How to make an earthquake

You may not realize it, but I majored in “Mad Science” in school [1]. And one of the things that every mad scientist needs to know (other than “never tell Bond your plan” and “last door on the left”) is how to make an earthquake. Being a genially evil sort, I thought that it might be fun to share that information with you in light of Oklahoma’s recent series of quakes [2].

To start with, you have to remember that the surface of the Earth is broken into pieces that are in constant motion. The pieces are proportionally about as thick as the wrapper on a Tootsie Pop and are known in the biz as “plates”. The plates [3] slide around on a solid but easily deformed part of the mantle [3]. Sometimes the plates move apart, creating ocean basins [4] and the world’s longest mountain chain [5] and earthquakes. Sometimes, the plates move side by side, creating strike-slip faults [6] and earthquakes. And sometimes they move together, forming mountains [7] and earthquakes.

Just in case you missed the common theme of plate tectonics, I’ll spell it out: e-a-r-t-h-q-u-a-k-e-s! You see, when plates change shape (as they must when moving around), they become stressed and store up strain [8]. And when rocks have too much strain, they snap! and you get an earthquake [9].

Now this is all well and good if you want to wait a couple of hundred years between earthquakes. But what if you need one now? That’s where science is your friend. You see, we learned how to turn earthquakes on and off back in the 60s. And, as most such things come about, it was discovered by the military by accident. You see, the Army had all of this lovely toxic sludge left over from making chemical weapons and needed to put it somewhere that civilians weren’t likely to stumble over it, so they decided to inject it into the ground. However, after pumping 165 million gallons of the stuff into the ground, the folks living in Denver noticed that they were having earthquakes. A lot of earthquakes. A lot of fairly big ones, where they had never had any before.

Of course the Army denied that it was at fault. And they even sponsored an experiment to prove that it wasn’t them. Naturally, the experiment showed that it was; the Army had triggered the earthquakes. But how?

In order to understand what happened, we need to think about what happens along a fault line. There is shear stress trying to create movement along the fault line and there is friction which works against the shear stress. As long as the friction is greater than the shear stress, there is no movement and no earthquake. But the friction is created by the normal stress (which is at right angles, or “normal”, to the fault); if there were a lower normal stress, there would be less friction.

We can take this general idea and plot it up using a fairly simple mathematical relationship. The combination of normal and shear stress creates a “Mohr’s circle” of stress. The amount of shear stress needed to create an earthquake increases linearly with the amount of normal stress, so we can plot a line separating the stable region from the unstable one. (Actually, this is two lines because of how the physics works.) The angle of the line depends on the amount of friction in the system; the higher the firction, the steeper the line. As long as the Mohr’s circle is totally within the stable region, no earthquakes can happen. But if something were to happen to reduce the normal stress and shift the circle to the left, we could get an earthquake.

And we already know what that “something” could be: water! You see, when you add water to a formation, it adds a “hydrostatic stress” to the situation. In effect, this is the force per unit area that would be exerted by a column of water extending all the way to the surface. That may sound like a lot of nasty physics, but you are already familiar with the idea if you’ve ever gone swimming. When you dive to the bottom of a swimming pool, you feel the pressure of the water above you. And when you float in the pool, part of the normal force of gravity is offset by the hydrostatic force of the water; that’s why you feel so light in a pool.

And the same thing happens in the rocks. The hydrostatic stress reduces the normal stress, so that the friction is reduced. It also reduces the shear stress, so our circle both moves to the left and gets smaller. But the stability line doesn’t change. So if there is enough hydrostatic force, we can reduce the normal stress enough to allow movement on the fault; we can trigger an earthquake!

OK, so this is just cool. Not only does it turn out that James Bond was right, but we could use this to trigger earthquakes at will. And it isn’t just injecting fluids that can do this; we have seen the same thing happen after dams fill up with water [10]. Unfortunately, this is not a good way to prevent earthquakes [11]; we can only trigger, we can’t avert.

And that brings us to the (literally) million dollar question: Was the Oklahoma earthquake triggered by fluid injection? You see, there are a lot of new wells being drilled in Oklahoma, taking advantage of the boom in “unconventional gas plays” [12]. And those wells typically “frack” or hydraulically fracture the formation in order to get more hydrocarbons out. Since they are pumping fluids into the rocks, could they be causing the earthquakes?

Probably not. You see, the amount of fluid used in fracking is on the order of five million gallons. Remember that the Rocky Mountain Arsenal Disposal Well pumped in 165 million gallons, or 33 times the amount pumped in a fracking operation. And the water in the Rocky Mountain Arsenal Disposal Well stayed in the formation, whereas the water in fracking goes in, creates fractures, and then flows back out. And then there is the question of depth. Most earthquakes happen several miles below the surface; the one in Oklahoma was 3 miles deep. In general, most shale plays are shallow and less than one mile deep. Thus, the water from a fracking operation is too little, too brief, and too shallow to cause a typical earthquake [13].

