Fortunately, because of their remote location, there was little or no damage associated with these earthquakes, and they didn’t make a lot of headline news. But they illustrate some interesting earthquake and plate tectonic processes of the Scotia Plate and the Sandwich Plate.

Click to enlarge.

The Scotia Plate/Sandwich Plate region is a tectonic region between the South American Plate and Antarctic Plate, and stretches from the South Sandwich Islands to the southern tip of South America. The Scotia Plate moves eastward relative to the South American plate, but the motion is complicated by the presence of a divergent boundary in the eastern portion of the Scotia Plate, delineating the western edge of the Sandwich Plate.

Click to enlarge.

Click to enlarge.

The two recent earthquakes had motion that is consistent with the long-term (millions-of-years time scale) plate motion (shown by the blue and white arrows in the top figure).

The second earthquake might be a remotely-triggered earthquake, an earthquake that is too far away from the first quake to be an aftershock, but (maybe?) too close in time to be just a coincidence…

New Opportunities for Exploring Global and Regional Earthquakes in the Classroom and Beyond

August 16, 2013

Alan Kafka, Justin Starr, and Anastasia Moulis
Weston Observatory
Department of Earth and Environmental Sciences
Boston College

Seismological observatories operate a variety of types of seismographs, each “tuned in” to some aspect of watching the Earth quake. At Weston Observatory, we monitor earthquakes recorded by “research seismographs” of the New England Seismic Network (NESN) and by “educational seismographs” of the Boston College Educational Seismology Project (BC-ESP).

The BC-ESP offers opportunities for students of all ages to collaborate with research scientists as part of their experience in school and beyond. Having a seismograph in a classroom, or other publicly accessible location, gives students of all ages direct experience with recording earthquakes. Educational seismographs are inexpensive and can be easily installed and operated in schools, libraries, and any other places that want to have their own seismograph. But these educational seismographs are limited in terms of the quality of seismic recording compared to what can be achieved with much more expensive research seismographs.

The figure below shows two different recordings of a magnitude 6.6 earthquake that occurred in Colombia on August 13, 2013. The seismogram on the left was recorded by an educational seismograph located in Stoughton, MA and the seismogram on the right shows the same earthquake recorded by an NESN research seismograph located at Weston Observatory. The earthquake is much more clearly recorded by the research seismograph, but the educational seismograph is much less expensive and is easily installed and operated at BC-ESP sites.

Colombia_081313Click on image to enlarge.

A new development is now enabling us to integrate these two aspects of our seismic recording at Weston Observatory. New software (currently in beta testing phase) called “jAmaSeis“, being developed by the Incorporated Research Institutions for Seismology (IRIS), in collaboration with Moravian College, makes it possible for us to bring educational and research seismograph data together in the same seismogram viewing and analysis environment. With this new software, students can have their own classroom seismograph, while simultaneously viewing and analyzing seismographs recording data at remote earthquake research observatories.

Our BC-ESP students, and other people who visit Weston Observatory, often ask us why we didn’t record some of the California earthquakes that were reported in the news. Being on an active plate boundary, California earthquakes of course occur more frequently than New England earthquakes, but many of those California earthquakes are too small to be seen all the way across the country at BC-ESP sites. But with the new jAmaSeis software, we are able to display the recordings from our own educational seismograph on the same screen as recordings from an educational seismograph located in California. This makes it possible for BC-ESP students to see real-time recordings of these smaller California earthquakes.

Thus, with the new jAmaSeis software, students at a school in New England could monitor earthquakes on both sides of the country, at a plate boundary and in the middle of a plate (and vice versa for students in California). In the figure below, the top seismogram shows what we recorded on August 4, 2013 at Weston Observatory (station WOBC), and the bottom shows what was recorded on the same day at Sitting Bull Academy, located in Apple Valley, CA (SBCA). At SBCA, you can see a magnitude 3.9 California quake and a magnitude 5.3 quake in Western Canada. At WOBC, you only see the bigger, magnitude 5.3 quake.

CA_Can_080413Click on image to enlarge.

These new developments in software and web-based networking are opening up new opportunities for students of all ages to explore global and regional earthquakes in the classroom and beyond, and to learn about science by participating with research seismologists as together we watch the Earth quake.

Relative Sizes of Some Recent Seismic Events

March 24, 2013

Alan Kafka
Weston Observatory
Department of Earth and Environmental Sciences
Boston College


A few things about this plot (all numbers are approximate, but they are correct enough for this level of analysis):

kT = kilotons

2013 DA14 refers to how big an event the near-Earth asteroid that missed us on Friday, Feb. 15, 2013 would have been if it crashed.

Japan quake is the 2011 mega-quake, and Haiti quake is the 2010 quake that devastated Haiti.

Haiti quake (magnitude 7) is very small (in terms of magnitude) compared to Japan quake (magnitude 9).

Everything else is very small compared to Japan quake.

In very rough numbers/guesstimate: The meteorite that killed the dinosaurs was equivalent to something like magnitude 10 to 12. If we use the number 11, that’s at least 100 times bigger than Japan quake. So, if we plotted the dinosaur meteorite, everything else here would probably be too small to see.

Hurricane Sandy Recorded by Seismographs: Interdependency and Interrelationships Within the Earth System

November 2, 2012

Alan Kafka
Weston Observatory
Department of Earth and Environmental Sciences
Boston College

Seismology provides an interesting window into the interdependency and interrelationships within the Earth System.

The seismograms of Hurricane Sandy shown below were recorded by the Boston College Educational Seismology Project (BC-ESP) on our BC campus seismograph. This is a good example of seismology as a window into the interdependency and interrelationships within the Earth System. Hurricane winds and waves generate seismic waves that are recorded by seismographs. And by coincidence, this is not only a fascinating recording of an historic hurricane, but it happens to also include one of the most well-recorded earthquakes I have ever seen on our educational seismographs. Plus, we just happened to record aftershocks of that well-recorded quake on the same seismogram as the main shock.

And, if you look very carefully near the beginning of the October 30 seismogram, you can see a magnitude 6.2 aftershock “hiding” beneath the hurricane waves.


Get every new post delivered to your Inbox.