Schneier on Security
A blog covering security and security technology.
« Surviving a Suicide Bombing |
| Security Fears Drive Iran to Linux »
March 26, 2009
A Solar Plasma Movie-Plot Threat
This is impressive:
It is midnight on 22 September 2012 and the skies above Manhattan are filled with a flickering curtain of colourful light. Few New Yorkers have seen the aurora this far south but their fascination is short-lived. Within a few seconds, electric bulbs dim and flicker, then become unusually bright for a fleeting moment. Then all the lights in the state go out. Within 90 seconds, the entire eastern half of the US is without power.
A year later and millions of Americans are dead and the nation's infrastructure lies in tatters. The World Bank declares America a developing nation. Europe, Scandinavia, China and Japan are also struggling to recover from the same fateful event—a violent storm, 150 million kilometres away on the surface of the sun.
It is hard to conceive of the sun wiping out a large amount of our hard-earned progress. Nevertheless, it is possible. The surface of the sun is a roiling mass of plasma—charged high-energy particles—some of which escape the surface and travel through space as the solar wind. From time to time, that wind carries a billion-tonne glob of plasma, a fireball known as a coronal mass ejection (see "When hell comes to Earth"). If one should hit the Earth's magnetic shield, the result could be truly devastating.
The incursion of the plasma into our atmosphere causes rapid changes in the configuration of Earth's magnetic field which, in turn, induce currents in the long wires of the power grids. The grids were not built to handle this sort of direct current electricity. The greatest danger is at the step-up and step-down transformers used to convert power from its transport voltage to domestically useful voltage. The increased DC current creates strong magnetic fields that saturate a transformer's magnetic core. The result is runaway current in the transformer's copper wiring, which rapidly heats up and melts. This is exactly what happened in the Canadian province of Quebec in March 1989, and six million people spent 9 hours without electricity. But things could get much, much worse than that.
Posted on March 26, 2009 at 12:44 PM
• 56 Comments
To receive these entries once a month by e-mail, sign up for the Crypto-Gram Newsletter.
Technical error in article: If the magnetic field is moving, then it's AC current, not DC. And unless the field (or conductors) is moving, it doesn't induce current in any conductors. It may be low frequency, but it's still AC.
I guess Edison would be laughing from the grave, if we had only gone for small DC generation facilities all over we'd be at lesss risk for a large scale cascading failure.
I don't believe that there would be a World Bank to declare the US a developing country afterwords.. If we were hit by such a plasma wave.. I think 70% of the Earth's population is toast due to everything from lack of food deliveries and other technological collapses. And while having 'hardened' centers to help rebuild is probably a good idea... its a level of collapse that I don't think one could plan for.
On the other hand, it is also a good solution to global warming.
"Flare", by Zelany, is a decent SF novel that goes into some of the effects of this type of event.
I'm currently reading "Aftermath", set in the near-future after a supernova event causes an EMP to wipe out all the microchips in the world.
"The report outlines the worst case scenario for the US. The "perfect storm" is most likely on a spring or autumn night in a year of heightened solar activity - something like 2012."
Translation: we know this can only happen when things are really really bad, and even then there's only a small chance. 2012 is supposed to be bad, so clearly it will happen then.
Perhaps I have a cold heart, but I have no particular sympathy for those who depend on perishable medications. Those who depend on them are dependent on a massive massive infrastructure to get said medications. If everything goes out... that's that.
They claim not to be fear mongers, but anyone else find the discrepancy that their claims of coal running out in 30 days are directly opposed to the previous statement that 90% of electrical distribution is going to be FUBAR?
This sort of situation is what a president is for. You need one person who can decide the right order to "reboot" the US in. Hopefully, when this flare happens: a) the US is still around b) the US president is ready for it.
On a different note, there's interesting groups online who concentrate on how to be ready for anything... even this.
I was thinking this sounded similar to an EMP attack myself.
If the US is starving, then so is the rest of the world since we are a massive food exporter.
But as far as power plants not coming back online because the grid is out, I suspect in very short order they would isolate large segments and create "mini-grids" where a single large power plant was directly powering its surrounding areas. So the only real unlivable places would be large "hi-rise based" cities like NY, Chicago (which I would posit are already unlivable, but thats IMHO) where they pushed the power plants out into the countryside.
