Oklahoma is not OK.
Over five years ago, at the end of 2009, we already had the data to show a dramatic increase in Oklahoma earthquake activity. We would soon know why.
Having recently been reminded of the issue, I again retrieved the data (even up to the present) on Oklahoma earthquakes from the U.S. Geological Survey. Let’s ignore the small stuff, counting only quakes with magnitude 3 or greater. From 1974 through 2008, most months by far brought no such earthquakes at all in Oklahoma, and only once came more than one in the same month. But the single year 2009 saw quakes in 9 of 12 months, and as for months with more than one, there were 5 of them.
The increase is even more striking in annual counts, with no fewer than 20 in 2009:
Run the numbers any way you want, the change is significant. Something was up, causing an increase in Oklahoma quakes (and other states too), but what?
Naturally, scientists (in particular, the seismological type) studied the issue. Meanwhile the earthquakes continued, with the following year exceeding even 2009:
Oklahoma topped 40 quakes in that fateful year 2010.
Indeed Oklahoma wasn’t the only area affected, so was much of the western U.S., and when scientists from the U.S. Geological survey examined the data they found a striking coincidence: that the biggest increases were happening where the oil and gas industry was prospecting most aggressively. They presented their results at the 2012 meeting of the American Seismological Society, concluding that the increase is “almost certainly man-made.” Of course, by that time we had data through 2011; here are updated monthly counts for Oklahoma:
Surprising that Oklahoma exceeded 40 quakes the preceding year? In 2011 she saw more than 40 in a single month. Annual counts:
Oklahoma exceeded 60 quakes in 2011, and the evidence implicating the fossil-fuel industry was solid as a rock. Yet for some reason none other than the Oklahoma state seismologist felt compelled to say that the increase might be “the result of natural processes,” even invoking a rather vague reference to “natural cycles,” thereby exercising two of the old canards so often used to dispute global warming.
By 2013, further research had found the specific cause: wastewater injection, mainly related to that “fracking” thing. Yet some powerful lobbies still disputed it, so of course doing something about it wasn’t really considered. Meanwhile 2012 data were in for Oklahoma:
Not as many quakes in in 2012 as in 2011. But still far more than there used to be, and far more than there would have been without fracking.
By 2014 the evidence was beyond “rock solid.” It was beyond a reasonable doubt. In particular, research had pinned down the connection between the location of earthquake swarms and the locations of “injection wells” where the wastewater is deposited. There just wasn’t any doubt at all, yet still some powerful lobbies denied it, and the Oklahoma state government just wouldn’t acknowledge it. Meanwhile, the 2013 data were in:
For the first time, Oklahoma saw over 100 earthquakes in one year.
This year, the Oklahoma government has finally acknowledged that their earthquake rise is due to the fossil fuel industry. Meanwhile, the state legislature seems bent on outlawing any interference with fracking by local governments.
It’s what we’ve come to expect from Oklahoma. What else from a state whose senior senator’s understanding of climate change seems limited to the fact that it still snows during winter?
All the while the data keep rolling in. Like last year’s:
Almost 600 in Oklahoma last year. Remember when 20 quakes in one year was a surprise increase?
Earthquake frequency has always changed It’s “natural”!
Reblogged this on Hypergeometric and commented:
Funny how convenient it is to paint over inconvenient truths.
LOL!!! Your screen name is perfect for this. The increase certainly appears to be beyond geometric.
Volcanoes are dormant for centuries or millennia before suddenly blowing for natural reasons.
While I suspect these OK earthquakes almost certainly are related to fracking — and should be treated that way until we know otherwise — do we really know enough to rule out a natural cause?
We have similar issues where I live, caused by gas drilling. They made a nice animation showing the increase in frequency and magnitude that presents it more visually. http://www.dwarshuis.com/aardbevingen-groningen/
Not sure how accurate this is, but a fun visualization of earthquakes around gas extraction wells in the Netherlands is here:
That’s quite astonishing. From less than 5 per year, to 600.
Time to roll out Step 3 of the denial process: “Yes it’s happening, yes we’re causing it, but it’s not a problem! Earthquakes are a good thing!”
Indeed. They loosen the soil. It’s good for agriculture. Yeah… THAT’s the ticket.
of course it is a good thing. people pay big bucks for vibrators. :+)
There is a major fault zone here.
The first known earthquake near Oklahoma City was in 1882.
Tamino, my entire post was relevant. Your deliberate edit deliberately gutted substantive context.
