Does Fracking Induce Seismicity?
I’m honored this paper has received such positive feedback, that it was circulated nationwide, and that Forbes’ David Blackmon suggested that it “should be required reading.” Since this paper was first published in June 2014, there have been a significant amount of scientific studies published by dozens of universities and government agencies that have nearly pinpointed the direct cause of our earthquake swarms in Oklahoma (and possibly nationwide). Subsequent studies have largely proven the conclusions made when it was written.
Summary
This paper will give you an answer to two energy issues being discussed today in America: does hydraulic fracturing (“HF”) induce seismicity, and do Class II injection wells induce seismicity? This paper will also briefly explore the environmental impacts of HF.
For the non-industry reader, HF is a process that occurs during the completion stage of an oil/gas well. The EPA defines HF as “the process of directing pressurized fluids containing any combination of water, proppant, and any added chemicals to penetrate tight formations, such as shale or coal formations, that subsequently require high rate, extended flow back to expel fracture fluids and solids during completions.”
This process lasts a matter of hours, not weeks or years. That is to say, HF doesn’t continue on down the road and cause seismic events years later; you should see a correlation between HF and seismic activity almost instantly.
PART I: Has HF caused (induced) seismic activity?
Hydraulic Fracturing (“HF”) has become gained celebrity status among American political commentary to the point it has become quite divisive. Oklahoma in particular has been the celebrity, garnering national attention due to its number of felt earthquakes – far exceeding that of California. We’ve been published in the LA Times, Time Magazine, and other major publications as well as news cycles like the Today Show. We will focus our attention here as a result.
HF was pioneered in 1947. Since that time, this completion technique has been used in over 1.2 million wells. From 1900-1928, Oklahoma was the world’s largest producer of oil and gas, and subsequently, Oklahoma remained one of the top producing regions in the world. Geologically, Oklahoma is one of the oldest and most faulted states in the nation giving ample amounts of opportunity for seismic activity.
Figure 1 shows Oklahoma appearing seismically dormant for nearly 30 years (1947-1980s); during that time over 291,000 wells were completed.
Some might argue a delayed correlation but, as we noted earlier, HF only lasts hours, not weeks or years. Further, the scale of seismic events is greatly skewed in comparison when you add in the 2000s. Properly put into proportion, the trend looks quite a bit different (Figure 2). Oklahoma is now completing less than one-quarter the amount of wells from its peak in 1981, and seismic activity is all the sudden spiking without any real rhyme or reason. Historically, it’s safe to say that HF didn’t cause seismic activity.
Has HF induced seismic activity during the modern era?
The modern era can be defined post-Devon Energy’s acquisition of Mitchell Energy in 2002 and shortly before. As chronicled in The Frackers, George Mitchell’s company spent the better part of the 90s attempting to utilize fracture stimulation to unlock shale gas in the Barnett before stumbling upon the “slickwater frack” and, thus, beginning the Shale Revolution as we know it today. A star was born when Oklahoma City’s Devon Energy realized its potential and took it to the next level, combining horizontal drilling with the new slickwater fracks designed by Mitchell’s team. Since the modern era began with George Mitchell’s Barnett Shale in North Texas, let’s quickly look for a correlation between HF and seismic activity there.
As you already knew, or now know, every well drilled into the Barnett formation from 2000-2008 utilized fracture stimulation as part of the completion technique. Ironically, there are zero seismic events during those eight years (Figure 3); there was actually an inverse (negative) relationship between the number of wells and seismic events from 2000-2007 and 2009-2013 – only one year, 2008, show any potential correlation (increase in number of wells and increase in number of seismic events). As with the Oklahoma example, there doesn’t appear to be a strong correlation between HF and induced seismicity.
Let’s look at Oklahoma again, this time only during the modern era. As Figure 4 illustrates, the number of total permits awarded in Oklahoma during the same timeframe as the Barnett Shale graph (Figure 3). As you can easily see, there is no correlation between wells permitted and seismicity. Again, we see inverse relationships nearly every year.
Figure 5 focuses only on horizontally completed wells in Oklahoma. This time, the chart indicates a seemingly delayed correlation. We’ll discuss that a little later in PART II – Disposal Injection Wells
Now, this is not to say that there is NO correlation between HF and felt seismic activity because there certainly is. As Columbia University’s Roger Anderson states, “hydrofracking by its nature causes tiny earthquakes because it involves fracturing of rock—but these are largely imperceptible, as the process takes place in relatively weak, shallow shales that crack before building up much strain.”
