The Art of Fracking a Volcano II

The Krafla Power Plant was built and taken online during the Krafla Fires series of eruptions. It did not cause the eruption.

The Krafla Power Plant was built and taken online during the Krafla Fires series of eruptions. It did not cause the eruption.

In this part of the series I will take a look at the risks of harvesting geothermal energy. Let me begin with stating that so far there has been no large accident caused by harvesting geothermal energy. On the other hand, there is no such thing as totally safe ways to harvest energy.

We all know the dangers of nuclear power plants, oil and coal destroys the biosphere and accidents in mining and drilling operations have been many. Hydroelectric dams have failed and drowned people, wind turbines have catastrophically failed and killed people and large scale solar arrays changes the heat pattern around them killing vegetation.

So then, what are the dangers of geothermal energy? Well, one big risk is actually economical. It is risky business. First you need to spend a lot of money before you even know if you have an economically viable power plant. But, that is not what this series of articles is about, it is after all only money and jobs involved in that risk. Not the lives of people.

Below I will try to give as unbiased information on the risks as possible, both on real and fictional dangers of harvesting geothermal energy.


Geothermal energy causes eruptions

This is the hole that drilled into magma. No eruption happened except that geologists made somersaults and played lip banjo out of joy.

This is the hole that drilled into magma. No eruption happened except that geologists made somersaults and played lip banjo out of joy.

The idea is that drilling into active or dormant volcanoes would cause an eruption to happen. In one scenario the magma would start to squirt out if the drill accidentally went into a magma reservoir and a new volcanic vent would be born.

An eruption occurs as the pressure in the magma reservoir goes beyond the threshold of what the rock containing the magma can withstand. If that happens there is nothing that can stop the eruption. And if the pressure is not high enough there will just simple be no eruption. If a borehole enters a magma reservoir the magma would indeed trickle into the borehole where it quickly would lose temperature and solidify, forming a perfect plug.

This has actually happened in Iceland during the International Deep Drilling Project, exactly as I described it. The only result was that a few volcanologists did somersaults while playing lip-banjo out of joy since they received a set of nice samples of magma directly out of a magma reservoir. Otherwise nothing happened.

Another proposed idea is that the harvesting of energy would cool down the lid of a magma reservoir so that it would be able to withstand a higher magma reservoir pressure. This in turn would cause a larger than natural eruption to occur at a later stage. This one is intriguing and slightly plausible actually, but as we will see it is not likely to happen.

To get a good answer we only need to look at the last eruption out of Bárdarbunga volcano. A magma reservoir is 3-dimensional, at Bárdarbunga the lid of the volcano was situated below 700 meters of ice functioning as a thermal sink lowering the temperature over time on a scale that any geothermal energy harvesting would never be able to do.

The ice there often quenches proto-eruptions from going into large eruptions and slowly the pressure in the chamber rose. In the end the pressure went beyond the capacity of the surrounding rock to hold and the side ruptured and a very long dyke formed and the eruption at Holuhraun followed.

At Bárdarbunga nature did the cooling and hardening on a scale that is quite honestly well beyond what even a bunch of the world’s largest geothermal power plants could ever do. From this we can learn that a power plant situated on top of magma reservoir would minisculy harden the roof of the reservoir, and minisculy increase the risk of a flank eruption.

If we turn to physics instead for an explanation we will see that any eruption is caused by an increase of energy (thermal) and that what a geothermal power plant does is that it leaches out that thermal energy on a minute scale.

Now it would be simple to see the geothermal power plant as a safety valve delaying or hindering an eruption. But the amount of energy in even a small volcano is so high that we are talking about fractions of a second of delay at best. But, at the same time the physics way of looking at things clearly proves that it does not raise the risk of an eruption happening.


Geothermal energy harms nature

It is true that geothermal energy can be harmful for nature. At the same time it is the safest way of harvesting energy for nature. This does not sound like a compatible set of statements does it?

Well, you are either drilling into a very hot water body or you are pumping down water into very hot volcanic rock. The water body would be filled with minerals and gasses, and if you pump water down it would also dissolve the same minerals and gasses. And a lot of those minerals and gasses are poisonous. So, yes a lot of harm could come out of this.

Both the drilling into a natural water reservoir and the pumping down of water gives polluted water as a result. If you use an open system and let that water come out into the local biosphere you basically have a toxic spill.

The solution is to use a closed loop system and pump the same water down into the natural reservoir, or into an injection borehole. Basically you are sending the minerals and gasses back where they came from. You are not adding anything new to the rock down below, you are just returning what has come up.

Another good thing with repumping the water is that the risk of land subsidence disappears. This is especially true if you are drilling into a natural hot water reservoir.

Has there ever been a large spill of toxic water from a geothermal power plant that uses closed loop technology? No, there has not been one. But, there has been a spill of cooling water on a large scale. This was turned into a tourist attraction by the Icelanders, who happily continued to pump hot cooling water into a manmade lake creating the world’s nicest industrial accident.

To be honest, at any other place the Blue Lagoon would have been a disaster, but in barren Iceland it was a boon for the tourist industry. In the end the industry learned a lot from that mishap.


Borehole blow out

Reykjanes Power Station. In front is the Gunnuhver geothermal field. The pock marks are from natural blowouts.

Reykjanes Power Station. In front is the Gunnuhver geothermal field. The pock marks are from natural blowouts.

This is exactly what it sounds like. In rare instances there has been sudden increases in steam pressure that has suddenly caused the borehole to blow out. This can be caused by sudden fracturing of bedrock so that over-pressurized steam ruptures the top of the borehole where it connects to pipes leading to the turbine.

The same thing happens in nature. You can have a gently steaming hydrothermal well that suddenly explodes as the underlying water comes into contact with extra hot material such as magma.

The danger here is not the general public, but it is indeed a work environment hazard.