And lack of power to pump gasoline would quickly be overcome by other ways of pumping it; the actual problem would be getting new deliveries from the refineries.
A moving magnetic field induces a direct current in a conductor; an alternating current requires a reversing field.
Ben Bova wrote a short story, later expanded to a novel "Test of Fire" (I don;t remember if the story's title was the same or not), where a solar flare actually hits the earth. It laster 20 hours and sterilized everything except a strip up the west coast of North America.
In support of pointing out the technical errors: The large distribution lines you see criss-crossing the country generally carry electricity DC. The substations and transfer stations that support this distribution system are there to boost (kind of like a repeater) the power to recover from resistance loss and convert the DC to AC and back as needed. The transformers at these locations can handle pretty large spikes and there are overload safety switches that will trip and prevent major damage.
Of course, that was a movie-plot, right? So, there is no accounting for technical detail or accuracy.
Hey, would more polution in the air slow it down? If we polute more maybe we can mitigate this problem. Let let Mr. Gore at it.
Obviously, it's time to declare the War On The Sun(tm)! We've been terrrorised by the silly yellow gas ball for far too long already.
What you describe is just a love-pat. For a vivid description of what a star can do to its planets when it gets upset, see Charles Stross' novel "The Iron Sun". First you feel a momentary nausea and disorientation, and at the same time all your computers and electronics are fried. A few minutes later you notice that the Sun is getting brighter and brighter, and then you notice that everything in direct sunlight is catching fire... From memory, just an ordinary nova would ablate the surface of our planet at 200 metres per second - boiling off the solid rock until there is nothing left but a dissipating cloud of gas.
That's a nova. Now a supernova is *really* violent...
For scads of threats that could kill us all, see the Exit Mundi website.
@ RH who said "Perhaps I have a cold heart, but I have no particular sympathy for those who depend on perishable medications."
I sincerely hope that you only mean this in the scope of an infrastructure destroying event and not as an every day feeling of malice toward those with issues requiring daily medication.
I agree that if something happens that puts our resources in danger and makes it nearly impossible to meet those needs, it should be our priority to focus on those who are able-bodies while helping the other at least die peacefully.
Something like diabetes or schizophrenia may force a person to be dependent on medication to live a normal life.
Cancer or AIDS require massive doses of medication that may have terrible side effects just to keep the person alive.
Other things like migraines or anxiety or depression may be tolerable without medication.
If you view the reliance on medication to treat certain conditions as a weakness, then I hope you never develop a chronic condition.
@Muffin: "Obviously, it's time to declare the War On The Sun(tm)! We've been terrrorised by the silly yellow gas ball for far too long already."
You're onto something. We could send a group of combatants to terrorize the sun. May be the way to close Gitmo. :)
On a serious note, the sun is obviously something we cannot do anything about. Just one more case where we need a response to results. In this case, losing electricity on a massive scale, which could be caused by solar disruptions, EMP (if someone could actually manage to get a nuke up high enough, good luck), earthquakes, hurricane, meteor, tsunami, polarity disruption, nuclear war, etc. Rather than worry about crazy threats, have a response plan for the effects, which is independent of whether the cause is likely, plausible, unimaginable, or just plain a wacky thing to worry about (cuz wacky sometimes happens).
The arguments about why this 'can't' happen misses the rather large data point that it has happened. Repeatedly. We don't have small scale disruptions any more since the equipment has been beefed up based on experience, but a larger one still knocked out power in Quebec.
The Carrington event shows that much larger events are possible, but we have no idea how common they are since we don't have a large baseline of observations. Maybe it was a one-in-ten-thousand years event, maybe it's something that usually happens every 50 years and we've had an unusual lull. We don't know.
I keep coming back to an analogy to hurricanes. We know they happen and we know they can be devastating. We can't live our lives in constant fear of them, but it would also be reckless to pretend that they can't happen.
Historical sidenote: the reason Florida was Spanish instead of French was because the captain of a small French fleet sent to beef up the forces at the French outpost didn't believe the wild tales of massive storms. He went sailing off to engage the Spanish... and right into a hurricane. His fleet was grounded and sailors executed, meanwhile Spanish soldiers got to enjoy a forced march through hurricane conditions to surprise the French outpost as the storm ebbed. He was sure that such massive storms didn't exist and everyone paid for it with their lives.