[Response: You started out by asking “do we really know enough to rule out a natural cause?” Yes, we do. Suggesting otherwise is either not paying attention, or attempting to muddy the issue. As for the rest of your comment, it was just more muddying the issue.
You’re always welcome to comment here. When I regard them as attempts to confuse issues, or rooted in confusion, I’m always welcome to delete them.]
Reminds me of how state Republican politicians in Pennsylvania – that are so concerned about “centralized power” and “statism”- enacted laws to prevent municipal governments from regulating fracking in their own communities (just as they had prevented residents in Philadelphia from limiting legal gun purchases to 10 or less per month)…
Is this a phenomenon that is associated in particular with the geology of OK, or is it a generalizable issue?
Hey Joshua, it seems to be the fact that there are there faults (previously inactive) in the OK areas discussed. It’s pretty evident that this isn’t a nationwide issue, since the same thing isn’t happening everywhere that fracking is going on. It really stresses that fracking needs to.be conducted smartly in approved areas away from faults.
Hate to think of whats happening to the casing failure rate.
I was wondering the exact same thing. Haven’t seen any info on it yet, though.
I may be wrong, but earthquakes do not necesseraly affect much the casings in depth. If I understood it right, casings (if properly done) are coupled to the surrounding rock, and thus shaking are not amplified ; casings are for the most part made of steel, which withstands quite well deformations ; and finally the wave amplitude is amplified only at the surface because of the waves bouncing back at the atmosphere/rock interface.
Of course, that implies that 1) the operator cemented correctly the well 2) the cement is still working properly 3) the well does not cross an active fault and is therefore not cut properly by the fault movement 4) surface installations are on the surface and suffer more.
To be checked once again, but this is what I remember from post-earthquake surveys in Japan.
It’s just wind turbine albedo change.
This just again proves that fracking on fault lines is not a good idea. It should be common sense, and the fact that the frackers are able to continue unabated in fragile zones is unconscionable and despicable.
However, this says nothing about fracking outside of fault areas, which seems to be safe enough to pursue. This whole situation underscores the fact that there needs to be federal regulation concerning fracking in safe areas. This is important because it’ll also impact regulations in pursuing enhanced geothermal technology, which could be an excellent and robust source of environmentally conscious energy production.
icarus62 is probably, sadly, prescient. Some nutcases will find the “good” side of poisonous fracking and earthquake frequency. Now OK is ahead of CA in earthquakes! Isn’t that just wonderful?
I don’t think Oklahoma wants more earthquakes, but how much damage are they actually doing? They’re pretty small quakes.
My main objection to the fracking is not the earthquakes, but the fossil fuel extraction that they help with.
You’re right 99.9% of the time, Tamino, but your truncation of the context that made a relevant point was simply dishonest.
[Response: We disagree about your comment making a relevant point, but that’s not a big deal. What is a big deal is that your reaction was to conclude that I was “dishonest.” If you’re so upset about having a comment deleted that you have to resort to that kind of sleazy smear, then your moral compass needs repair.]
Remarkable data! Thanks for another lucid post. I’ll be sharing this one.
Also, I want to let you know I appropriated a graph from this site for use in an online article. It’s attributed of course, and I know you’ve been welcoming in the past of such reuse, but let me know if you’re not OK with this one for some reason:
Check out these two excellent NYer articles on Oklahoma earthquakes (and politics):
Goodbye OK earthquake observatory
Tamino, the “dishonest” observation was dead-on-balls correct.
You said of my comment, “You started out by asking “do we really know enough to rule out a natural cause?”
That’s simply a lie, Tamino, and you know it, But you deleted what I actually said it so nobody else does.
Much closer to the truth, Tamino, as you KNOW, I started out by saying “While I suspect these OK earthquakes almost certainly are related to fracking — and should be treated that way until we know otherwise…”
[Response: I admit that what I quoted from you wasn’t the first words in your comment. But I did quote you. When you have to ask whether we can really rule out a natural cause, one wonders whether you actually read the post. Because that’s not a reasonable doubt.
And it’s not all I found unreasonable in your comment. Rather than have the thread go that way, I simply deleted it. My blog, I decide what goes, what stays. You disagree, that’s fine, I could be wrong. As for name-calling, shame on you.
Still my blog.]
Hey, the constructor said earthquakes are not entirley a bad thing. And the oil industry said fracking is not bad either!