Geothermal energy causes earthquakes

Geothermal plant boreholes with muffler and distribution domes.

Geothermal plant boreholes with muffler and distribution domes.

Absolutely true however you look at it, but there are actually advantages to some of these earthquakes.

First rule of geothermal power harvesting, do not drill into an active faultline. I will get back to this.

If you are harvesting dry steam out of a natural reservoir of water you must pump the residual water back into the reservoir, otherwise you will cause subsidence. The subsidence would cause damages to houses and roads. Subsidence would also cause earthquakes in the area. On the other hand, pumping back water will also cause earthquakes. These earthquakes would be small, generally it would be felt as a vibration and not as distinct earthquakes. Maximum size earthquakes would be at around M3 in this context.

If you are going to harvest from dry bedrock you first need to fraction the bedrock at the desired depth to make it more porous for the injected water. This would cause a series of earthquakes ranging up towards a maximum of M3.5 and when you repressurize the well with water you would get tremor or discrete earthquakes at a maximum of M3.

Neither of these earthquakes would be dangerous, but they would be a nuisance to the local residents. It is therefore essential to inform them, and that you do this during daytime to not hamper normal sleep patterns. Also, and this is important, the engineers can at any time shut down the injection of water if they feel that the injection is causing undue amounts of earthquakes for the local population or lower the amount of water injected per second.

Now back to the first rule of geothermal power harvesting. It is very hard to avoid doing this close to, or in a faultline, since almost all volcanoes exist at faultlines. In Iceland geothermal energy is harvested at faultlines or even triple-junction faultlines. This practice causes tectonic earthquake swarms as pressurized water enters the faultlines.

This can’t be good can it? No and yes is the answer to that. Yes you risk having series of earthquakes from the faultline as the pressurized water lubricates the faultline. Because it is true that the water would cause the rock on both side of the faultline to slide easier and this causes more earthquakes. On the other hand, all of these earthquakes releases energy that otherwise would have been dispersed in one large earthquake.

Yes, water-injection into a faultline would cause hundreds to tens of thousands of earthquakes ranging from M1 to M5. But on the other hand neither of those would cause widespread destruction. And it would lower the energy level of the next big natural earthquake, or even hindering a big natural earthquake to occur at all.

One thing is though clear, no large earthquake (M6+) has ever been caused by water injection. The reason for this is that the water injection is taking place locally and a large earthquake happens as a long distance of the faultline rips open.

In the end one thing is clear about this last issue, more research needs to be done in the future. Only science can give us the answers needed, and at the moment scientific research seems to be in favor of lubricating faultlines even though the subject is highly controversial.


147 thoughts on “The Art of Fracking a Volcano II

  1. Cool. Good article to stumble on. 😀

    Re: Triple junctions. Already the most seismic active regions due to the naturally unstable fault geometries. The size quakes that triple junctions naturally produce are orders of magnitude greater than what hydro injection can cause. Witness the geysers geothermal plant in California. The nearby mendecino triple junction and plate shard easily throw M7.2 quakes all by themselves. Far stronger than what the plant has ever produced.

    • My point is that fluid injections can release an earthquake even in a triple junction if you inject into them, but the earthquake will not be more powerful. Instead it is likely to be smaller than it would eventually have been.
      Due to lubrication it would be more likely that you would cause a series of smaller earthquakes instead and that the earthquakes would come at an earlier point since the lubrications would help to overcome the natural “locks”.
      Of course this would take place locally around the injection site.
      Take for instance a 300 kilometer long faultline. If you drill into the middle of it and set of a series of smaller localized earthquakes it is quite likely that when the “big one” starts to rip down the faultline it would end at the mid point thusly halving the maximum capacity earthquake.

      This is though a bit hypothetical, I am not suggesting drilling a hole every mile down the San Andreas to inject fluids… I do not think anyone would allow that regardless of it probably saving lives in the end. The 100 000 M3 to M5 earthquakes would scare the shit out of people and force a closedown of operations pretty much emediately.

  2. Super interesting article Carl, I look forward to future installments.

    I expect that you are waiting for someone to bring up the subject of the Indonesian Lusi mud volcano that is like the Eveready Bunny that keeps running and running:

    As many know, this mud volcano caused huge loss of farmland and significant loss of life.

    The article above indicates that the jury is still out whether the eruption was a natural event or did nearby exploratory drilling for natural gas (some 150 m from the mud volcano) have something to do with the eruption? The mud still continues to flow from Lusi.

    Well, I will not try to speculate as to the cause of the eruption but it does indeed point out that in some locations unexpected events can take place in the quest for geothermal energy.

    You have already pointed out some steps that have been taken to mitigate effects from drilling and water injection.

    FWIIW mud volcanoes often are found in areas of underground gas/oil formations.

    Occasionally there are spectacular incidents: AKHTARMA PASHALI , AZERBAÏDJAN – mud volcano spouts flaming petrol. See link below:

    Hobbes, thanks for your help re posting, PM me and I will explain.

    • Lusi was not a fluid injection. Well, not intentionally. What happened there is that a really large grabben was being supported by a fluid rich strata and the well released that pressure. It flowed back up the bore hole, which was poorly cased and got into a higher strata where it drove it’s way to the surface near, but not in the well bore.

    • Hello!

      Sorry, I will not go into Lusi in this context. It is something quite different from the subject at hand since it does not involve volcanoes or geothermal energy. It does not even involve fracking.