The Oklahoma Geologic Survey (OGS), for example, has suggested two separate events were triggered by HF: the Eagleton #1-29 (2013) and the Picket Unit B #4-18 (2011). More recently, Skoumal et al published their research, directly linking HF to seismic events in Poland Township, Ohio (2015). The authors of the study, Skoumal and Brudzinski, subsequently published in the National Review Online “The Fracking Fracas over Earthquakes”, aimed to calm the sensationalized response. As they state “an outright ban [on fracking] wouldn’t be appropriate… if you’re right on top of it [a 3.0 earthquake], it’s like a milk jug falling off a counter, and that’s it. A 3.0 is not a danger to anybody.” As we just read above, it would take 1,000 of these earthquakes to equal the energy release of one single 4.0 magnitude earthquake. It is, thus, readily acknowledged herein that HF has been linked to rare cases – so rare that earthquake expert Cliff Fohlich told National Review Online he is aware of “maybe ten other instances on the planet” where HF triggered an earthquake. The science has certainly begun to mount to that end. So much so that Harvard University’s Chair of Earth & Planetary Sciences Department stated “[f]racking is probably not a direct cause of associated earthquakes.” As the Barnett and Oklahoma charts on pages 2-4 illustrate, those rare exceptions certainly aren’t the rule and we shouldn’t treat them that way either.
We’ve chosen to focus on Oklahoma and Texas for relevant case studies to show that HF isn’t to blame for increased seismic activity – at least generally. What about all those areas in the United States absent of oil and gas activity? What is causing their significant increases?
The State of Idaho, which is an oil and gas powerhouse, has seen a significant increase in seismic activity. In case you missed my attempt at humor, Idaho ranks 42nd in energy production by EIA’s most recent data, yet we are reading “Central Idaho Rocked by Hundreds of Earthquakes.” We even saw an abnormally large 8.0 magnitude earthquake rock Alaska’s Aleutian Islands (you’d need 32 MILLION earthquakes the size of the one reported in Poland Township, Ohio to equal the energy release of that single earthquake). Again, no oil and gas activity is present in either area, so it’s more likely we are experiencing a greater trend – earth is waking up and becoming more tectonically active. .
[A Quick Tangent] Environmental Impacts of HF
Whether intentionally or not, media outlets routinely strike fear into their uneducated, innocent readership when discussing fracture stimulation. They list the ingredients of HF as “water, chemicals and sand” without providing any basis for what that mixture actually looks like. While technically truthful, the reality is that 99.5% of the mixture is made up of water and sand, with a very small 0.1% to 0.5% being chemical in nature. Admittedly, the Wall Street Journal – a pro free market paper – still gets this grossly wrong in their article, “Energy Companies Try New Methods to Address HF Complaints”.
Freshwater Use Further, they use phrases like “millions of gallons”, again without providing any context. As reported by Energy In Depth, “The Facts on Hydraulic Fracturing and Water Use”, a single frack uses about 4 million gallons of water. In comparison, New York City uses 4 million gallons of water every 6 minutes and golf courses in the United States use 2.1 billion gallons of water per day – that’s equivalent to 525 oil and gas wells being fracked in a single day. Energy In Depth also published a fantastic two-part story on freshwater use in California. They point out that HF operations in California used about as much water as a single California golf course, and there are 800-1100 golf courses depending on who you ask)! So, oil and gas uses a minuscule amount of freshwater compared to other, non-essential industries like Recreation. Let that sink in… freshwater use for all HF wells in California = freshwater use for a single California Golf Course!
Wastewater As you’ll see later in this paper, Duke University demonstrates the ratio of energy per unit of wastewater has dramatically increased due to hydraulic fracturing (i.e. hydraulic fracturing has decreased the amount of wastewater produced while increasing the amount of energy harvested). Not only is this an environmental success, it also hints that hydraulic fracturing might also decrease the amount of freshwater needed to recover each unit of energy – another environmental success. Behold, we happened across a study from the American Chemical Society published September 18, 2014 on the subject and our gut was right – conventional oil production, which yields less oil per well, requires more water usage than unconventional oil wells. Environmentalists should, thus, be singing its praises.
