Hacker News new | past | comments | ask | show | jobs | submit login
Near Miss: The Solar Superstorm of July 2012 (nasa.gov)
234 points by jamessun on July 24, 2014 | hide | past | favorite | 157 comments

I wish the article spent some time discussing how we can defend against the effects of storms like this. It mentions a 12% chance of getting hit by one in the next 10 years (and I assume that percentage is much higher for a probability space of the next 50 or 100), but makes no mention of potential mitigation strategies.

Can you imagine how disastrous this would be? People would likely wake up without power, and think little of it at first. "Oh, it's a power outage. Better light some candles and wait it out." Then hours, days would pass. There would be no internet, television, or even newspapers to report what happened. We would be reduced to hearing the news from Shakespearian harriers!

I wonder if this would cause mass revolt. What happens when you completely disable modern technology on all of earth? If this is as likely as they say it is, there needs to be some serious research into mitigation strategies not only of the immediate technological effects, but also of the sociological effects.

Edit: maybe not all of earth, as the NASA page compares this to a localized blackout in Quebec. Can anyone comment on the geographical range of damage?

I attended a lecture about this a few years back. NOAA has a satellite system to warn us about large solar flares... but it's old and severely underfunded. Which is somewhat surprising, because that 20 minute heads up is the difference between GPS satellites and electric grid components being safely taken offline vs permanently damaged.

Yeah, when I started getting involved in web operations tools, and realizing all the levels of redundancy that the big players like Google and Netflix have, it really got me thinking about our hydro and communications infrasture.

For instance, Netflix has it's chaos monkey, than randomly kills servers and availability zones and generally wreaks havoc. They set it loose, and learn to keep things operating. I really thinkg it would be a responsible societal effort to introduce chaos monkeys into our archaic infrastructure outside digital.

And as for that underfunded satellite, it's sitting in the sweet spot between the gravitation tug of the earth and sun called the L1 langrangian point, where it takes no effort to keep it exactly in place. But we really need more drills to learned whether the hydro/comm infrastructure on one hemisphere of the earth can actually shut everything down in 20 minutes, day or night.

The spirit of the idea is good, but it would not work well for hardware. With software, if something breaks, you restart the servers. With hardware, if something breaks, you have to buy a new one. Throwing chaos monkeys to see which transformers blow up would end up quite expensive. :)

Then at least simulations on how it affects electric grids coyld help quite a bit. Are this simulations taking place?

I interned at Areva T&D in 2006 (now Alstom Grid). While I was not working on that specific project, I remember they had a very powerful EMS (Energy Management System), which let them simulate incidents (e.g. "what happens if these power lines are destroyed").

Doing a quick search, this seems to be called a "dispatcher training simulator": http://en.wikipedia.org/wiki/Dispatcher_training_simulator




Haha, was thinking the same thing. This would make a great xkcd.

The 100-year old underground electric wires running under the street where I work have caught on fire three times in the past year. Something tells me our archaic infrastructure has quite a backlog of existing chaos monkeys to deal with already.

The relevant department within NOAA is the space weather prediction center, http://www.swpc.noaa.gov (beta site: http://origin-www.swpc.noaa.gov).

I toured their facilities in Boulder, CO a few years ago (I develop a satellite data product, http://link.springer.com/article/10.1007%2Fs11207-014-0529-3, that is used in forecasting). The forecasting is done partly with numerical models, partly with observational data, and partly with old school intuition and experience. It partly involves a small group of grizzled experts sitting in front of a bunch of monitors displaying images and time series.

They are trying to automate it as much as possible and bring in new data sources, especially near real time data sources, from NASA and NOAA satellites. It's too much to go into in the scope of a comment, but suffice it to say that, with as many photons as the Sun puts out, you can gather a lot of information about the entire 3D solar atmosphere and magnetic environment. One day, these measurements would feed a forward model of the same type used in terrestrial weather prediction.

There are a lot of modern facilities providing solar flare predictions, many are old and many are new. Some are space based and a lot are ground based (radio telescopes and coronagraphs) Government agencies amalgamate data from all the sources to come up with the best predictions. We always have more than 12 hours to prepare after spotting something from surface on the Sun, the accuracy of predictions improves as time goes by, the 20 minute mark is generally the time it takes for a fast CME to go from L1 Lagrange point to Earth. L1 is where multiple aging solar observatories collect insitu data, but we usually know before then if a strong event is Earth directed.