  3. According to my electricity bill, 40% comes from geothermal..i wish it was more! The power stations are located in Fogo volcano (Agua do Pau – Azores)

    But like @Carl wrote, geothermal has its issues, while its hard to get direct relation from geothermal exploration to dormant volcanoes activity, i have manage to find some direct and indirect situations:

    -In the 80s, when first geothermal holes was made, a seismic episode started, closing/open the hole changed the seismic activity intensity. Unable to find a paper confirming the relation.
    -In the 90s, a GPS survey to the volcano showed a deflation. “The deflation can be attributed to several processes, such as pressure decrease in a shallow magma chamber beneath Fogo, or it can be due to an extraction of hot water and steam by a geothermal power plant that is located on the volcano”
    -Mid 00s, large seismic episode in the center of the island, with epicenters located in Fogo caldera and East of it. Deformation was observed during this seismic crisis.
    -2009, when drilling a new hole near a small village, a pocket of gas was hit, destroying the geothermal equipment, quickly they tried to seal the hole. But the damaged was already done, some houses were inhabited, its not recommend to stay too long on the area.
    CO2 anomaly last december (RG4, stands for the location of geothermal hole that was sealed).
    2011, VT swarm that last some months, when subsided a LF swarm lasted for at least a year. GPS showed inflation.
    2011, Geothermal shuts down one power central, due analysis in one spring that feeds a near city (~6km) showed chemical and temperature differences. Its confirmed that geothermal fluid contaminated the spring.

    Maybe the seismic activity and deformation is normal behavior on the volcano, but maybe the geothermal exploration is having a lot of impact on the volcano (for better or for worse). Only on the last 20 years, it was installed monitoring equipment, so we dont know how the volcano behave before the geothermal exploration.

    • Hello Azost!

      Thank you for a very interesting comment.

      First I would like to say that you should never perform geothermal drilling in a system that is not well monitored before commencement of drilling.

      Your comment raises many questions from a professional capacity.
      1. Is it an open system or a closed loop system where they pump back the water that they use?
      2. Have they drilled into a underground reservoir and extract the water?
      3. Or is it a dry system where they have fracked the bedrock before injecting water?

      Judging from your comments I would say it is an open system that is utilizing an underground reservoir. That would explain the subsidence and the following seismicity. This is not the prefered way to operate a geothermal plant. Closed loop and reinjection is the prefered way as I have stated in the article.

      The gas pocket was a case of bad luck or bad preparation. Normally you would deploy a large array of seismometers to get a high definition tomography. This way you can avoid most surprises like that.
      What you normally do is to just keep the drill going until you have passed the gas pocket. Stopping is not good since that leaves a channel down open for the gas to continue in.

      Let it here be said that I am not familiar with this system and that I am not generally criticizing those who performed the drilling or operate the plant. I am just raising questions in light of this article and your comment.

      I will study this plant closer if I get the chance.

      If the well has really been affected they must be operating at very shallow depth indeed. I have never heard of a fresh water spring being affected by geothermal energy harvesting before.

      • Come think about it.
        If it is an open loop it could be water discharge that is released in such a way that it polluted the spring.
        Slake water that is not reinjected is a serious health hazard as I wrote up above. A power utility company letting slake water out into the aquatic system is basically causing an industrial accident.

        • From a brief search, i think it runs a closed circuit/loop.
          Translated from portuguese:
          “In the cicle, the organic fluid works on a closed circuit, never entering in contact with the geothermal fluid or with the atmosphere. The geothermal fluid, after passing through a pre-heater, its conducted to a line of discharge in a concrete box, which its connected to re injection well.”

          Only the CO2 incident its confirmed to lead to some seismic activity, according to a paper, the geothermal team mentioned tremor being felt in the area, it was confirmed by seismometers, intermittent tremor was recorded on the following days the hole was closed (degassing?). But cant confirm that most seismic activity (VT/LF swarms) in the following years are related with this incident. While there were recorded some events near the geothermal exploration (some very superficial), most are located in the other flank of the volcano, which might be related with regional stress or normal volcanic behavior.

          • Any VT/LF swarms would not be caused by geothermal energy harvesting.

            I would though like to say that one of the points of this article series is to point out real and imagined dangers with geothermal energy. And the most common dangers is to be found during the drilling of the boreholes. It is then you risk blowouts or hitting unexpected things. That is why drilling is surrounded by stringent safety protocols today.

            On average geothermal energy is though both safe and environmentally friendly.

    • Wairākei is carefully studied.
      Many of todays practices such as borehole safety procedures and reinjection comes from the knowledge that the power plant of Wairākei gave.

  4. Is there a reason why 1809 has to be in the tropics, rather than farther south in the Chilean Andes?

    • Viability mainly. At the time, the Andes had a lot of people hanging out that probably would have noticed. Also, with the requisite power, there would have been quite a bit of evidence from the ash-fall/destruction that something had happened.

    • As I understand it, because the 1809 sulfate ‘spike’ registered in both the Greenland and Antarctic ice cores which implies a source somewhere in the tropics. An eruption in the Chilean Andes would probably have registered only in Antarctica. Which is a shame, because the volcanoes in/near the Patagonian icecap are among the most remote in South America and would otherwise have been very much in the frame

    • Comparing how much sulphate was deposited in the south and in the north suggest a location of 5 degree south but this has considerable uncertainty. Sulphate from volcanoes at high latitudes tend to stay in their hemisphere. That from tropical volcanoes can spread to both hemispheres.

  5. Question: is it possible to have an eruption with a subsiding volcano and what does it mean (a subsiding volcano -> magma is compact, cooler (less likely to erupt?)

    • I assume you mean with a subsiding edifice. In this case, a lateral fissure could easily be fed as the mass of the mountain pushes down on the chamber and empties it.

      Compacting doesn’t necessarily mean cooling. One scenario that does involve crystallization, is that as the crystals grow in the mush, less room is available for the still liquid melt. This could increase the chamber pressure to the point where it can overcome the hoop strength of the chamber and propagate a dike. As a general rule, hoop strength would be less at the top of a chamber and the dike would tend to work it’s way to the surface as the cracks would preferentially be towards the weakest part of the rock. The hoop strength of the chamber would depend on the tensile strength of the rock augmented by the lithostatic stress of the overburden.