Groundwater Contamination With respect to groundwater contamination, the Environmental Protection Agency’s Lisa Jackson herself admitted in May 2011, “[I am] not aware of any proven case where the HF process itself affected water.” Since that time there have been no links of freshwater contamination to HF, only poor cement jobs and circumstantial gas seepage in active areas.
A group of “often accused anti-HF” researchers from Duke University released a definitive study on this very subject September 15, 2014. The Duke study set out to answer two questions (a) “Are elevated levels of hydrocarbon gases in drinking-water aquifers near gas wells natural or anthropogenic?” and (b) “if fugitive gas contamination exists, what mechanisms cause it?” After studying 113 Marcellus wells and 20 Barnett Shale wells, their very straightforward answer was that “our data do not suggest that horizontal drilling or hydraulic fracturing has provided a conduit to connect deep Marcellus or Barnett formations directly to surface aquifers.”
Even Harvard University’s Chair of Earth & Planetary Sciences Department made statements to this effect, “…fracking itself probably does not release fluids into the water [acquifers]…”
Then, there is always the EPA’s nearly 1,000-page study on the relationship between groundwater and HF, which stated that HF has “not led to widespread, systematic impacts to drinking water resources.” This is a convoluted way of saying, “after years of dissecting every report on the subject, we didn’t find anything.” EID published a fun Top 10 on the study so you don’t have to read it all, the biggest of which greatly expands the definition of HF from the EPA’s own definition provided on page 1.
Harmful Chemicals Finally, the University of Colorado published first-of-its-kind research on the chemicals used in HF fluids, which many proponents of HF bans throughout the country rely upon to support their notion that HF is bad for the environment. This graphic from DangersofHF.com illustrates how proponents of HF bans illustrate the “facts”:
Earthworksaction.org takes the fear-mongering to the ultimate level making statements such as, “Many fracturing fluid chemicals are known to be toxic to humans and wildlife, and several are known to cause cancer.”
What did the University of Colorado’s first-of-its-kind research say? The University of Colorado analyzed surfactants from HF fluids found from Colorado, Louisiana, Nevada, Pennsylvania and Texas. “The results showed that the chemicals found in the fluid samples were also commonly found in everyday products, from toothpaste to laxatives to detergent to ice cream.” They go on to further state, “What we have learned in this piece of work is that the really toxic surfactants aren’t being used in the wells we have tested.” Science is continuing to mount on the side of hydraulic fracturing being a safe technology.
HF Conclusion
Keeping in mind HF operations last a matter of hours and seismic activity should ensue shortly (days, not years) thereafter, I believe we can now put to bed the idea that “HF” is causing felt seismic activity.
Even though we know with a large degree of certainty that HF doesn’t induce seismic events, the anti-fracking hysteria is spreading across the United States. Republicans in Louisiana have moved to ban HF, New York has banned HF using debunked research - one study cited was immediately disavowed by the executive director of the publishing organization well in advance of New York’s decision touting peer-reviewed science which also appears to be in violation of ethical conduct. New York now faces a major lawsuit, which could find punitive damages in favor of mineral owners’ rights being infringed upon, and the Southern Tier trying to secede from New York and be absorbed by Pennsylvania as they fight for their economic life – admittedly a long shot.
The City of Denton, Texas, also passed a HF ban November 4, 2014. This ban is somewhat unique to all others since it was originally placed on the ballot as an initiative by those who took issue with how close gas wells were to their homes. Of course, after it was found on the ballot, environmental and earthquake fears began to swirl. The Denton HF ban lead to economic damage, loss of property rights, and legal problems that proved to be very costly to Denton taxpayers as the head of the Barnett Shale Energy Education Council suggested it would prior to the election. Regardless, the vote was overwhelmingly passed (60%-40% in favor) and it only cost $70,000 to strike that kind of fear into the hearts of innocent constituents – not even $700,000 backed by numerous university studies could sway them. Thankfully, the Denton City Council voted 6-1 to repeal their fracking ban, which has cost the taxpayers between $500,000 and $1 million depending on whom you talk to – The Denton RC for example.