What could we do to satellites in 20 minutes that would protect them from a coronal mass ejection? It doesn't seem to me that simply powering them down would protect them.

under adverse space weather conditions, launch personnel may delay a launch, and satellite operators may postpone certain operations (e.g., thruster firings). For the spacecraft industry, however, the primary approach to mitigating the effects of space weather is to design satellites to operate under extreme environmental conditions to the maximum extent possible within cost and resource constraints.


We can't save satellites but when those begin to fail, we have 20 minutes to power down things before earth itself is hit.

You can save satellites - you can order them to rotate instrumentation away from the flare vector, and you can shutdown sensitive electronics or high-gain radios and the like (most satellites have many levels of shutdown for exactly this reason).

> Can you imagine how disastrous this would be? People would likely wake up without power, and think little of it at first. "Oh, it's a power outage. Better light some candles and wait it out." Then hours, days would pass. There would be no internet, television, or even newspapers to report what happened. We would be reduced to hearing the news from Shakespearian harriers!

> I wonder if this would cause mass revolt. What happens when you completely disable modern technology on all of earth? If this is as likely as they say it is, there needs to be some serious research into mitigation strategies not only of the immediate technological effects, but also of the sociological effects.

I really doubt it would be that terrible. It's not like electricity would have completely disappeared for ever (like in this great novel apparently called in English "Ashes, Ashes" [0]).

Important things and urgent things like communication means would quickly be operated by emergency generators, so people would immediately have information through radio (and most of the receptors are battery operated), and probably also TV, although they would probably have to go to the local townhall, hospital, school or whatever place would have a working generator to watch it.

Cars would still work so people could move around, at least while there are refined gasoline reserves, but refineries would probably be one of the first industries to get powered again anyway.

Probably for a good while there would be restrictions on electricity usage, depending on how hard existing plants would have been hit and whether they are fixable or have to be completely rebuilt, but I really don't think it would "destroy life as we know it".

I wonder how nuclear plants would handle this, though.

[0] http://en.wikipedia.org/wiki/Ravage_(novel)

Batteries only last so long. And last time a hurricane took out power in my city, the gas stations quit working.

The biggest problem would be high-voltage transformers. They're made mostly be hand, by people who spend several years learning to do it, and we don't have many spares. If they go down continent-wide, we could be without grid power for years. I'm a little less sanguine than you about how well we'd do in that situation.


> _Cars would still work so people could move around.

Any car that was running at the time of the CME would have all of its electronics fried and become inoperable.

> at least while there are refined gasoline reserves

As another poster hinted at, any gas station in operation at the time of the event would have its pumping equipment fried. Even if the pumps weren't harmed, you still need electricity to run them, so you'll have to wait until the grid comes back up.

> I wonder how nuclear plants would handle this, though.

Yes, could all those aged reactors deal with large scale grid or component malfunction? Fukishima didn't do too well. FUD: Would we have hundreds of Fukishimas on our hands after such event?

A power plant makes its own power, so the grid going down wouldn't affect it. For other components, nuclear reactors have many failsafes. Fukushima exhausted at least 4 that I know of. Since a CME wouldn't cause physical destruction like the earthquake and tsunami did, there's a much lower chance that all the failsafes will fail at once.

A nuclear power plant cannot run at say 10% to just power its own systems after it was separated from the grid. That's why diesel generators and batteries are so important to keep the reactor from melting.

It's hard to find a source on this, I apologize. The reason is that it only has large turbines to generate power, and these cannot be operated for arbitrary power output.

To defend against it you just need a fuse, or a relay.

Put one on each transformer that cuts power if it exceeds limits, this would save the transformer and would allow us to restore function pretty quickly.

That seems like a cheap solution that would protect not only against solar storms, but lightning strikes as well. Are these fuses common in transformers? If not, why?

Here is a video on high voltage circuit breaker evaluation and maintenance. More then you ever wanted to know I'm sure.


These systems are complicated, hard to maintain and expensive.

Typically, your fuses at that scale aren't fuses per se, but mechanical separation of conductors.



I don't think they have them - just watch any transformer fire fueled by electricity.

I say "fuse or relay", but for the power levels involved these are NOT simple components.

I work in commercial electrical power. No they don't. Because they aren't required to. It's all about costs. So unless it's about known risks (read PRA), no they don't protect against it. You need to spend money where the risks happen.

