Water injection to defuse big quakes

  1. In Science 4 July 2014, there is an item that I saw summarised online:
    Injection wells blamed in Oklahoma earthquakes.
    ...The pressure can reduce forces acting to keep faults locked and trigger earthquakes...

    Now, I don't know how likely this is, and the discussion was commendably non-committal, but the idea is not fundamentally implausible. Let us assume that there is a strong causal effect and consider the implications.

    Right. Either there is
    1: essentially no significant accumulation of stress energy in the region, or
    2: an alarmingly growing, but still moderate amount, or
    3: a dangerous accumulation, a BIG ONE in waiting. OK?

    In the first case, no one cares and the tremors are irrelevant.

    In the second case surely the one thing we would want is to bleed off the energy ASAP and as effectively as possible, to prevent a dangerous accumulation?

    In the third case, suppose that we not only refrain from injecting water, but pump out all the excess we can find; that might delay the BIG ONE no?
    But to what effect? So that all the politicians and big businessmen could quickly quake-proof every big city around???

    I don't THINK so!

    And the longer the delay, the bigger the BIG ONE!

    And it might be a bit unnerving to tickle the dragon's tail by injecting water, but every joule we lubricate out of the BIG ONE and dissipate as tremors, will reduce the impact when it does hit. In fact wouldn't it be conceivable that if we could make the entire West Coast feel like a diesel truck on a country road for a year or so, we could bleed the BIG ONE so effectively that when it does hit, everyone complains about the fuss and wants to know why these scientists keep shouting wolf?

    And although it might be too difficult to fend off the BIG ONE and all that we did would be to hasten its coming, then surely the sooner the better, because the later the bigger?

    So why complain about a bit of water injection or any amount of water injection at all?
    Why not start injecting a nice big slice of the Pacific in the hopes of ending the suspense with a smaller imminent BIG ONE than a deferred REALLY BIG ONE, hm?

    Any comments?
  2. jcsd
  3. davenn

    davenn 4,356
    Science Advisor
    Gold Member
    2014 Award

    Hi Jon

    you have overlooked one other and the main reason for the Oklahoma ( and other regions where fluid injection is done) quakes

    The term is called fracking .... This is where fluid is injected into the rock fracturing it to release natural gas and oil to be recovered.
    The quakes are a direct result of the rock being fractured

    The other fluid injection process is purely for waste disposal, but the result is the same, causing the fracturing of the rock

    I have fellow seismology friends in the USA and a couple of them in the OK region that are regularly recording quakes caused by the fracking and waste disposal processes. Its a worry and a subject that is regularly discussed on our Public Seismic Net email forum.

    here is a reference to fracking in CA and the calls to ban the process ....

    In response to concerns about this issue Santa Cruz County became the first California county to ban fracking and any future exploration that might cross county lines such as slant drilling from other counties.

    http://rt.com/usa/160316-santa-cruz-ban-fracking/ We have a very active and progressive Board of Supervisors that really stay on top of the real issues.

    here's another reference for you to have a look at ....


  4. Simon Bridge

    Simon Bridge 15,474
    Science Advisor
    Homework Helper
    Gold Member

    You are saying that blaming injection wells, as in the article, is silly on all levels?
    I'm going to agree - scientifically.

    Socially and politically there is a blame + litigation game going on for short-term gain.
    This looks like the usual snafu to me.

    ... there is the fracking issue though... did the article cover that?
  5. davenn

    davenn 4,356
    Science Advisor
    Gold Member
    2014 Award

    Unfortunately its really a bad idea and false logic.