  6. And again another question, sorry. For a large – very large eruption, do we need a stratovolcano. They are talking about possible eruption there, possible eruption there, possible eruption at long valley, possible eruption at yellowstone. But is it possible to have a massive eruption without the being of a stratovolcano (what not automatically implies it will happen, i’m not a sensatic journalist), but i’m just wondering. Was there a stratovolcano when campi flegrei erupted 35000+- years ago?

    • campi flegrei? Dunno, I wasn’t around. This period saw a huge uptick in large caldera events and bonafide stratovolcanoes going poof. This was the time-frame of the Laschamp event and lots of stuff was going on everywhere at the time.

      My pet theory, (with no proof what so ever) is that it was caused by a subducted plate detachment event. (a part of the Tethys Ocean crust detaching) About the only thing that halfway supports my idea is that over behind Montenegro, is an area in the mountains that shows evidence of hypersubduction. (a region where there is evidence of stacked ancient subduction zones that were running counter to each other)

      It’s likely that when the Farallon plate detached 20+myr ago, that it triggered the Columbia Flood Basalt formation.

    • Lakagigar, good questions.

      There are many instances of massive eruptions without stratovolcanos. More often than not however, volcanic systems start by forming stratovolcanoes first before they blow apart the edifice, leaving a caldera (typically smaller than what you see at Yellowstone). This often results in a volcano without a huge edifice that can still create large eruptions.

      Similarly, some volcanoes that form in rift valleys where the crust is spreading (such as the African Rift Valley, or Taupo Volcanic Zone) never form large stratovolcano edifices, but still can create some of the largest eruptions in the world.

      There is a good example close to where we’re looking for this missing 1809 eruption in the VEI-6 Huaynaputina eruption. This eruption more or less came from a vent that was mostly sitting at ground level (albeit, in the high mountain region of the altiplano).

      As for Campi Flegrei, chances are, there at one point was a stravolcano similar to Vesuvius there, but that obviously was torn apart in the many large eruptions that have occurred.

      Second, I wouldn’t worry about what “they” are saying in regards to Yellowstone or Long Valley. Unless you hear it from a volcanologist who works for the observatories / USGS close to those volcanic systems, chances are they’re just trying to drum up fear (for ratings/clicks), or play into some doomer fantasy they may have.

  7. Are locations like the Upper and Lower Rhine Grabens and the volcanic Eifel plateau in Germany suitable for the exploitation of geothermal energy?

    • I do not know about the Rhine Grabens, but I know that there has been discussions about the area around Laacher Zee.

      • I would expect that in the Upper and Lower Rhine grabens and in the Lower Rhenish Basin the heatflow would be considerable higher, the lithosphere is thinner than in the rest of Western-Europe.

  8. I hadn’t realised that the earthquakes in the Sea of Alboran were centred on a volcanic island

    • Yes, and following some deep EQs (20 and 26 kms) a few days earlier. Coincidence? Interesting location, too, in the middle of the caldera.

      • I would though like to point out that the deep earthquakes and the M3 is unrelated and due to different processes. The deep ones out at Kistufell are though interesting.

        • OK, blocking for one, magma injection for the other being what you have in mind, I presume.

          • Thanks. What I had in mind was that as the magma moved upwards it could have caused the mag 3 in the caldera – though I can now see that that contains its own muddle-headedness. And I have tried, but so far failed, to teach myself to read seismic signatures…

          • One thing we do know is that the area will continue to rift.

            There is a fresh new “crack” running 45km out of BB and if it continues to open due to rifting, even if is a small amount, would this continue to drain material out of the caldera? This could explain the pattern of deep and then shallow, the same pattern that happened as the eruption was ongoing.

          • Since it is likely that this spread has caused a previous eruption at the same place it is probable that it would do so again.
            I would say that the deep earthquakes are related to fresh magma ascending from the mantleplume into the fissure swarms. But, at the same time it is also probable that at some places we have earthquakes caused by shrinkage as the magma cools down.

            It is a very interesting phenomenon.

            I am currently more concerned about what the sudden release of strain in the area has as an effect on the nearby parts. In the last 50 years strain has been released at Holuhraun and Krafla. The release at Holuhraun occured mainly via Grimsvötn Fissure Swarm. This leaves two northern areas that has not had any release of strain for a very long time and those are Theistareykjarbunga and Fremrinamur. Sooner or later that strain will have to be released somehow.
            Also, we have an area in the south running from Thordharhyrna to Katla and back to Bárdarbunga. This zone hase been called the dead zone for a very good reason since it is pretty much aseismically dead even though Icelands 3 largest fissure swarms run through the area.

          • To reuse your analogy.
            The interesting thing here would be what speed the tyre pump is reflating with.
            If we assume that it is the same as for Grimsvötn we would get somewhere between 0,05 and 0,1 cubic kilometers of “pumping” per year.

          • Well there is not really a rate of recharge as such,more a constant state,with influxes, which mostly go towards building crust and more evolved magma’s and occasionally breach the surface as eruptions.If it were as simple as filling a chamber,then when it overfills an eruption occurs, then eruptions would be more consistent in occurrence,but it seems more like a closed system failure causes an eruption?ls Iceland more active now than in the past,or is it just the large system of Bardarbunga that has accumulated stress to failure point?

          • Thing is that those intrusions tend to build up to what is an averageable intrusion rate. Obviously it is both cyclical and chaotic, but over time it is fairly stable.

            You need to separate between crustal intrusions and intrusions into a magma reservoir. One does not build up towards a future eruption, the other does. The rate of evolvement of magmas is so slow that for the frequent erupters like Grimsvötn very little actually becomes evolved magmas.
            Yes it is all about reaching a critical level of stress. Without that you do not get an eruption. But, there is a wide variety of factors driving when that critical point will be reached giving variations in frequency and size.