Wish I had better news, but hey you protect against "real" risks, like LOOPs, trips, downed trees, lightning, regulators, and floods. Well I guess the quotes are unfair, these things happen often. So these are real things you want to mitigate. End of the world stuff, no one takes that seriously sadly.

Also, don't forget a lot of the protective relays are digital protective relays. So even if you have them, well they might not work in This case. So aside from the digital PRs being "better" companies still sell them and few sell the older SD models, so we end up replacing the old solid state relays with digital drop in ones.

So even if you have a PR it might not work in this case. And frankly few people in the industry, or with the regulators, takes a geostorm seriously in commercial power, even though they should.

It's not "cheap" when you have to install a breaker for each one. For the companies it's cheaper to have a transformer blow up every now and then.

It'll only be considered an investment when the damage can out-weight the cost. So it's likely they'll only learn the lesson after the first solar storm hits, unless legislation demands they put it now.

>>no mention of potential mitigation strategies

The biggest risk is if you think of it on the very long term, staying on earth is like risking it all on staying on one rock. There are far too many risks, and there is little you can do mitigate them on the long term.

The long term solution is to invest money to build technologies to develop inter stellar travel and related technologies to settle humans outside earth. And that really wasn't that difficult to achieve. The problem is we haven't made serious attempts beyond the moon and mars missions. The funding is too limited, and there is no political consensus around the world to work towards these goals. We have become far too busy with wars, conflicts and other pointless spending on earth. While we could have used the money to do something like this.

>>Can you imagine how disastrous this would be?....

It can be far more disastrous than you imagine. Loss of electricity and water are just primary effects. There are many secondary effects of these scenarios. No electricity means you can't pump out fuel. Bulk of your critical production[Food/Medicine/Sanitation etc] Will come to grinding halt because of inter dependencies. Expect mass starvation, riots, loss of law and order kind of scenarios. There will likely be powerful groups controlling critical resources required to survive.

I wouldn't say its a extinction scenario. But it will come pretty close.

It would be far cheaper to build underground bunkers on Earth, or fund prevention of specific risks like this one. Building a permanently self-sufficient colony isn't just ridiculously expensive, it may not even be possible with our technology for a very long time. And it's still only a last resort. Civilization and everyone on Earth still perishes in the apocalypse.

It would take significantly more than this to wipe out humanity btw. Even the worse case scenarios, civilization still recovers eventually.

Sounds exactly like the now canceled TV show "Revolution".

That's it. I've had it with cancelled shows. I'm no longer starting any shows until the series finale has aired.

You could build an oven in your backyard and stock up with a load of wood ;) and make friends with your local livestock dealer. Just have a plan for how you're going to still make money once everything shuts down.

Start here: https://en.wikipedia.org/wiki/Maslow's_hierarchy_of_needs and work upwards.

Hint: you'll need guns. Lots of guns.

Cities are stocked with food and water for few days, but once the supplies are gone, you'll have waves of hungry and desperate people radiating from cities to nearby farms and villages.

I think you're right that guns are probably a necessity to some extent, but I would rather spend the money on feeding my neighbors as much as possible rather than fighting them. I ran some quick numbers on webstaurantsupply.com, and for a simple "donut diet"[1] you can feed 100 people 1800 kCal/day for nearly a month for $1000 (the price of a decent quality rifle where I live).

I've been trying to work out a way to organize neighborhood-level basic essentials storages on the cheap, so that people in that kind of situation can afford to take care of themselves until the power comes back on. Food is pretty easy, but water and sanitation are a bit more problematic so far.

[1] - just calories to make it over the hump, this is not health food. Ingredients are white flour, white sugar, vegetable oil, and some white rice for the people who can't do the wheat products. Add vitamin pills to prevent deficiency diseases and you can stay functional for a long time at about $0.30/day.

the point is, money stops being worth anything in this scenario. you can't eat it and you can't drink it. money is based on trust and when civilization collapses, trust disappears.

Everybody realises that, the point he was making was that $1000 is better invested in storage now than a gun.

Everybody assumes that in an end of the world scenario that we all end up fighting for food, however the better plan, and to be honest the one that served society well in the past, was that we become farmers. Without tractors we'll have plenty of farmland that needs manual labour.

It would be nice if it was codified into law that our fallback plan was to redistribute farmland. Then there is no need for roving biker gangs fighting for the last can of tuna. We'll just step away from the now dead computers and pick up a hoe.