    A few dozen smaller events isn't going to fend off a major event
    Consider this ....
    10 x Mag 2.0 events = 1 x M3
    10 x M3 = 1 x M4
    100 x M3 = 1 x M5
    1000 x M3 = 1 x M6
    10,000 x M3 = 1 x M7
    100,000 x M3 = M8 ( 1906 San Francisco = M7.9)
    1000 x M5 = 1 x M8

    You definitely don't want 1000 x M5 events occurring anywhere around any urban area
    The damage is going to be significant

    That's going to result in $billions in lawsuits for deaths and damage against whoever allowed this hairbrain scheme to be initiated

    Since we have no real idea of how big a quake will be caused by lubricating a major fault line like the San Andreas Fault up through California. It may be that the section that was decided on to start the process was on the verge of letting loose and suddenly you have a M8 event starting that you cannot stop and all hell is going to occur

  6. AlephZero

    AlephZero 7,248
    Science Advisor
    Homework Helper

    Whatever you "blame", something has clearly changed, from a statistical analysis:
    From http://www.usgs.gov/blogs/features/usgs_top_story/man-made-earthquakes/:
    From http://www.sciencemag.org/content/341/6142/1225942 it seems nobody even knows what is being done, at the right level of detail to do any modeling:
  7. davenn

    davenn 4,356
    Science Advisor
    Gold Member
    2014 Award

    Not sure how to take that, Simon ??

    There is no doubt that fluid injection induces seismicity
    You only need to ask my fellow seiosmo guys. They can tell you when the injection processes start and stop by watching the coming and going of recorded events

    The real worry is if a significant event is induced, say an M5+
    M0 - M3's can be lived with, bigger ones could cause damage.

  8. mfb

    Staff: Mentor

    Smaller earthquakes also have smaller areas of influence - out of those 1000 M5 events, most won't influence you at all. If you could produce 30 M3 events per day in a controlled way (hypothetical example), you would get rid of the energy of an M7 within a year or an M8 within 10 years (so basically all the energy released in major earthquakes), the influence range would be limited to the fault line and damage to buildings would be rare.

    Even a (somewhat) controlled release of the same magnitude as a "natural" earthquake could be interesting, then you can prepare for the results.
    Sure, there is no way this would get allowed...

    @AlephZero: What is the detection rate for those earthquakes? Did they find all M3 quakes in 1970?
  9. davenn

    davenn 4,356
    Science Advisor
    Gold Member
    2014 Award

    The whole problem is.... you have no control over what the size of the quake you produce is going to be or of its exact location. You have no way of knowing if the next quake you set off will be a M2 a M3 or the big M7 that's been getting closer to striking and you have just changed the stress field patterns just enough to tip it over the edge
    M5 quakes can cause a lot of damage ... .as was widely experienced in my old home city of Dunedin, NZ, back in 1974. It was the quake that got me into seismology.
    As I said earlier, you really don't want M5 quakes going off anywhere near an urban area

  10. davenn

    davenn 4,356
    Science Advisor
    Gold Member
    2014 Award

  11. Borek

    Staff: Mentor

    You are perfectly right on some level, at the same time I have a feeling you are ignoring the big picture - if the energy is there, it is going to be released. And the later it hits, the larger it will be. So while yes, series of smaller ones is dangerous, the final one will be even worse.

    If the injections can work (which we don't know), it would be a matter of choice. Not doing anything to avoid smaller ones is a sure way of having to deal with a huge one sooner or later.
  12. Phew! I seem to have tickled the tail of quite a few dragons! :bugeye: Thank you folks for your thoughtful responses. Clearly however, I do need to clarify, not so much the points I intended, as those that I did not intend. I won't quote the texts, because they overlapped multiple responses.

    1: I specifically do NOT address fracking as a theme. I am deeply nervous of fracking, not so much because of its short-term potential for harm as because it seems to be heading for one hell of a short-term energy bubble, followed by a long-term energy crisis. (But don't get me started! :rolleyes:) However, the main reason for avoiding the topic here is that it is too fraught and off the earthquake point.

    2: I do NOT address the question of whether injection promotes quakes; I accept that it could, but am not equipped to deal with the technical aspects on a quantitative and analytical basis. Even if it could be shown that injection has no effect, the same question remains in force as long as it can be shown that SOME accessible process, injection or not, exists that can promote quakes.