To add to the list of SF stories in which this has already been done to death: "Inconsistent Moon", by Larry Niven.
In fact, the MMORPG EVE Online just had a major story line involving a CME hitting an inhabited planet two weeks ago.
Seems like a lot of folks have had this in the zeitgeist recently.
On a more amusing note, someone pointed out (correctly) that the long-haul power lines are DC. Saves on losses. Consider if we manage to get microwave power before this happens (think Simcity 2000):
DC from solar cells
DC long-haul power lines
AC short-haul power lines
@Guy: Sorry to say, but pretty much everything you said is wrong. Most large distribution lines are AC, not DC. There are a few large DC lines, which can run at higher voltages and are thus more efficient. (They can go higher because corona discharge limits your peak voltage, and DC runs always at peak voltage while the average (RMS) AC voltage is sqrt(2) smaller than the peak. Also the skin depth at DC is infinite, so you can make 2" diameter wires and actually use all of the cross-sectional area.)
Also, substations can't boost the power, they just convert the voltage.
Finally, tripping the overload protection is what generally leads to blackouts, so that doesn't help.
The article specifies or longest highest-voltage lines as being candidates for the doomsday effects. These longest highest-voltage lines are also the most likely to be DC.
You're right about the substations not "boosting" power, but tripping overload protection is a good thing. The US can handle a 1week blackout while we replace every single overload protection device. The doomsday situation is where these mechanisms fails and we have 2month+ blackouts
RH wrote: "Consider if we manage to get microwave power before this happens (think Simcity 2000)."
No, no, no, no, no! Microwave power is not cost effective! It's best to go straight from Oil to Fusion and then start building archologies.
Solar storms do not last for a long time. Electronic nations would recover. Not all modern technologies require a continuous supply of electricity. Refrigeration, at the least, can be done easily using clay pots. ;)
The debate about DC -v- AC longhaul is a mute point. In some parts of the world it's AC and in others it's DC.
On a historical note most longhaul has been AC for a very simple reason. Converting from one voltage to another is simply and very efficiently done by a large lump of iron with a couple of coils around it known as a transformer. It has a further advantage of isolating one side from the other so segregating the network from natural effects like static discharge, the effects of thunder storms and solar flares etc.
DC longhaul systems were first used where power was transfered from one AC network to another incompatable AC network (be it frequency or phase incompatability). The original method of conversion using Mercury Arc rectifiers etc was something truly scary to see.
The other historical use of medium haul DC networks are for railways and telephone systems.
Due to the likley effect of a solar flare, the older AC networks with mechanical safety mechanisms are likley to be more resiliant than the modern DC systems.
One another note I'm surprised not to see someone mention tje "Ring World" novels one of which had an induced solar flare used to restabalise the ring around the sun....
Oh and we do have quite frequent reminders of what effects the sun can have on electrical supply systems and electronics in general. As you approach either pole you get rather charged particles coming down close to the earths surface, at times of high solar activity (see sun spot cycle) the particles get sufficient to actually knock out power and communications systems in some places far away from the poles (aura have been reported as far south as Rome).
Some solar flares in the past have actually knocked out communications satelites with little difficulty, however in the vast majority of cases they do come back on line a short while later.
What we need to remember that whilst the energy involved with a solar jet is many many times the anual energy usage on the earth it is however very small beans compared to the sun's normal output.
Oh one thing I forgot to mention.
In a lot of places those over head power cables realy are transmission lines just like twisted pair cables. The effect of the solar flare or "glob" is likley to be "balanced out" over any given "full twist".
For it to have the "devestating effects" described the transmission line would have to also have a "phantom circuit" associated with it (ie the transmission line side of both transformers would have a center tap to ground). A protection network to deal with this eventuality is generaly fitted for other reasons.
I guess the question is would the rise time be sufficiently fast to allow enough energy in the phantom circuit to do significant damage in the length of time it takes to activate?
If it was then we would have a number of other problems to be worrying about.
Remember that lightening has an effect on the earth's atmospher that is detrimental to most forms of life...
Then there is the question of what would be the effect of a "billion tones" of matter (charged or otherwise) hitting the earth (think of one cubic kilometer of water) at that sort of velocity be?