Bootstrapping the reeducation of enough people to effectively get them intensive-gardening in that kind of a situation would be a massive challenge, but if enough people did the feed-your-neighbors preparedness approach we might have enough time to get it done. If we were looking at a long-term scenario I think I'd want to have at least a year of grace period to get people rolling. Thinking back to all the mistakes I made starting my own farm, if my life had been depending on success I'd most likely be dead, or would have survived solely by eating my sheep...

Cuba is an interesting example of a place that had to rapidly adopt a low-power lifestyle, and they actually seem to have done it with grace. There might be a good case there to study for clues about what works. They have the advantage of not having to contend with long, cold winters though, which is a critical difference for most of North America.

Was your lack of success partly due to no seeking the knowledge of other or not being give the help?

I would imagine that secretly farmers want other local farmers to not do as well, so that there is less competition for their produce, however you'd know better than me.

Except that you need to survive the months between planting the seeds and harvest. Also when 90% of crops fail because of lack of modern agriculture tech and farming experience, don't expect big part of people to just lie down and die from hunger. You will have fighting until the food supply stabilizes (which does involve cutting down population a lot).

That doesn't make sense, why would the current crops in the ground not be usable? Why would the food in storage right now not be usable?

Saying that 90% of crops would fail is kind of an over statement. Organic farms that don't use fertilisers are quite successful. The issue is planing and sowing, all of which the extra man power can assist with. Instead of a tractor you have people.

All that's needed is a plan, where farmers us the extra labour until machines can be gotten back on line. No need for a lawless society.

Other than the breathless headline, WaPo adds nothing to the original reporting. Suggest to change the link to http://science.nasa.gov/science-news/science-at-nasa/2014/23... and use that article's title (“Near Miss: The Solar Superstorm of July 2012”).

I agree. Although reading some of the comments on WaPo makes me happy to be on HN.

I couldn't agree more.

Reading the comments on wapo wherein individuals ridiculed the NASA report as pseudoscience and compared it to global warming (as deniers) utterly depressed me.

Then I read the HN comments where there is an evidence driven debate over the extent of the possible impact of such an event (generally concluding utter catastrophe), and my hope is restored.

I previously did research forecasting solar storms.

original comment: Highly unlikely, they compare it to the Carrington storm of 1859, which didn't destroy any life (or even life as we know it), so even if it were more powerful, it probably wouldn't have destroyed all life. It may have interfered with a lot of satellites and communications systems though.

edit: Any damage from a Carrington level event would likely have been localized and temporary. For life as we know it to be destroyed, it would be irrecoverable or permanent, which is why I am arguing against the statement "life as we know it would be destroyed."

Many satellites are exposed to the open solar wind all the time, and ones in low Earth orbit, though shielded to some degree by the magnetosphere are designed to deal with large solar flares either by baffling sensitive instruments or shutting down temporarily (some may even change their orbit).

National Academy of Science predicted a Carrington-level storm would inflict $1 to $2 trillion in damages to infrastructure and take four to ten years to recover from.

I don't know what "life as we know it" means, but it would be a significant disaster.

Here's the report: http://www.nap.edu/openbook.php?record_id=12507 (it was funded by NASA)

I imagine that would be an extreme upper bound. But yes, it would cause damage, economically and possibly damage communications. We'd recover and I don't think it will take 10 years. We're continually reducing the costs of launches which is a huge factor in the monetary damages (potentially relaunch all communication satellites).

We'd be out billions of dollars in spy satellites too so maybe we'd triage on what we decide is important to launch after such an event.

One significant detail that has been mentioned in these discussions has been the provisioning of new transformers. Apparently power companies don't keep a huge backup supply of them, and they take a lot to build. If every transformer in a metropolitan area exploded all at once, then we'd be in quite the pickle. Add that up with every transformer on a single hemisphere exploding at once, and it becomes a little more grim.

The phrasing is "life as we know it", which is much less severe than actual life.

Though I imagine hospitals and other medical facilities might encounter serious issues, including inpatient deaths, were such a big solar storm to occur.

I still don't think this is accurate. I say this as someone who used to study solar physics and worked with people on the STEREO team.

We've seen lots of powerful solar events strike the Earth and so far, even the ones that have struck in times of modern technology have only caused localized issues, and those localized issues were usually a result of some level of poor design.