What’s the bottom line? Fear is a powerful motivator and you don’t have to spend very much money to strike fear into the public regardless of what the science says. Politicians should know better than the populous, however. They are professionals elected to educate themselves on the issues concerning their constituencies. As you’ve just read, your politicians should also have read. Any hyper-aggressive regulation of hydraulic fracturing supported or proposed by any politician should highlight them as a phony, sell-out politician who is in someone’s back pocket. They think of you as an idiot, incapable of understanding the numerous studies presented to you herein. Secretary of Interior Sally Jewell, nominated by President Obama, might sum it up best, “There is a lot of misinformation about fracking. I think localized efforts or statewide efforts in many cased don’t understand the science behind it… [bans are] the wrong way to go”.
There is a seemingly delayed correlation in Figure 5, which needs to be evaluated, however. It’s one that implicates a potential culprit for increased seismic activity – disposal injection wells.
PART II: Disposal Injection Wells
For those not familiar, disposal injection wells and HF are completely separate events in completely separate wells. Unfortunately most reporting you see on the subject – similar to StateImpact and Time Magazine – report “disposal of drilling wastewater used in HF has now been scientifically linked to earthquakes” and suggest that disposal and HF are so closely tied together they are one and the same. They also write about disposal wells as if they are something new – they are not. Disposal wells, or Class II injection wells, have been directly linked to seismic events as early as 1968 in Denver, Colorado, as this article in Science Magazine reads. It’s nothing new and we’ve learned a great deal already about their impacts.
Disposal wells exist due to EPA Regulations of the oil and gas industry. The EPA requires the oil and gas industry dispose of all produced fluids (fluid that comes out of the well) and treat it as hazardous waste. So, when fluid comes to the surface [out of a formation underground], the oil and gas industry then trucks or pipes it to a disposal well where it’s then re-injected back into the earth to be disposed of safely and prudently. Of course, new technologies like Ecosphere could possibly antiquate this practice if the EPA would update its rules to the 21st century. While the fluid injected back into the earth does contain some frack fluid, the majority comes from existing fluid residing within the reservoir prior to drilling and completion (i.e. non-frack fluid). So, HF is a process used while completing an oil or gas well and a disposal well is a completely separate well that is used to dispose of produced fluids from multiple oil and gas wells. In many cases, a single disposal well services multiple oil and gas wells.
While some cases have been directly linked to disposal wells, they are a relatively small percentage. Out of over 140,000 currently active Class II injection wells, only a handful of wells have been associated with induced seismic activity. Not all of those injection wells are disposal wells either; approximately 60% of the active injection wells in Oklahoma are wells used for enhanced oil recovery, not disposal. Further, Oklahoma’s oil and gas regulatory body, the Oklahoma Corporation Commission, has stated “In most of the cases examined thus far [since 2011], there have been either no injection wells operating in the area of interest [seismic activity], or the focus of the events were at depths that varied greatly from the operational depths of the injection wells in question.” Columbia University’s 2012 report actually placed a probability on the chance of an injection well causing a seismic event in a dormant fault at 1:200 or 0.5% chance. An examination of another major producing shale play, the Bakken, indicates that not all states are created equal because unlike in other oil-producing states, North Dakota experiences little or no seismic activity.
Still, steps are actively being taken on the State level to ensure seismic events are more closely monitored and located off-fault. Of course, not all the faults in the United States are known and that is one reason why some wells have caused seismic activity. Ohio, for example, experienced a cluster of seismic events due to injection wells drilled into unknown, dormant faults. Since then, Ohio has put into place new permit conditions: companies that drill within three miles of a known fault or area of seismic activity greater than 2.0 magnitude are required to install seismic monitors that are linked to Ohio Department of Natural Resources (ODNR) – full story on the event can be found here. Unfortunately, the story fails its readers with the ever-present lack of understanding between HF and disposal injection wells.
ODNR even published a study back in 2012 that stated the ingredients that must be present to induce a seismic event:
“a fault must exist within the crystalline basement rock,”
“that fault must already be in a near-failure state of stress,”
“an injection well must be drilled deep enough and near enough to the fault and have a path of communication to the fault,” and
“the injection well must inject a sufficient quantity of fluids at a high enough pressure and for an adequate period of time to cause failure, or movement, along that fault (or system of faults).”
What was the ODNR’s conclusion? “All the evidence indicates that properly located Class II injection wells will not cause earthquakes” since “[g]eologists believe it’s very difficult for all conditions to be met to induce seismic events.”