Looking at the comments so far I wonder if note has been taken of Tamino’s very clear statement that the quakes are associated with wastewater-injection wells. We see far too often the jump from that observation to “fracking causes earthquakes” thinking. There are plenty of problems associated with the fracking process itself but that is not the topic here.
Injection of waste water from a number of activities, some not associated with oil or gas production, can boost earthquake frequency.
Your prolonged absences of late have been noted with some dismay, and I will continue to highly recommend Tamino to family and friends regarding climate and statistics. Only now with a caveat.
Tamino – any reason my post hasn’t gone up. Realise it was a little critical but it was balanced
Earthquake statistics are tricky Tamino — because of the huge dynamic range (the log scale), and because of the very strong spatial and temporal autocorrelations (many earthquakes happen close to previous ones, often relatively soon afterwards — minutes to years). I’ve little doubt your conclusion is valid here, but intraplate locations can transition from quiet to active for natural reasons. That’s usually in response to one or more large events though, and the spatial clustering is then a clear give-away. See for example the 1988 Tenant Creek earthquake in the Northern Territory, Australia. I’d be interested to see the Oklahoma spatial earthquake distribution for the last few years.
Has any harm resulted? Have damaging earthquakes become more likely? By now much? (Tough questions perhaps, but important ones. The frequency of small earthquakes generally correlates with the frequency of rarer, large and damaging ones, but usually with some location-specific asymptotic magnitude cutoff.)
It’s interesting to have a look at the change in distribution over magnitudes, too. The UGS site has a nice API to download a taylored data set without having to go through the GUI.
Here’s an animated histogram
Greg’s comment above led me to have a closer look at the data. Event parameters include magnitude type, which identifies type of measurement and analysis, and the depth of the quake. It turns out there were major changes occuring in 2009/2010, as can be seen from the depth distribution:
This change is not simply due to a changeover between magnitude types, as e.g. the magnitude type “mblg” has data over the entire time range and only recently gained improved depth resolution. I have no idea, whether or not this change influences the quake counts. As the 5.0 km peak also shows the strong increase, I am inclined to think that the strong increase in quake counts is real and not just an artifact of closer scrutiny.
This is also discussed at Daily Kos
without Tamino’s clear graphs but with much other information.
Quake software including HAZUS and open source possibilities are mentioned in the comments.
Gerg asks “Has any harm resulted?”
There may well be effects on foundations and pipes, but anyway Gerg what should we do, press on for the big one? See this comment for more:
Just recently I saw the expression “Dumb state death spiral.” I think it referred to economics, but multiple use is OK even if OK isn’t ;)
One issue with earthquakes in Oklahoma is that the buildings are not built to withstand them. A 5.0 earthquake in California will not damage anything. In Oklahoma there are pictures of the facings cracking and falling off houses from 3.0 earthquakes. In California it is required to have lips on shelves in laboratories to keep chemicals from falling off during earthquakes. In Florida (where I live now) lips are not used since there are no earthquakes. Houses in California are securely bolted to the foundation so earthquakes do not knock them down. In Florida they tie the roof on really tight to withstand hurricanes.
When you start to get many small quakes you wonder if a bigger one will come soon. Do you want to bet your house the big one will not come? I have seen reports that many people in Oklahoma are buying earthquake insurance. How much are they paying to subsidize the oil industry?
Huge departure from the norm in recent years. How can they ignore this?
I tried and failed to access the data for Oklahoma in a form that I could plot. Tamino, does it make much difference if smaller magnitude quakes are included? I was thinking that fracking should lead to smaller, local quakes (I’m in no way an expert), so I wondered if the stats were any different looking at smaller quakes (inclusive and exclusive of other magnitudes).
Depth is usually the least-well resolved parameter in earthquake data Bluegrue. The reason for the large peak at 5 km is that, by convention, that is the depth assigned to shallow earthquakes — so it means little. I’m not familiar with the Oklahoma situation, but it’s possible that the apparent improvement in depth data is due to closer scrutiny as you say, or closer scrutiny plus better instrumentation, or even just the larger sample of slightly bigger earthquakes available recently, for which depth may be better resolved. (Classically, depth is obtained from the double arrival at a far-distant seismograph, requiring a large enough earthquake to be resolved at that distance. Better instrumentation can resolve depth locally.)
The association of earthquakes with fluid injection came to notice in 1962, after an injection well (drilled to 12,000 feet in 1961) in Colorado began to be used for disposing of waste from the Rocky Mountain Arsenal. A number of eports on that situation were published.