OMG! Why did they pick 2012? Now we'll have more doomsday woo-woo.
Dr. Ian O'Neil over at Astroengine has done a stellar job (sorry bout the pun) of debunking 2012 stuff. See http://www.astroengine.com/?cat=52
Tom Welsh, the book is called _Iron Sunrise_. The hundreds of m/s rate for the ablation of the Earth is assuming that the Sun became a supernova (which it cannot), assuming that in so doing its brightness increased by a similar factor to real supernovas (a paltry matter of ten billion times).
Figures from Steven I. Dutch's excellent paper at http://www.nagt.org/files/nagt/jge/abstracts/... (Charlie Stross linked to this on his blog a few years back, which is probably where you saw it.)
Thankfully this is not a meaningful security threat (I'm not sure what a meaningful response would be, unless vaporizing counts.)
We can demonstrate the effect of a moving magnetic field on a static conductor simply, needing only a small strong rare-earth magnet, a length of copper tubing, and a penny.
Holding the tube upright, drop the penny into it and notice it falls as we would expect. Then drop the magnet in: it will fall very much slower.
The retarding force is back-EMF, opposing the motion of the magnetic field.
No one, no one challenges the great Nostradamus! Hister (aka Hitler) - I mean how much more evidence, how much more proof, how much more of "Are you sure?" do you require of The Man? Visionary, I mean.
As Nix said, the book is _Iron Sunrise_.
In the book, Moscow's sun is induced into going nova by manipulations of cosmic string and/or extra dimensions and/or causality violation.
When it comes to SF-plot threats, that one's way out there.
You can still burn out a transformer with low-frequency AC. The lower the frequency, the easier it is to saturate the core, and the more iron the transformer needs to withstand it.
This smells to me more like a black swan than a movie plot.
I don't know much about power grids and their design, but it seems like designers made assumptions about conditional independence of failures, such that no single event is presumed to be large enough to cause failures that can not be handled through redundancy. The fact that such an event is possible, and has been observed in the past, makes it quite reasonable to explore its impact.
(I would expect that grid designers do consider large-scale events, but most such events probably fall in the realm of "we've got other problems to deal with", e.g. nuclear war. A solar event is alarming because, a) we have no way to predict or prevent it, and b) there are no other long-term side effects to distract us.)
Check out Phil Plait's excellent book "Death from the skies" for detailed and rational discussion of this sort of astronomical (in the literal sense!) risks.
Also the "bad astronomy" blog at http://blogs.discovermagazine.com/badastronomy/
Let's not forget the threat of alien invasion, Communist world domination or the appearance of flying pink elephants either.
On the one hand, some of the effects they describe are out of all proportion. A massive solar storm or flare coming our way would cause a lot of problems, yes, but it would not turn the US into a developing nation.
On the other hand, I don't think it's a question of *whether* we'll ever ever get hit by a massive record-breaking solar event, but *when*. We've only got a couple hundred years of decent recorded data on solar activity, and if one thing is obvious from that data we do have, it's that the sun doesn't run through its entire repertoire of behavior in a timeframe that short.
Turns out the answer to your phantom circuit. question tends to be yes from some studies I have seen. Though the specific threat was EMP from nukes. Solar flares was discussed however just like weather prediction is discussed re military planing.
@everyone or noone.
However even a really really big flare will not cause irreparable damage. Most places will have power back on with 24 hours. Lets face it short term electricity outage is not a big deal. And if it is you would have backups (which are unlikely to be affected).
The rest is pure bunk up there with climate change will kill us all. Or that magnetic pole reversal will burn us (aka the core). And what is up with 2012? I guess a lot of people were disappointed with Y2K not taking us out and need a new date. They should pick a prime number! 2017!
There are quite a few large level risks from space. However by far the most likely is a old fashion asteroid strike. Also pretty close to that is "super volcanoes".
But I don't usually sweat the small stuff.
If power companies are aware of the threat and it takes weeks to replace the damaged equipment, then why don't they put more focus on building more backup equipment and training installers? It occurs to me that building transformers and other infrastructure equipment now while we have the power is vastly more prudent than trying to deal with it during a blackout.
Movie plot threat indeed! Very timely, given that "Knowing" just debuted with Nicolas Cage and is in theaters right now.