SMU released a report as well, in conjunction with the University of Texas at Austin, on the link between injection wells and “earthquakes”. Their report indicates there are more than 2,400 injection wells in the Fort Worth Basin (Barnett Shale area) and noted the vast majority do not have seismic activity associated with them. Still, they were reluctant to rule disposal wells out of the equation even though Texas’ regulatory body, the Railroad Commission of Texas, found no data that linked hydraulic fracturing to earthquakes and no significant correlation between seismic activity and injection wells. You can read more on the SMU-UT study here.
More recently (July 4, 2014), geologists from Cornell University, together with the University of Colorado and the United States Geological Society (USGS), issued a study titled Sharp Increase in Central Oklahoma Seismicity since 2008 Induced by Massive Wastewater Injection. This peer-reviewed study states, “fluid migration from high-rate disposal wells in Oklahoma is potentially responsible for the largest swarm” and “Our earthquake relocations and pore pressure models indicate that four high-rate disposal wells are capable of increasing pore pressure above the reported triggering threshold throughout the Jones swarm. Although thousands of wells operate aseismically, four of the highest-rate wells are capable of inducing 20% of 2008-2013 central US seismicity.” This study doesn’t come to any direct, concrete conclusion that these four, high-rate wastewater injection wells are to blame for increased seismic events in Oklahoma other than in title of the study, which doesn’t appear to be appropriately named. Of course, the LA Times and others write headlines like, “Study links Oklahoma earthquake swarm with HF operations.” Can you see what’s wrong with this kind of journalism? First, nowhere does the study say anything about HF being linked to wastewater injection wells. Second, the study shows that four wells potentially triggered a swarm of seismic activity with no clear, definitive statement that is the case.
Of note, the Cornell-Colorado-USGS study did indicate that fluid migration from disposal injection wells could travel up to 19 miles from the well site and documented that the 2010 earthquake swarm ruptured an “unmapped” fault. We will need time to understand fluid migration. These are two important annotations because they get to the crux of the problem: we don’t understand underground fluid migration quite yet and we don’t’ have every fault mapped across the United States.
Since the study, the evidence has continued to mount, however, that disposal injection wells are, in fact, the cause of much seismicity – at least in Oklahoma. The Oklahoma Geological Society (OGS) updated its statement on earthquakes appropriately in April 2015 to state two key findings:
“Based on observed seismicity rates and geographical patterns of migrating seismicity in Oklahoma, which follow major oil and gas plays with large amounts of produced water, these rates and patterns of seismicity are very unlikely to represent naturally occurring rate change and process.”
“It is often stated that disposed water is waste-water or “flow-back water” from hydraulic fracturing. While there are large amounts of waste-water generated from hydraulic fracturing, this volume represents a small percentage of the total volume of waste-water injected in disposal wells in Oklahoma.”
Stanford University released the most recent study (June 18, 2015) in Science Magazine to more precisely state that, “In three study areas that encompass the vast majority of the recent seismicity, we show that the increases in seismicity follow 5 to 10-fold increases in the rates of saltwater disposal. Adjacent areas where there has been relatively little saltwater disposal have had comparatively few recent earthquakes.”
Oklahoma’s earthquakes appear to be due to disposal injection wells located in the Arbuckle formation, overlying crystalline basement and that the Arbuckle formation appears to be “in hydraulic communication with potentially active faults in crystalline basement, where nearly all the earthquakes are occurring.”
Regardless, we are beginning to see some clear conclusions from all these studies – (a) don’t inject into known faults, (b) increased monitoring of wells in areas of known seismic activity is a good idea, and (c) don’t inject too close to crystalline basement whereby you could possibly have hydraulic communication. It wouldn’t be crazy to suggest shutting down Class II injection wells that end up drilling into unmapped faults either – a recommendation that has now been put into effect in Oklahoma and other states.
If HF increases wastewater disposal volumes, we should stop HF, right?
This argument has started to surface lately: HF increases wastewater volumes that are pumped into disposal injection wells, a handful of disposal injection wells induced seismic events, and therefore HF is bad. From the studies done on wastewater thus far, this couldn’t be further from the truth. In 2012, Duke University released an extensive report on wastewater generation and disposal from natural gas wells in Pennsylvania.