So, if you want to be an evil scientist, you now know all you need to know in order to trigger an earthquake. Just let me know before you do, OK? I have some insurance that I want to buy…


[1] Actually, in Physics and then Geophysics – but what’s the difference, really?
[2] Oklahoma is actually a fun place, seismically speaking. They have isostatic rebound from glaciers in the northern part of the state and fossil stress from the alacogen in the southern part, giving them an average of an earthquake a day! Of course, most of them are just tiny little things (about 1/1,000,000th the energy of a Fukushima event), but you have to start somewhere.
[3] Which is known as the aesthenosphere, which is Greek for “weak layer” [i].
[4] Like the Gulf of Mexico, the Sea of Cortez, the Red Sea, and the Atlantic Ocean.
[5] The Mid-Atlantic Ridge, which runs from the North Pole nearly all the way to Antarctica.
[6] Such as the North Anatolian Fault, which gave rise to the earthquake in Turkey in October, and the San Andreas Fault, which gave rise to the Great Earthquake of 1906 [ii].
[7] Such as the Appalachians, the Rocky Mountains, the Andes, and the Himalayas. In general, if it is a mountain range, it was probably formed this way [iii].
[8] This is one of those science-geek type bits of terminology. Stress is the force that you put on an object. Strain is the change in the object’s shape. You can think of it as being like a person and food. If a plate (person) is put under stress (eats too much) then you’ll see strain accumulate (changes shape).
[9] Here’s a simple demonstration of how accumulated strain makes an earthquake: Take a piece of uncooked spaghetti by both ends. Start pushing your hands together so that the spaghetti begins to bow (in science: “stress the spaghetti so that strain accumulates”). At some point, your hands will get too close together and the spaghetti will snap (in science: “the accumulated strain will exceed the compressional strength of the spaghetti”). That snap is an earthquake.
[10] Actually, that is the more common way to create this sort of earthquake. Fluid injection tends to involve small amounts of water over short periods of time, where impoundments are by definition large volumes of water being held for long periods of time.
[11] Nor should we use it as a way to avoid a big earthquake by having a lot of small ones. Think about the math for a moment: a magnitude 7 earthquake has the same energy as 33 magnitude 6 events or 100 magnitude 5 ones. So, if we have a magnitude 7 every century, we’d have to have a magnitude 6 every three years, or a magnitude 5 every month or a magnitude 4 every day. It is just easier and simpler to prepare for the magnitude 7 than to constantly repair from a magnitude 4. And it is a lot less nerve-wracking!
[12] In other words, oil and gas that are trapped in shale instead of being trapped in sandstone (most commonly used) or limestone (less commonly used) [iv]. Most oil and gas starts its life in shale and migrates to nearby sandstone. But a lot of that hydrocarbon never makes it to the porous (and therefore easily drained) sandstone. Until recently, it was thought that we couldn’t get them out of shale. But now we can drill multilaterally and directionally and get more out of one borehole than we used to get out of ten. Ain’t progress grand?
[13] That’s not to say that fracking hasn’t caused the occasional tremor. There have been a few cases recorded in the literature that showed a fracking job triggered a local event. But they are the exception and not the rule.

[i] Who says science is tough?
[ii] Only a geophysicist could call a disaster that killed 3,000 people “great”…
[iii] The only exceptions are for sea mounts and island chains like Hawai’i that may have been formed by hot spots. But those are just weird, so we’ll ignore them.
[iv] There are also other types of unconventional hydrocarbons, including deep gas (gas that is really, really deep {duh}), tight gas (gas locked in non-permeable formations), and clathrates (methane-ice crystals). But the main play is in onshore shale formations. For now, anyway…

8 thoughts on “How to make an earthquake

  1. I knew if I held my breath for 24 hours or so you’d answer my question. I was pretty sure that 3 miles was too deep to be caused by fracking, but it’s always good to have an expert confirm my suspicions.

  2. Every time I hear fracking, I don’t think BSG, I think “John.”

    “And when rocks have too much strain, they snap! and you get an earthquake.” Through syllogism, this makes me a rock?

    Speaking of the Army proving themselves guilty, did you hear about the study funded by conservatives (probably The Brothers Koch) to prove global warming isn’t influenced by man that proved man has added to it? Thought of you then, too.

      1. I am so relieved you knew what I meant because “lil ol me” has such a hard time. I read these things, understand them as best I can but when I try to “disseminate” the info to friends or coworkers, they’ve NO clue what I’m saying and then I can’t even recall where I read it. ::sigh:: If I were smarter, I’d be smarter!

  3. I didn’t suspect the quakes were drilling related until after I read your blog and the report you linked to. That report basically says all evidence (from January, not recent quakes) points to the drilling, but then pulls the escape pod and says we can’t really know for sure so don’t blame the drilling. Huh?

    Wouldn’t the newer practice of horizontal drilling be much more likely to allow plate slippage versus the vertical bore? And isn’t it quite possible that changes at a 1 mile depth can give just enough wiggle room for plates to shift at a 3 mile depth or even deeper? (assuming horizontal holes and not verticals) I mean, it just doesn’t pass muster to say the OK drilling only go 1 mile, and the earthquake was clocked at 3 miles in, therefore they are not related. That’s not really how physics works, as you so graciously explained with the Mohr’s circle.

    Just really curious about all of this. Of course, blaming big oil for earthquake is TONS more fun than blaming HAARP, plus we get to say fracking a lot more. Or maybe it’s a 2-for-1 deal and the phrase “play that fracking HAARP” will become part of pop culture. 😉

    1. That’s not really how physics works, as you so graciously explained with the Mohr’s circle.

      In this case, it is. Remember that hydrostatic pressure is a diffusive process and so an increase in pressure at one point does not equate to an increase in pressure throughout the volume.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s