> Carrington storm of 1859

Isn't this storm supposed to have set telegraph wires on fire? Why would the magnetic field involved here not do the same to our electric grid? Wouldn't most transformers be overloaded / damaged by the additional electricity induced on the grid?

Or is this less likely to happen everywhere? It's been a while since I studied electro-magnetism, but IIRC a magnetic field induced a current orthogonal to it's direction. Does this mean that not all transmission lines would be equally affected?

The telegraph network back then had very poor shielding as did other electic components. We've come a long way in 150 years.

When there are ice storms in North Carolina where I grew up transformers would explode by the dozen each night but this didn't keep the power off for more than a few days. I doubt a flare would destroy transformers wholesale but even if it did we'd probably not be too bad off for long.

The phrase 'life as we know it' usually means "humanlike intellects made of meat", not 'comfortable Western lifestyle'.

I read it[0] as compared to "life, but not as we know it", meaning alien life, and therefore thought it meant all carbon-based, DNA-powered life of the type found on Earth would be destroyed. (Except for the damn cockroaches, obviously.)

[0] Edit: The HN story title has changed, and was originally "A solar storm nearly destroyed life as we know it two years ago"

Destroyed life as we know it means that it would've broken satellites and the power grid. It does not mean that there would've been a direct loss of life from it. Just that our lives without GPS and power would've been radically different than today with those things.

I understand, I'll edit to make myself clear.

For those interested in an in-depth analysis of the risk, Lloyd's insurance produced a detailed report last year looking at this issue:

Solar Storm Risk to the North American Electric Grid [PDF]


Summary here:


Some key points:

Weighted by population, the highest risk of storm-induced power outages in the US is along the Atlantic corridor between Washington D.C. and New York City.

The total U.S. population at risk of extended power outage from a Carrington-level [estimated to occur every ~150 years] storm is between 20-40 million, with durations of 16 days to 1-2 years.

The wide variation of expected duration is is dependent on the number of transformers destroyed:

If spares are readily available, the total transportation and setup time for a large power transformer can range from a few weeks to months depending on distance and logistical issues. If new transformers need to be ordered, the lead-time is estimated to be between 5-12 months for domestic suppliers, and 6-12 months for international suppliers.

I am wondering: if the large power transformers are the most critical asset, wouldn't it make sense to try to protect them in Faraday cages capable of sustaining high levels of current without melting down? The cost of such cages are likely to be much much smaller than the cost of power outage lasting months or years for millions of people.

A user (growupkids) from the industry posted on this below:


>I work in commercial electrical power. No they don't. Because they aren't required to. It's all about costs. So unless it's about known risks (read PRA), no they don't protect against it. You need to spend money where the risks happen.

That would not help. It's the wires attached to the transformer that get the current surge, not the transformer itself.

Plus a Faraday cage would not do anything anyway, the frequency is too low.

> That would not help. It's the wires attached to the transformer that get the current surge, not the transformer itself.

Wouldn't it possible to add automated circuit breakers in case of high current surge then?

> Plus a Faraday cage would not do anything anyway, the frequency is too low.

Low frequency means long wave length, no? I thought that the Faraday cage holes needed to be smaller than the wave length for it to work.

> Wouldn't it possible to add automated circuit breakers in case of high current surge then?

Yes, but a circuit breaker at the voltages and/or power levels involved is not a simple device. They don't have them (due to cost and complexity), and there is a lot of criticism that they don't.

> Low frequency means long wave length, no? I thought that the Faraday cage holes needed to be smaller than the wave length for it to work.

It's both. If the holes are too large then small waves can go right in, but if the wavelength is very large the wave doesn't even notice the cage is there.

A Faraday cage works by picking up and distributing the EMF on all sides of the cage, so the field inside is canceled out. But if the wavelength is too large the cage can not pick up the EMF, and the EMF is present inside the cage. (However with a wave that large is not clear it would do anything to devices inside. The nice long wires of the power grid are perfect for it though.)

Thanks for the clarifications.

So at the moment, there are no transformer spares.

The outage would be 5-12 months to produce them and weeks to months for transport them and install them.

Even with spares, the expected outage can be between weeks and months if the spares aren't near enough.

There are still no spares now.

From the conclusion:

"Since the 1989 Quebec storm and power outage, the Canadian government has invested $1.2 billion (about $34 per person) into protecting the Hydro-Quebec grid infrastructure, installing numerous blocking capacitors44. While these mitigation strategies can be expensive up front (estimated cost of $100k per blocking capacitor for a total of $100 million to protect the 1,000 most vulnerable transformers45), the cost of prevention is much smaller than the cost of the damage a single storm can create."