            Iceland’s volcanism is cyclical and we are entering a heightened cycle, this has been known for quite some time now, during such a cycle more magma comes up from depth. There is yet another cycle that relates to rifting. And currently those two cycles is coming at the same time and they will interact. Furthermore, this cycle peak will be affected by the increased isostatic rebound from the rapidly melting glacier.

            Due to these cycles Bárdarbunga was forced to erupt. The magma that arrived from depth at Kistufell was force injected at the northern side of the magma reservoir of Bárdarbunga via the fissure swarm. This in turn caused the magma reservoir to reach critical point and a NNE trending fissure opened. The rifting from Kistufell to Bárdarbunga and the weakness of the NNE wall, the subsequent rifting to NNE into the Grimsvötn fissure swarm and then up to Holuhraun was of course caused by the strain prior to rifting. So both cycles struck at the same time.

          • Possibly underlying all that is a planet wide cycle,where increased heat inside the earth as a whole has to be relieved via a cycle of increased volcanic or tectonic activity,although not necessarily at the same time,hence patterns of this nature seem to increase worldwide at similar times?

          • I guess it would be possible in some way. But, there is no evidence for that and the long trend is towards cooling. Regardless of that it would be so long cycled that we do not really need to bother with speculating upon it.

    • There was also an odd spike on the tremor chart around that time (i think).

  9. Unrelated to this post, but we need to stop referring to Campi Flegrei as a supervolcano. And by “we”, I mostly mean the news media. I saw a decent article in the BBC posted in the VC facebook group that begrudgingly referred to Campi as “Europe’s Supervolcano”.

    Is it a larger than normal volcano? Yes. Is it exceptionally dangerous due to the location proximity to Naples? Yes. But if Campi Flegrei qualifies as a supervolcano, that means we probably have over 100 supervolcanoes around the world. Campi flegrei is probably about average to slightly above average in terms of size of a volcano caldera system around the world.

    • Well, it just goes to show what the term was intended for. To scare up eyeballs to sell to the advertisers. It’s a horrible term and about as useful as “jumbo shrimp” and even less accurate. It’s nothing but an advertising creation. A “Super Mediocre” term.

      It’s also why I refuse to use the term in the manner they did. “Large Caldera Event” is much more accurate and covers things that did not necessarily originate as a stratovolcano. By my logic, “large” denotes events larger than the median, so even though I don’t have an actual numerical value, I do have a definition of what constitutes “large.” As for the caldera portion, well, if it’s generated as part of a collapse process, that fits. Unfortunately, that could encompass sinkholes. But, sinkholes don’t erupt, so that leaves them out.

      • It is though handy for the handfull of volcanoes that are active and can produce a VEI-8 eruption. All of those are so darnedly different compared to regular volcanoes that they are quite simply in a class of their own.
        But they are few, very few.

        Come think about it I can only come up with Amatitlán (Maria Tecun Eruption), Atitlán (has the capacity, but has not gone off yet), Taupo, Toba and Tondano. Prospect member is perhaps Uturuncu.
        All of those are so different from regular volcanism that they kind of deserve the epitet. Same goes for the future Sakakawea eruption in about 2 to 3 million years time (remember that name if you are still around).

        Using it on other smaller volcanoes is comparatively silly, I agree on that.

        • I also have a strong personal suspicion that the central kamchatka depression could be a future supereruptor if we’re talking about prospects. It has many characteristics you would look for when trying to identify an area where you may find a future VEI-8 eruption. Those being:

          -Ridiculously prolific heat source / energy input
          -Large tall edifice (leads to structural instability and increased likelihood of caldera event)
          -Located within a spreading zone over a subducting slab
          -Possible slab gap and rollback
          -Evidence of bimodal volcanism
          -Thinning continental crust

          Klyuchevskoy alone has pushed out more than 145 cubic kilometers of magma in the last 6000 years (that’s a conservative estimate too). Extrapolating that out extremely conservatively, you end up with over 240+ cubic km of magma from Klyuchevksoi extruded over 10,000 years – a ridiculous pace that may by itself rival Grimsvotn.

          The crazy thing about Klyuchevksoy is that it’s in a very dense cluster with other volcanoes that share very similar characteristics. It is quite likely that they’re all linked to a single deeper magma source. If you account for the volcanoes of the central Kamchatka depression as one singular entity, you end up with the single most prolific volcano on earth almost without doubt.

          Given, none of this will ever happen in any of our lifetimes, but its interesting to think about.

          • I agree, it could also be a spot of potential VEI-8 eruptions in the deep future.

          • Supervolcanoes are different in one respect: their magma chamber is large enough that it can break through 10 kilometer of rock without requiring a structural fault first. They can erupt just by waiting: no new magma influx is needed to initiate an eruption. That makes them harder to predict.

          • Albert, do you know any supervolcano that is not situated on a fault?
            I am hard pressed to come up with one.

          • How close to a fault? Volcanoes are normally associated with faults but are some distance away from them (100 km or more). But Toba is right next to a fault. I think in that case the region is undergoing some extension due to the fault bending, and this has allowed such a large magma chamber to build up. It has to be deep, since otherwise it will erupt before it gets huge. So, perhaps you should be looking for regions where locally some extension is taking place. In the case of a hot spot (Toba isn’t), you are looking for a thick, solid crust, thick enough to resist a lot of pressure, but not so thick it can’t get through at all. Yellowstone is situated in such a place, between areas where the crust is too thick. It has now moved to a thicker crustal region and I don’t think it will super-erupt again until it has gone past that.

          • I think you misunderstand me here Albert. What I meant is that all of them exist on pre-existing local faultlines.
            Take Amatitlán for instance. You have no less than 3 local faults, one of them transverse, and two longitudinal along the graben formed by a spread center.