    3: I do NOT discuss the political/commercial considerations. I accept that even if we could bleed off the energy with imperceptible magnitude 2 quakes and even if nothing bad happened, it got into the press, there would be hell popping as all the office chasers and ambulance chasers and notoriety chasers and snake-oil merchants and similar social parasites got word of it. I limit myself to the question of the rational course of action in a region like the USA west coast, or parts of Japan or a few similar places: how much better might it be to sit on the lid of the cooker, than let off steam, and what would it cost and how fast should we let it off?

    4: @Simon & others, I do not have access to the original article but only to an on-line teaser and I don't remember fracking being mentioned one way or another anyway. My thanks to Dave and Aleph for links to more detail that added a bit more substance to my comprehension, though without affecting the fundamental point, one that does not affect me personally, living, as I do, in an area of very low seismicity. However, the real question is independent of all that. The fundamentals are:
    a: There are various regions, possibly most dramatically the USA West, where everyone knows that there is a BIG ONE building.
    b: We don't know exactly where or when.
    c: There is a good likelihood that the longer it holds off, the bigger the bang when the elastic snaps. Certainly there is always a possibility that a slow, natural slippage could avert it, and in some places does, and rationally, even along the western Rockies we know that huge amounts of energy regularly are leaked in quakelets, even in sizable quakes, but afaik, the general trend is towards continuing accumulation of ever-increasing amounts of strain energy.
    d: The longer the accumulation the bigger the bang to expect, and the greater the probability of a cataclysm rather than a disaster or even a mere inconvenience.
    e: Irrespective of politics etc, even if tickling the dragon leads to a totally unacceptable disaster, the rational expectation is that to delay the blast simply would amplify that disaster.
    f: Whether we could do better to deal with a much larger quake (say one to two magnitudes larger) after twenty years, during which we prepared for the blow by quake-proofing everything from hen houses to the Hollywood walk of fame, or to deal ill-prepared with the consequences of a much smaller, but still disastrously large, quake right now, is something we could model, but not predict reliably. But I know how I would bet!

    5: I like mfb's remarks: "Smaller earthquakes also have smaller areas of influence - out of those 1000 M5 events, most won't influence you at all. If you could produce 30 M3 events per day in a controlled way (hypothetical example), you would get rid of the energy of an M7 within a year or an M8 within 10 years (so basically all the energy released in major earthquakes), the influence range would be limited to the fault line and damage to buildings would be rare."
    I admit that I had overlooked the significance of the range, and had in effect assumed that the only variable was intensity.

    6: @Dave, you said i.a. "The whole problem is.... you have no control over what the size of the quake you produce is going to be or of its exact location. You have no way of knowing if the next quake you set off will be a M2 a M3 or the big M7 that's been getting closer to striking and you have just changed the stress field patterns just enough to tip it over the edge"
    and I agree, in fact, even as a non-seismologist I had anticipated this. But the thing is that it would take less frequent circumstances for your M7 to anneal to say a bunch of M4s if left to itself; it would be likelier either to blow as an M7 anyway, or to build itself up to an M8. IOW, I am arguing, and inviting refutation, that no matter WHAT happens if you grease a possible quake, it was likely to happen anyway, only a bit later and probably worse or a LOT worse. In fact, if a really big quake accumulates, it is likely to entrain a lot of smaller quakes that would have occurred separately and relatively benignly if pre-empted. I do not represent this as inevitable, but I sure reckon it is the smart bet.

    7: And if you do prick the balloon early, you stand a good chance of not releasing large parts of the event that had not yet matured, making it go "Ssss..." instead of "Bang!" so to speak.

    8: As I see it this is one of those cases where the public interest is best served by grasping the nettle and taking what comes, but it also is a case where, no matter what happens, whoever throws the switch need not fash himself seeking hero badges from those he had benefited. So forget trying to persuade Big Business or City Hall to seek for favourable publicity or re-election that way! :biggrin:

    OK, anyone: what would be more rational than throwing the switch, no matter where in the cycle one starts. but as early and often as possible?
  13. davenn

    davenn 4,356
    Science Advisor
    Gold Member
    2014 Award

    no, not ignoring the big pic :smile:
    just looking at the current realities of not being able to control the sizes of the events that may be produced. As in that last article, in one situation a M 5.7 was produced. " In 2011, a small number of people were injured and 14 houses were destroyed in the town of Prague, Oklahoma by a 5.7 tremor".