Mind you, I'm not recommending it, but it wouldn't be the first time I saw a "wow, what a horrible threat!" article come out suspiciously close to a movie, lame or otherwise.
There's a lot of obstacles to preparing for an unknown disaster of unknown likelihood. First, it isn't immediately profitable. Second, power utilities are normally highly regulated, and it would be necessary to convince the regulators to allow the power company to charge more to increase preparedness.
We can compare that with the idea of building extra capacity, which will be worthwhile in a reasonably foreseen timeframe. Many utilities simply weren't interested. The main power company around here was proposing a slight rate increase to cover capacity increase, and had problems with the state regulatory office (I don't remember how that turned out).
I think this initiative would have to be at a government level, since spending money on such preparedness tends to be a government function (there not being a whole lot of profit potential in it). It would involve increasing the rates, and would therefore face a lot of resistance, particularly in a bad economy.
Have had a look at the final scene in the movie PREDICTIONS ?
High-rise cities like new york a) tend to have a surprising amount of generating capacity onsite or in very close reach and b) are designed for significantly less energy demand than sprawl cities, especially ones in the sunbelt. New york, of course, has already demonstrated repeatedly that 24-hour outages are not a big deal.
Re: HVDC networks -- there are only a few HVDC transmission systems in North America; the vast majority of high-voltage lines you see are AC (if there are three or multiples of three conductors, they are AC).
According to http://climatetechnology.gov/library/2003/... "There is no active U.S. program for HVDC development..."
DC current in an AC transformer will burn it out very quickly because the windings ordinarily generate back-emf which increases the apparent resistance at AC frequencies beyond what the pure DC resistance would be. Very low frequency AC would be equivalent to DC power for the purpose of this discussion, and that's what you would be getting in this case. (The Swiss trains would probably survive just fine since their grid is 16.6Hz!)
The existing protection systems would have a hard time breaking low-frequency AC or DC current, because they depend on the self-quenching property of alternating current to break the arc as the breaker or switch is opened. So existing protection mechanisms would probably not be adequate.
This is definitely a serious concern, and fits the profile of a Black Swan perfectly.
Continent-wide power lines are perfect antennas for picking up very low frequency high energy signals from aliens in outer space...
Although a real threat, this article does really take it too far. People will starve because vehicles wont have fuel to transport food since the pumps at gas stations wont work? I think if it got that bad they could get out good old hand-pumps.
"I think if it got that bad they could get out good old hand-pumps."
When I was young the way to get gas (petrol) out of a tank was a six foot length of hose pipe and the ability to suck hard ;)
Unfortunatly some people got so good at it that somebody went and designed a simple preventantive measure to stop them so it does not work any more...
"Continent-wide power lines are perfect antennas for picking up very low frequency high energy signals..."
Actually no the are not that good at ELF and below, they are much better at HF and above.
As I noted earlier they are often rotated round from pylon to pylon making it into a twisted transmission line.
For the transformer to be burnt out significant current actually has to flow in the windings.
And this is the problem in a balanced transmission line any external field inducess the same current in all the conductors and it flows in the same direction. Which means that effectivly the conductors are in parellel and shorted together at both ends. If there is no external connection to form a circuit back down the length of the transmission line then NO current will flow irespective of it's frequency so there is nothing to burn out the transformer.
It is only when you center tap the transformer to ground that you have a return circuit via earth. This is the "phantom" circuit I was refering to in my earlier post.
If there is no earth tap the only way current can flow is if the effective Potential Difference (PD) on the transmission line to ground is sufficient to arc across to ground.
Due to normal atmospheric effects this is a problem that is generaly already catered for.
Therefore without a "phantom" return for damage to be done by this senario the "solar flux" would have to have a very fast rise time to do damage.
And as greg has noted there has been "EMP" research that has come up with methods of dealing with fast rise times.
Which brings us around to the cost... If the event is rare (ie MTTF is large) and bringing the service back up is measured in hours (ie MTTR is comparativly small) the availability figure is going to be up in the five 9's or better. There are a large number of other events where the availability figure is going to be down in the four 9's. So guess where the money is going to be spent...
Stuart Clark's excellent book "The Sun Kings" is a fascinating historical account of how 19th-century scientists came to understand how the Sun causes aurorae and terrestrial magnetic storms.