          • There are also crustal weaknesses at Taupo, Tondano, Atitlán and Toba. I am not sure about Uturuncu, but I would be surprised if there is non.
            If we look at the defunct Yellowstone it sits at a local fracture too that caused the Hebgen Earthquake.

          • Remember that not even the largest volcanic feature on the planet, the African superplume could punch through a craton. Instead it followed pre-existing weaknesses. As it hit the Tanzanian Craton it ran into problems and the eruptions started to follow the boundaries of the craton.
            Come think about it, the definition of living in interesting times will be when it pops out on the other side of that craton.

          • How about thinking of these systems as like core plugs on an engine,when the block freezes the core plug pops out.So Yellowstone is the core plug of the North American landmass,maybe not the only one but the most recently active.This plug only fails when mantle stress on the crust cannot be released any other way,hence the very long eruption interval of an average 600000 years.OK must take my med now😊

          • Ilopango is far from a supervolcano. It has a caldera forming eruption reaching up to a medium VEI-6. Or other terms about 40 times smaller than a VEI-8 eruption.
            The ones on the list I gave have all had VEI-8 eruptions except Uturuncu (that has the magma influx suggesting a future capacity), Amatitlan (that has had VEI-7 eruptions and is having the right magma and high influx) and of course Sakakawea that has had a series of previous VEI-8 calderas prior to the upcoming one in 2 to 3 million years.
            True supervolcanoes are very very rare indeed.

      • Which is the beauty of the LCE nomenclature, it points more to the style of eruption and doesn’t hype the event.

    • If you are unfortunate enough to live to witness one of these large Calderai events,you will then realize that “super eruption”is quite a modest term.In fact even Krakatoa was a super eruption as there has been nothing remotely like it in an explosive sense since and that is modest compared to some of the big Calder events and miniscule compared to the likes of Yellowstone.So yes in the lifespan of a human being these events are really are extraordinary events and the term large does not do it justice.

      • My guess if you wanted to have the thankfully miniscule chance of a ring side seat for one of these events in your lifetime,then move to New Zealand.The mere mention of a Taupo eruption there is enough to instill a deep dread and Taupo is not the only candidate.Why because the evidence of previous events is clearly seen ,in road cuttings and large boulders a great distance from ground zero and the fact it happened as recently geologically as 1800 years ago and a “smaller”eruption in the Okataina Caldera as little as 120 years ago.

      • Witnessing such an eruption, should it occur, and living to tell the tale, is probably impossible unless you are on the ISS. Close enough to see what’s happening is also close enough to be fried.

  10. drilling in a volcano woke up a certain Apo Mallari,
    Don’t disturb the dragon!

    • Rubish!

      Apo Mallari is the local “God” of Merapi. It famously had a High Priest that denied when the worlds leading volcanologist said it was time to move away because there would be an eruption. The high priest in a snit galumphed off up unto the mountain with a few faithful.
      The high priest and his faithfuls are no more.
      The rest of the locals decided that the volcanologist, Surono, would make a better high priest of Apo Mallari.

      As soon as someone needs to invoke a god around volcanoes it rarely bodes well for scientific facts. Geothermal energy harvest did not in any way cause an eruption at Merapi.

      • In Indonesia, the only Higher Authority it’s worth to invest one’s trust in is Surono and his disciples. In Iceland, the IMO are a better bet for accurate information and forecasts than the three Norns; Urd, Skald & Verdandi. In Hawaii though, the locals with their offerings to Pélée seem to be doing just as well as the HVO, so one should never say never. 😉

      • I got the name a little wrong I meant Apo Namalyari, and no it is not Merapi.

  11. One comment on the post: the idea that the ice cap cool Bardarbunga is interesting but I am not sure it holds water. Ice can insulate. It slows the flow of heat out. But the heat flow should balance what comes out of the ground so that doesn’t work. The ground heats up until the heat flow through the glacier equal that from the ground. In the end, the ground heats up, not cool down. Melting the ice does cool the mountain down – for a while. A more important effect of the glacier comes from its weight. It adds to the pressure and increases the pressure gradient along the slopes of the mountain. This makes it more likely that the mama will go for a side exit as in fact happened in Bardarbunga. The dyke at every point chose the path of steepest pressure gradient.

    • Sweet Albert. You are of course probably right on this one.

      I should have thought of it. I once spent a weekend in a snow bivack in minus 48 celcius, so I kind of have the insulating properties of snow ingrained in me since it saved my life.

      • After further pondering.
        The initial volcano that the intrusion came up via was Kistufell. It lack a glacier all together. But it has not erupted for a long time.
        I think we need to take count of the fissure swarm itself. Both Kistufell and Bárdarbunga is being torn apart in an EW direction by the plate motion creating the fissure swarm. Due to some poorly known mechanical motion both Grimsvötn and Bárdarbunga is prone to create radial fissures towards each others fissure swarms. Probably due to the transverse fault running in from Hofsjökull to the area.

        If I am correct it would create a system that goes between rifting if the pent up tectonic movement is great, and caldera eruptions if it is already released. It is probably far from all of the factors, but I think it is one of the factors that creates those pesky rifting cycles.
        And following that it is probably more likely that the next Bárdarbunga eruption would be intra-caldera, at least if it comes fairly soon (within the next 50 years or so).

        • Are you sure the initial intrusion came up via Kistufell? If you watch an animation from the event, it looks like the swarm started at the SE corner of the Bárdarbunga caldera and propagated from Bárdarbunga towards Kistufell, pretty much following the gradient of the mountain in a similar way as the dyke that led to the eruption did. The quakes in the Kistufell area also were quite shallow.

          This interactive animation show it in a very nice way:

          Rotate the view so you are looking at Bárdarbunga from Askja, play the animation, and it looks almost as if the eartquakes come trickling down the mountain.