    They have no way of knowing exactly what is occurring several km's down. I see it as just way too risky to "play with fire" like that.

  14. I agree with dave on this. How exactly do you propose to trigger 30 Mw 5.0 events/day? We don't have the knowledge/technology to do it. Who would be prepared to risk being held responsible for triggering a destructive Mw 7.0+? The incentives are not there. If you get it right, you're just doing your job, if you get it wrong you've killed and likely face severe penalty .
  15. Borek

    Staff: Mentor

    No doubt about that part. But we are surrounded by technologies that didn't exist before, aren't we?

    And if you do nothing, kilodeaths when the huge one comes are just not your problem? I guess that means you don't live in Italy :wink:

    I do believe it makes sense to consider the idea and to run it through some checks with people knowing physics behind. Could be it won't survive discussion over a beer stage for reasons I am not aware of. Could be the cure would be worse than the disease. Could be it was already proposed/analyzed/rejected/put on shelf till we get more data. But rejecting the idea because someone can get hurt when we know people WILL get hurt, is putting head in the sand.
  16. Agree. But how do you incentivize someone to do it? The reward that you might save lives in the future is not enough IMO, at least not on a personal level, because you have to offset that with the risk that you could inadvertently take lives.

    Fair enough if we had the technology to make it 100% risk free. But I would argue it could never be risk free until we could 100% predict earthquakes. In that case we could take other strategies, like evacuation, to avoid disaster.
  17. AlephZero

    AlephZero 7,248
    Science Advisor
    Homework Helper

    I have no idea, but I assume the US Geological Survey would know if they suddenly changed the detection or reporting system in 2009.

    You don't need any high technology to detect a M3 earthquake. Members of the public can feel them, though they don't usually cause much damage. I've experienced several in the UK.
  18. mfb

    Staff: Mentor

    @billiards: That's the political side. Certainly a big issue for a possible practical application, but not the scope of this thread I think.

    Sure, the 30 M3-quakes per day are a purely hypothetical example, but the general idea is independent of it: earthquake damages are nonlinear with the released energy. For a given energy to dissipate, some intensities are better than others.
    I would expect that damage rises faster than linear in a large energy range until it reaches a point where many buildings get destroyed - afterwards the curve changes because you cannot destroy a building more than once.

    There is another nonlinear effect afterwards:
    A few broken houses are easy to fix, you have enough hospitals to care for a few injured persons and the remaining infrastructure is in place to clean up.
    If a town gets destroyed and thousands (or more) get injured, your local hospitals can't handle it any more, you can get problems with the supply of clean water, food, shelter, with diseases and so on.
    1 person likes this.
  19. I guess the other technical issue is what effect does doing this have on the stress regime somewhere further down the line? Are you not just shifting the problem onto someone else's doorstep?
  20. mfb

    Staff: Mentor

    How would that be different from natural earthquakes?
  21. @Billiards. "...shifting the problem onto someone else's doorstep? "
    Reasonable thought, but not persuasive, I'd say. Once energy is spent, it is spent, so whatever has been removed by rattling one backyard cannot harm another backyard. Sure, distorting (or removing the distortion from) the rocks under my yard might permit (or force) your rocks to perform somewhat, but they would do so anyway if the BIG ONE struck as a result of my doing nothing, and your rocks would then have the un-bled energy to deal with all at once. They would perform more violently from a full sledgehammer blow than from energy bleeding from your neighbour's rocks a jiggle at a time. Bottom line remains: it looks very much as though ANYTHING we do is likelier to be beneficial than whatever we neglect to do.
Know someone interested in this topic? Share this thead via email, Google+, Twitter, or Facebook

Have something to add?