As Duke’s study states, “overall waste volume is very different between Marcellus wells and conventional wells, while the waste composition is only slightly different if both wells types are hydraulic fractured.” If composition really isn’t any different, does the volume associated present negative impacts?
“In Figure 13-A, the pie charts are sized to show the difference in the amount of waste production, with Marcellus wells producing approximately 11.3 times more waste per well than conventional wells. However, Marcellus wells also produce approximately 40 times more gas in 4 years on average than conventional wells. As a result Marcellus wells are significantly more efficient than conventional wells in terms of waste production, with about three times as much gas produced for each gallon of wastewater. This ratio may go down slightly over time as gas production from hydraulically fractured wells decreases, but even with no gas production beyond the fourth year and steady brine production mirroring the fourth year brine values, Marcellus wells remain more efficient than conventional wells… Pennsylvania dramatically increased its production of natural gas producing 10 times more natural gas in 2011 than 2004, with only 4 times more waste production.”
Thus, modern wells have actually increased the efficiency by which oil and gas companies extract gas per unit of waste. As a result, less water will be injected back down into disposal injection wells than without the coupling of horizontal drilling and HF. The environmental movement should be praising the oil and gas community for such a development.
In Summary
Generally, anytime we see something detrimental happen (e.g. earthquakes, war, sub-prime mortgage crisis etc.), politicians run to the rescue with hyper-aggressive legislation aimed at taking advantage of available political capital. Of course, I’m being sarcastic when saying they rush to the rescue.
This paper has explored seismic activity in relation to oil and gas operations across the United States, with a focus on the most active states. More specifically, we explored HF operations in areas of increased seismic activity, HF’s environmental impact and disposal injection wells in areas of increased seismic activity. We have quashed the argument that “HF causes earthquakes” (a conclusion that Standford University's Mark Zoback has plainly stated on CNBC), shown that HF’s use of freshwater is environmentally dwarfed in comparison to other non-essential industries like recreation (a conclusion the EPA came to as well), only a statistical few disposal injection wells have induced seismic activity, and hydraulic fracturing actually decreases the amount of wastewater that needs to be disposed of via injection wells per unit of oil and gas produced. We also realized non-energy states like Idaho are seeing significant increases in seismic activity that obviously can’t be linked to oil and gas activity while other major energy producing states like North Dakota have been devoid of statistical increases in seismic activity. So, what do we do?
Arkansas and New York have displayed exactly what not to do; Arkansas banned disposal injection wells altogether while New York banned HF altogether. The City of Denton, Texas, gave us a perfect example of legislative oversteps and the undue financial burden it places upon their taxpayers. It should be noted that famed environmentalist Governor Jerry Brown of California even refuses to subject California to such a ban, saying “it doesn’t make a lot of sense.”
I’m proud to say Oklahoma is taking a very reasonable approach. The Oklahoma Corporation Commission (OCC), Oklahoma’s oil and gas regulatory body, says is is “keeping an open mind” about ongoing research, correctly stating “we are having to manage perceived risks.” Meanwhile, Oklahoma Governor Mary Fallin signed into law a bill that requires operators of injection wells to report daily, instead of monthly, injection volumes – a measure that took effect in September 2014. Subsequently, the OCC implemented a traffic light system in March 2015 that applies to 347 of the approximately 900 disposal wells in Oklahoma located in the Arbuckle formation – a formation that sits on top of basement rock. So far, the traffic light system has shut down more than 50 disposal wells and reduced injection rates of approximately 150 wells. In July 2015, the OCC updated the March directive and Operators have a matter of days to prove they are not injecting below the Arbuckle formation where the well could be in communication with basement rock. There is, thus, clear evidence that regulators are paying attention and taking real, proactive steps to ensure the public’s safety while balancing delicate economic considerations. Oklahoma should be proud.
Sources (not hyperlinked):
United States Geological Society (USGS)
Oklahoma Geological Society (OGS)
Oklahoma Corporation Commission (OCC)
Railroad Commission of Texas (RRC)
DrillingInfo
US Energy Information Administration (EIA)
EnergyInDepth
University of Oklahoma
Columbia University
Southern Methodist University
The University of Texas at Austin
Duke University
University of Colorado
Stanford University
American Chemical Society
Harvard University