          If you look over a longer time perspective, the part of Bárdarbunga that sees the most deep quakes is the SE corner where the swarm started. I think the original intrusion came from that point, continued as an intrusion towards Kistufell, stopped at Kistufell and the edge of the glacier. The pressure was not enough to cause an eruption at that point, instead, the dyke that propagated towards Holuhraun became the weakest point and we all know what happend next.

          • That starts to late and has a cut off point at 10km.
            The swarm started at 20km depth under Kistufell weeks prior to that animations starting point.
            So, yes I am 100 percent sure. Rememberthat I tracked the activity months prior to the onset of eruption.

            To expound further. About 3 months prior to the eruption a series of anomalous earthquakes took place at 3 different locations at more than 20km depth. As these continued we got a continously clearer image about that fresh magma was moving up from depth. During the last 4 weeks before the larger seismic crisis started we had a lot of backchannel discussions about it, but we decided to wait for an official response from IMO that never came.
            While we waited for IMO to act I wrote a few articles about the volcanoes involved treating them as separate events fully well knowing they where not. In the end I felt that I had to write a “watered down” article warning about a potential eruption since IMO was not on the ball. The day after the seismic crisis started.
            One of these days I will write an article about how I tracked it and how we reasoned in months and weeks following up to the eruptions. It was almost as exciting as a mystery novel in a way.

          • If you are interested in looking back at the events Carl is referring to I would suggest have a look back through time leading up to the August 2014 event.

            If you go to the link starting in Week 33 2010 and page through time using “naesta vika”, you can really see the activity is centered around Kistufell and seems to be activating the East side of BB. For a better look at weeks that look interesting or very active, click on the Bardarbungu link below the diagram and see a better picture. This area is very active for quite some time and then after the eruption really goes very quiet


          • How many times has that happened and no eruption has occurred?Predictions never can wait for evidence,by the the time there is evidence the event has occurred and prediction without evidence is as useless a whistling in the wind.Oh the dilemma😀

          • If you really must know Geyser my score is 7 eruptions to 1 false positive. The false positive being Hekla.

            To make it clear. No tealeaves are involved.
            I read a tremendous amount of articles on the volcanoes at hand, I have the right education, I do understand the squiggles on a seismometer by heart, I also understand what the rest of the equipment is telling me. I then use that to build a mental model of the volcano. And if after all that it checks all the prerequisite boxes I say it will erupt. For some reason not related to tealeaves, yetis, or seeing cryptothingamabobs where there are none; the volcanoes tend to do as I predict.

          • Geyser. Let me state this once.
            Your opinion about yourself is not evident in your comments.

            I will though answer this comment for the benefit of others.
            Evacuation orders are issued by the authorities in the various countries affected. In Iceland it is IMO that issues the warnings and Allmannavarnadelid that issues the evacuation orders. Similar sets are in place in most volcanic countries.
            Do I work for any of those? No. I have though turned down employment at two such authorities since I have another daytime job.

            Now back to you Geyser. When you stop being an internet troll and take the time to spend ten years studying physics, geophysics and volcanism then you will start to understand things better. Until then I actually suggest that you listen and learn instead of constantly trolling.

          • @Carl

            It is amazing that as soon as the eruption started the area circled in green, which was very active for years, went completely silent to this day, nothing happening at all!

            Too bad all things are not as silent since the eruption

          • Cool image Ian.
            The lineament from Grimsvötn is really standing out in it.

          • Yes, I noticed that after your see that line become active, it seems to energize Grims and even west for Grims.

            Another place you can see how Kist is very quiet until 2005 and then becomes very active is here

          • Really good video.
            One thing I found interesting beside the Kistufell thing.
            You could clearly see the fissure effect at Gjálp in 1996, again in 1998 as there was a swarm running between Grimsvötn and Hamarinn, but no such things in 2004 or 2011.

        • Presumably the connection between the deep magma reservoir of Kistufell and the shallow is blocked? As it hasn’t erupted for a long time. That would make it plausible that it feeds towards Bardarbunga. At shallower levels, Bardabunga can feed into Kistufell so if it every erupts is may be magma that traveled the long way around? Underground this place must be a maze. All magma routes go through Rome Bardarbunga.

          • The fun thing is that the fissure swarm is the deep magma reservoir. So in a sense all of it is Bárdarbunga.

          • The place I’m wondering about is the frequently active swarm at 64.6N 17.2W (circled in the figure below). That’s just a little south of where the dyke started and, like Kistufell, it had some deep quakes just before things got interesting.

            Blue dots are all registered quakes from the beginning of 2014 up until today (thanks IMO). Red circles are all quakes deeper than 20km. We can see that the area circled in magenta hosts quite a few deep ones.

            Plotting time vs depth shows that the area had a few deep quakes just before the start of events. Most notably, just before the uptick of activity in May, there were two quakes in that area; one at 15km and one at 20km. In June there was a deep swarm, and another one in July. During the eruption, the area was quiet, but activity started again right at the end of the eruption.

            Blue dots are all registered quakes in the Bárdarbunga area (same as top figure), green is Kistufell and red is the area I’m wondering about.

            So, what is that place and what is happening there?

            Friendly disclaimer: I’m not trying to prove anything and I’m aware that there are more things to look at than just hypocenters. This is not my field of expertise and I simply would like to know more.

            Also another note on the Kistufell activity. Ian pointed out that activity started there in 2005.I wonder if that is not just an effect of the SIL network being upgraded with more equipment in the Vatnajökull area?

          • Not gonna get into the Crapple fight, but I would like to point out that there was a recent security hole “backdoor” found in KitKaT 4.4. It seems that it was put there to troubleshoot connectivity issues with some countries cell network and they never closed the hole. I recommend upgrading past that version O/S if you can.