It seems only Canadians are doing at least something?

In II-1-5 [1] we learn that magnetism is a relativistic effect whose strength is v^2/c^2. This ratio is a very small number for the electron drift velocities associated with typical currents in copper wires, yet we perceive magnetic forces because positive and negative electric charges in the wire are balanced to better than v^2/c^2.

Can anyone indicate how + and - charges are distributed in a CME? Are there strong electric fields due to local charge imbalances? For the magnetic field, it's tempting to say that the drift velocity of charges in a CME is high (as much as 0.001c) but then wouldn't charges of both polarities be moving with that (average) speed? Perhaps there is some initial motion when the charges are ejected from the sun?

[1] http://www.feynmanlectures.caltech.edu/II_01.html#Ch1-S5

Thank God we live in a 3 dimensional space, where things rarely collide because the space is so "big" compared lower dimensions [0].

[0] http://en.wikipedia.org/wiki/Random_walk#Higher_dimensions

And yet:

> There is a 12 percent chance of a Carrington-type event on Earth in the next 10 years

From the linked NASA article:

> "... knock modern civilization back to the 18th century ..."

and later:

> "According to a study by the National Academy of Sciences, the total economic impact could exceed $2 trillion or 20 times greater than the costs of a Hurricane Katrina."

Is that figure for the US, or the world? Also, while that's a lot of money, it's not "knock modern civilization back to the 18th century"-money. So which one is it? Neither the NASA or WaPo article seems to give a clear answer.

It would have wiped out the power grid, no power to anyone for months. (Unless you had a generator and fuel.)

Although we would fix it. Assuming current levels of production it would take years to fully repair, but I assume after a disaster we would massively increase the rate of production and it wouldn't be as bad as some make out.

Poorer counties, and those without a manufacturing base, would be in much much worse shape though.

>>I assume after a disaster we would massively increase the rate of production

There is a chicken-egg problem here. You can't massively increase production because, you don't have enough energy to do that. And that is because your existing infrastructure is fried.

For a few months/years at least you will have to bootstrap a lot of transformers to increase supply to produce more transformers. And only when you break even on a good enough energy you will go back to the normal.

However note, in order to increase production you also need to mine quickly for you don't really have energy. So there are a lot of sub dependencies.

But by any measure lack of energy for a such a long time, would have already set back world economy to way back in the past.

> You can't massively increase production because, you don't have enough energy to do that.

That's not really true. The grid might be down, but this isn't a SciFi story - regular generators will still work just fine.

The larger generators the power the grid will also still be there and working fine - so just figure out a way to jury rig a connection to one.

No power grid = lots of difficulty getting fuel into vehicles = much less transport availability = very, very serious problems for modern societies. Reduced transport capacity isn't just reduced ability for people to get to work - it's also reduced ability to distribute food.

You'd also have people panic-buying at supermarkets and hardware stores - but without the ability to electronically transfer money... you've got a lot of public panic to contend with. Food riots, crime skyrocketing. Lots of ugly stuff would happen before you would be able to do much manufacturing to recover (and most manufacturing these days needs a lot of power...)

It would also mean that many people with chronic diseases--people you all likely know, yes?--would die because supplies of medicines would dry up rapidly, existing stocks of some drugs would spoil (insulin requires refrigeration), and further production would be offline for months or years. So yes, I'm concerned.

To the naysayers here, don't your jobs require access to electricity? How's your Node startup going to survive months or years without power? (No elevator pitches if the elevators aren't working!)

The article suggests that we'd get up to a day of warning. Would it be practical to physically disconnect transformers and other equipment from the grid to protect against damage? Also, how exposed are subterranean power wires?

You would have to shut down the power grid of the whole world for 2 days.

That would feel like the apocalypse if it happened. It could certainly be done in time, but doubt you could convince people it was real.

There would be a ton of deaths, so it's not something to do lightly.

I must admit that I have a pretty high degree of confidence that if the director of NASA briefs the president that this is happening (which he will know with a very high degree of certainty), such a shutdown would be effected - and similar for pretty much every other civilized country on the planet.

Any deaths that would happen from shutting down the power grid would happen anyway, but this would allow for mitigation strategies to be put into place. Hospitals already have emergency generators, so it not like they're strangers to the concept. Also, the 2003 northeastern blackout affected 55 million persons for up to two days and only ten fatalities were attributed to the outage.