“…press on for the big one” … I don’t reckon I suggested that Pete; the opposite. The last frame in Bluegrue’s animation displays the classical power law relationship between earthquake magnitude and frequency. Extrapolating that, increased frequency of small earthquakes simplistically implies increased frequency of larger, more damaging earthquakes (say M5+). Plotted on a log scale, there is usually some regional upper bound magnitude to the power law’s straight line. Where is that for Oklahoma? Does that still apply for fracking-induced earthquakes? More difficult and important question.
(BTW, as one who has designed dangerous things (large dams) in remote places with poor earthquake data, I can say that the modern approach is that there is NOWHERE on earth that has “no earthquakes”. Anyone designing hazardous things for less than say a close-range M6.5, even in the quietest region, is negligent IMO. See that Tennant Creek earthquake — an M6.7 in a previously silent place with decent earthquake data.)
That’s this one, I presume?
Guess not, none of them are 6.7.
Yep … Geoscience Australia has the largest in the swarm at M6.7: http://www.ga.gov.au/scientific-topics/hazards/earthquake/basics/historic#heading-8
“…a 35km long fault scarp with up to two metres vertical displacement was formed.”
It’s been a while since I studied this, but I seem to remember the lots of little quakes correlating to an eventual larger quake was because as stresses built up small movements could dissipate the stress for a time, but eventually the little slips (low-hanging fruit?) were exhausted and the continued stress could only be dissipated in a much larger tremor.
If this recollection is correct, it suggests that there is not a great need for small injection quakes leading to a larger quake. But that’s just a naive analysis based on one mechanism. Perhaps injection quakes lead down a different path capable of causing substantial quakes. I’d guess modeling this is no trivial matter…
I have to point out that most of these Oklahoma quakes are in northern Oklahoma where they’re associated with dewatering wells. When companies dewater a formation, they remove a lot of water before oil starts coming out with it. In the end, they produce a lot more water than they do oil and this water has to be disposed of, which is done in wastewater disposal wells (a relatively small set compared to the number of production wells there are).
Thus, there’s lots of water going into a relatively few disposal wells. Because of this large volume, the pressure wave can move miles from these injection wells and change the pore pressures around nearby faults. Of course, this can put faults already close to being critically stressed over the edge.
However, none of these dewatering wells has ever been fracked, so it’s misleading to say fracking is the big problem there. Despite what they say, Oklahoma isn’t a big hotbed of fracking. It’s been happening, but it’s a pittance compared to what’s going on in places like Texas, North Dakota, and Pennyslvania.
That’s not to say that disposal of fracking wastewater can’t cause problems, because disposal of any wastewater underground can cause seismicity, but it’s an overall oil and gas industry problem, not just a specific fracking problem.
A good article on Oklahoma with standard industry denial/ignorance. It’s actually very well established in the scientific literature that wastewater injection can cause quakes. But I’d wager this dick was never big on science in the first place considering he “never found a drop of oil through textbook geology.”
I also understood that part of the earthquake link with waste water injection is that the injected water can have a lubricating effect on existing fault lines. Geothermal energy production has similar issues associated with it and has been linked to small earthquakes.
Fracking has its issues but earthquakes isn’t one of them. Yes there is a link in that it is the waste water from fracked (and conventional) wells that is being injected and is likely causing the earthquakes but there are other ways to deal with fracking waste water which don’t cause earthquakes. RO is probably the best approach environmentally but it is more expensive and the waste sludge still needs to be disposed of.
“RO” being reverse osmosis…
RO creates several liters of waste water for every liter of fresh water it processes. The sludge is not very concentrated. It is also energy intensive.
“Fresh” is a relative term. In this case perhaps I should have used the word “raw” or “unprocessed”. You wind up with more polluted water than you started with (and it is more polluted). But you do get a trickle of drinking water…
Dang I screwed it up after 2 tries…
You still get a lot of waste water with RO. It takes a lot of energy.
I’ll stick with that.
arch, You can get much better recovery rates than that (over 70% for commercial applications) and the produced water is pretty good quality. Household units I understand have much lower efficiency due to pressure issues which is where your figures may be coming from). Energy use is also being reduced through new tech improvements. The sludge is more concentrated with these levels of efficiency.
Yes it does require energy but the advantage of using it with gas production is that there is a ready supply of energy close to the point of demand (i.e. the gas). Sure, it’s more expensive but it also has a useful byproduct that has value and where the key concern about fracking (injection related earthquakes aside) for most communities is the potential impact on water supplies, being able to provide water to them at little or no cost may be a positive thing for the industry.