            “Budget smartphones from Lenovo, Huawei, and other largely Chinese brands contain an accidental backdoor that grants intruders root access.

            Android KitKit is still the most popular version despite being first released in 2013 and superseded by the much-improved 5.0 Lollipop and 6.0 Marshmallow releases.”

          • That is one of the 4 intrusions I followed prior to the eruption. It meandered up to around 5km before the eruption. I think it is what created that odd “knee” where the dyke turned around it could also have partially fed the eruption.

          • “We hypothesise that the low-frequency earthquakes
            at Torfajo¨ kull are expressions of a shallow cryptodome
            in some sort of state of activity. Our interpretation is
            supported by the following facts: 1) seismicity is very
            localised, 2) the low-frequency activity coincides with
            the locations of highest temperature geothermal areas
            within the volcano, 3) Torfajo¨ kull is a rhyolitic volcano
            dominated by dome-building activity.”Hmm interesting?

          • I just think it is so notable how the quakes have just stopped in that location since the eruption started, this has some significance to being somehow connected I would think

          • Strange thing with the images. I thought all browsers supported png. I don’t like using jpeg for graphs, since it is a lossy format that blurs the details. I can try using gif next time I post any plots.

          • @tomas

            These plots are fantastic!

            There are a few spots that are interesting including the one you pointed out. It looks to me like the dike has deep connections in a few spots? appear to be one where you point out with the red circle, another just a bit further down at 66.66 on the video I posted, again at 64.74 and 64.80 on the video where the quakes stall.

            regarding the lack of quakes pre updated SIL, if you look at the video right at the beginning in 1996 there is a swarm in Kistufel just before the eruption and then it goes quiet, then if it is not my imagination, every time it swarms there the activity east and south of BB down towards Grims seems to get going.

          • The dike is finished, just a cool plug of magma,the question is what is going on that led to that event and is still occurring at some level now?The overwhelming amount of energy released had little to do with that dike.

      • Seems like that old lug of magma that arrived a few years ago decided that it is time for a walkabout.
        This might become interesting.

      • First one is clearly a clean tectonic break. Will be fun to see if this evolves into other types of quakes.
        So far all the aftershocks seem to be tectonic, and among the M2+ earthquakes there is not clear trend of movement up or down.
        But the site is still damned interesting.
        Could be that the lake just drains again.

  12. Leave it to technology to mess with your head.

    For the last couple of month, my outdoor temp sensor has been unresponsive. I haven’t really wanted to look into it because that would involve digging through pages of convoluted instructions about how to set the base station up. Its been easier to just open the door and look at the weather.

    Today. Its a bit chilly for florida and while padding around the living room nursing a cup of coffee, I notice that the outdoor sensor has come back online.

    • Ours did that when moisture got to the battery. ‘Watertight’ just meant that water can’t get out. Apparently it can still get in.

      • Hah – my external wind and temperature unit stopped functioning. The so-called watertight unit was so full of water that when I opened it a stream of mouldy green water poured out of it! You are so right.

      • “just meant that water can’t get out.”

        I like that. Very pragmatic. 😀

        After a yard period, I found that condensation from the cooling unit in upper ECM would not drain out of the condensation pit. Looking into it, we found that a DC plug had been left in the pipe where it discharges over the side. We sent a guy from 1st division over the side with a harness and a mallet to remove the plug. One tap and the plug came out, and he caught 6 decks worth of water from that 3 inch diameter line square in the face and chest. Had it not been for the safety harness, he would have fallen. As it was, he just dangled there spitting and cursing.

        Most of that foul water had been sitting in that pipe since we had charged the cooling unit when we came out of the yards.

    • Speaking of enclosures.

      One of the biometric units that I occasionally work on had the manufacturers designed (and expensive) environmental enclosure. Every other site had some kludged together plywood gizmo that was made on site. Anyway, the manufactures enclosure did not survive being hit with a pressure washer. About a week later, they had taken it down and built one of those PoS wooden gizmos as well. Very tight, almost no room to work. It is no more protective than the OEM enclosure, but it’s not as expensive.

  13. We are waiting for the confirmation from INGV and Boris Behncke, but there seems to be a bit of a chance that it is time to pull out the wine and cheese.

    P.S. Boris recommends wine, but says there is no eruption as of now. We will have to sit and wait for a few hours before we know. D.S.

    • Hopefully the weather will clear before Etna puts on her show, right now there’s barely any visibility on the cams. Buy some extra wine, Carl, she may tease us for a few days before she decides it’s showtime.

  14. Somewhat interesting – Ticsani volcano in Peru has been rumbling quite a bit apparently. I find this interesting because I thought for a short bit that it could have been a candidate for the missing 1809 eruption (read at the comments in the last post). Turns out, it’s not likely to be a candidate for this eruption, but it is quite an interesting volcano on its own.

    Geyser, since you seem fixated on lava domes, this should be a volcano you would be interested in 🙂 .

  15. Looks like the Gremlins were in at the cables again.

    We can only apologies for the outage, this was nothing to do with us, it was a WordPress issue apparently.

  16. Been sitting watching the FHP traffic site this evening. A fatality was listed at mm 42, westbound on the interstate. The listed hazard was that all eastbound lanes were blocked. Now, it was either a bona-fide mess, or else the rubberneckers were out in force. They just now got the interstate cleared for traffic….4 hrs later. Usually, if you have to bring in an air ambulance you just cordone off a clear sport of roadway and set up up your LZ there.

  17. In response to Dawmast, above: a map of surface heat flow is below, from Davies et al. 2003. It shows that Iceland is about the best place on land in the world. In Europe, southern France has possibilities. Elsewhere, the usual suspects: New Zealand, Indonesia, .. By far the best way to use the heat flow would be to but a ground heat pump down in the mid ocean spreading ridges. But the human demand for heat there is pretty low


    — Reuters

Comments are closed.