Anyway, this seems like a good thing to have a contingency plan written up for. The NASA article suggest there isn't, but that just might be that they don't know about it - they wouldn't be responsible for executing it anyway.

It seems much more likely that the grids around critical infrastructure would be taken offline, but major metropolitan areas would be left connected. In a TEOWAWKI scenario, we would expect a government to prioritize its military capacity ahead of the comfort of the civilian population. Also, the riot and crime risk would heavily weigh against any actions to take cities offline, not to mention the PR nightmare of being wrong.

It seems "the grid" is a singular thing here. Either you shut it down completely (and then back on later), or you lose it completely.

If you leave cities connected, and the solar solar storm hits, then the cities are without power for a much longer period of time than if you took preemptive measures - with all the risks of crime that follow. Also, I'll refer to the 2003 blackout for a dry run of the mayhem that will unfold in the face of a vast power outage (ie. not much), and in this case, you'll even be able to provide a quite fair number of hours warning, so most people should be able to get home - you won't have a million office workers stranded in Manhattan. Hoarding will happen, though, and it won't be pretty. That survivalist nutcase guy at the end of the street will have a field day when he struts out of his well-stocked bunker after the power comes back on.

The PR nightmare of having shut down the grid for a few hours with warning (presumably, you'll shut it down immediately before the hit is expected, so if you're wrong, you can turn it back on fairly quickly[1]) is much, much lighter than frying the entire nations electrical grid, causing trouble for years because you acted against advice from your NASA director. Also, if he was wrong, the NASA director will make a pretty good fall guy and it won't even be all that unfair. If your PR guy can't spin "I listened to the best scientists in the country and followed their recommendations, I'm sorry they were wrong, but better safe than sorry", fire him too.

1: Spinning a grid back up isn't anywhere near instant, I know, but it'll be a lot faster than having to rebuild the grid first.

Nukes can shut down pretty quickly but take a long time to get back online.

Also, if enough of the grid goes down you'll have black-start issues in places.

Not true. You can turn off the power in a nuclear plant pretty much instantly. You can turn it on pretty much instantly. What you can't do is efficiently turn it on and off - it takes time for the reactor pile to warm up after you've dropped the control rods to shutdown.

But there's no specific reason you need to do that - the reactor won't be effected, the control rod circuits are heavily shielded and fail safe these days (i.e. if power does go, the arms holding the rods drop them in automatically when the electromagnets fail).

For a day or 2 of shutdown, you could leave the reactor hot.

Sorry but you're simply wrong here about the conditions of the shutdown. These are not planned months ahead shutdowns they are near instantaneous emergency shutdowns and it takes a while to recover from those. After the 2003 blackouts the nuclear plants took the longest to be brought back on-line.

This isn't a reactor SCRAM though - or at least, it doesn't have to be. It's a momentary need to disconnect the reactor generators from the electrical grid.

If you planned in advance (to manage the reactor's thermal state - i.e. keep the heat exchangers going) in such a situation, then all you're doing is disconnecting the transformers.

The problem is no one's planned ahead for this type of shutdown of a nuclear plant. The procedure you follow is the one you have since you definitely don't want to make it up on the fly with a nuclear plant.

I think that we can agree it's much more preferable than catastrophic damage to the grid that will take years to recover from.

>but I assume after a disaster we would massively increase the rate of production and it wouldn't be as bad as some make out

Unfortunately it's not so easy. A number of the raw materials have quite limited supply and even after production, transportation is very time consuming.

Transportation can be shorted by producing locally, but in 2010, only 15% of large power transformers (>60 MVA) deployed in the US were supplied by domestic producers[1] (although domestic production capacity is increasing):

LPTs require a long lead time, and transporting them can be challenging. The average lead time for an LPT is between five and 16 months; however, the lead time can extend beyond 20 months if there are any supply disruptions or delays with the supplies, raw materials, or key parts. Its large size and weight can further complicate the procurement process, because an LPT requires special arrangements and special rail cars for transport.[1]

For more than you ever wanted to know on the subject, see the following:

[1] http://energy.gov/sites/prod/files/2014/04/f15/LPTStudyUpdat...

I don't think it would really be this dire if push came to shove. A whole country without transformers and lacking ability to import them would be like a war effort. It would certainly be a huge disaster, but it wouldn't be cataclysmic. People would come out of their regular job roles and work with rebuilding.