It includes a vivid description of the disruption which major solar storms wreaked on the telegraph system, and the very real hazard of electrocution which telegraph operators faced if they tried to use their equipment during such storms.
Clark's book may not be disaster-movie material, but it's a fascinating story, and science is sound.
@clive: "they are often rotated round from pylon to pylon making it into a twisted transmission line"
In 50 years of infrastructure-watching (mostly in the U.S.) I have yet to see an example of this. For the >100kV latticework pylon lines that abound in the States (one passes about 1 km from my house), this would be quite a trick to do, both in terms of mechanical loads, and in maintaining the required spacing between conductors.
I would imagine that a high-tension pylon which makes such a swap would look obviously different from its neighbors, and that if the swap is made between two pylons, the pylons at both ends and the intervening span would be odd-looking.
Won't you please post a link that shows the design (or better yet, a photo) of such a switcheroo arrangement?
I thought most high-voltage lines were 3-phase. How does that affect the "balanced transmission line" theory?
"In 50 years of infrastructure-watching (mostly in the U.S.) I have yet to see an example of this."
I think I've seen these, but it never occured to me they were making "twists" in the conductors.
IIRC, the pylon has the insulators, hence the lines, arrange vertically, where the typical pylon has them arrange horizontally (roughly speaking). To produce a twist, or half-twist, consider the horizontal lines at one pylon as ABC left to right. At the vertical pylon, they are arranged ABC from top to bottom. This gives a quarter-twist clockwise. At the next pylon they are arranced CBA left to right, and have now undergone a half-twist.
I always thought the vertical pylons were intended for sharper dog-leg changes of direction, such as following a road or railroad. You can't have horizontal conductors make a sharp direction change without reducing the distance between the conductors as they approach the pylon. But if the pylon has the conductors arranged vertically, you can make as sharp a direction change as you want without affecting conductor distance.
For those watching at home, if this doesn't make sense, it's a soda-straw and string kitchen-table experiment.
New Scientist is turning into the Daily Mail of science reporting.
Looks like we need a good humanity disaster recovery solution in another solar system. Does sunguard have something yet?
Shadowfirebird is in the absolute right of it. It's 'Inconstant Moon' , though
First appearance: 1971 short story collection All the Myriad Ways.
Stan, the narrator, notices that the moon is glowing much brighter than ever before. The people he meets as the story begins all praise the moon's increased beauty but lack the scientific background to understand its cause. However the narrator surmises that the Sun has gone nova, the day side of the Earth is already destroyed, and this is the last night of his life. He then calls and visits his girlfriend Leslie, presuming her ignorant of the situation, but she realizes it independently when Jupiter brightens with appropriate delay; they then enjoy their last night on the town, before rain and winds start.
Later, he realizes one other possibility. In case he is right, they find appropriate supplies and seek refuge from the coming natural disasters in Leslie's high-rise apartment. The second possibility turns out to be correct: the Earth has merely been struck by an enormous solar flare. The vaporized seawater leads to torrential rains, hurricanes and floods. Most (if not all) people on the Eastern Hemisphere are presumed dead. The story ends at the break of an overcast, gray morning, with the narrator "wonder[ing] if our children would colonize Europe, or Asia, or Africa".
At this time, we appear to be just emerging from a very deep lull in solar sunspot activity. Looking at the past records I see that this similar to those of 1811, 1823, and 1856 -- three years before the last Carrington event. While Carrington super-flares are thought to strike the earth, on average, every 500 years, we may have a higher than average probability of being hit in the next three years.
Let's hope the sun is pointed in another direction when the next one goes off. As far as I can tell, there has been no effort to prepare for any such event.
Listen, we all are just conjecturing here. We really don't know our sun as well as we'd like. The next few years will open up an incredible amount of new stuff, as we send up more labs into space. There may be an effect by our sun which can create a massive power collapse. What if circuits were wiped out? How long would it take to replace them? What if these kinds of CME's continued for up to an entire week, or month? We' be at minimal power output, at best. What if terrorists took advantage of this somehow? We need to get smart about our grid, and not be so hasty in thinking things will be fine. We could help by constructing independent power grids/sources. Each house can be supplied by most of its own power, that way, if some massive cme strikes, the main grid won't be fried!
Schneier.com is a personal website. Opinions expressed are not necessarily those of BT.