If you read up about e.g. the crazy efforts that went on during World War II, this would be a good example of how focused a large economy can be in times of crisis. Of course there would be less time available to make the next Instagram, but it would probably have some positive side effects too.

> raw materials have quite limited supply

The high quality (i.e. efficient) version (electrical steel) has limited supply, but if you didn't care about that, and just wanted something to work now, there is much more supply. All the other materials are common and widely available.

> Its large size and weight can further complicate

I would expect the military to step in with some heavy lift vehicles.

All of what you say is for normal times, but in an emergency (i.e. cost is no object) we could do things a LOT faster.

The power grid where? In every country? Or just one/a few?

Probably every country.

So why was the 1989 storm localized? Just because it was much smaller? Do those metrics for measuring the six did the storm correspond to geographic spread?

It was not localized (It affected both the US and the UK), but rather Canada was uniquely sensitive to it because of the very long transmission wires and that there is no conductive soil to earth the power.

These days everyone is "uniquely" sensitive to it.

> Is that figure for the US, or the world?

Given it's the National Academy of Sciences (not International), given how inward facing the US usually is and given that it's only 20 times greater than Hurricane Katrina, I'm thinking it's only the figure for the US.

Of course, if the US happened to be facing away from the sun when it happened (night), based on other comments here I'm assuming the damage would be much less severe.

Let's see: Say that one of these storms arrives at a point in time. Then over time, these storms form a stochastic point process.

Now it should be fairly clear intuitively that at least roughly increments, that is, arrivals of points, are stationary in time and independent.

Presto: It follows that the stochastic point process is a Poisson process about which we know a lot.

So, we'd like the arrival rate of this Poisson process, say, in arrivals per year. Okay, we can get a good estimate just by dividing the number of arrivals we have observed over the number of years we have been observing. Then with this estimate of the arrival rate, we can calculate the probability of an arrival in, say, the next 10 years.

Since, when was it, 18?? or some such, we've actually had no arrivals at all that hit Earth, the arrival rate is low and the chances of an arrival in the next 10 years is low.

Relax! Good news, right? I mean, the sky is not really falling this time after all!

Bad news for this model: it's not stationary, due to the 11-year solar cycle, and to other behaviors at longer time scales superimposed on the cycle.

But if you average out over at least the solar cycle, you could come up with a rate of superstorms per-cycle. There have been about 15 cycles since the 1859 superstorm, which would mean that perhaps you expect ~ 1 chance in 15 of such a storm in the coming cycle.

Of course what I mentioned is also called the 'axiomatic' derivation of the Poisson process. In part it's nice because it starts with 'qualitative' assumptions and ends up with the full algebra of the process and its distribution and independence of times between arrivals.

Details are in E. Cinlar, 'Introduction to Stochastic Processes' in his chapter on the Poisson process. There he also touches on the non-stationary case, and what he has may (I'm in hurry this morning) be equivalent to your "average out over ... the cycle".

Since a finite sum of independent Poisson processes is a Poisson process, your 1 in 15 over the next cycle may be equivalent to 1 in 2014 - 1859 years over the next year.

I do the same for asteroids: We have a good ballpark estimate of the rate of killer asteroids, and the rate is low enough to f'get about it.

Yes, you described a case ("homogeneous P.P.") where the intensity, lambda, is constant over time. In the inhomogeneous P.P., there is a time-varying rate, lambda(t). The derivation of the properties in the inhomogeneous case is almost the same because the key property is independent increments.

When you add up the number of catastrophic events over an 11-year cycle, you get a sequence of "numbers of catastrophic CMEs"

  N1, N2, ..., Nk
each of which is Poisson with intensity

  \int_{cycle} lambda(t) dt    [*]
If you ignore the larger-than-a-cycle variations in intensity, this integral is the same for each cycle. So the N's above are iid Poisson with some common parameter lambda_0, say.

Observing the N's for many cycles would allow you to estimate lambda_0. That's pretty much where the 1 in 15 comes from (indeed, lambda_0 is so low that there are only 0 or 1 events per cycle, so "estimation" is trivial).

The "easy" problem is that, even averaging over a cycle, the activity ([*] above) is not constant, e.g. this graph of sunspot numbers vs. time:


The "hard" problem is that the counts are not, in real life, independent, even across weeks-to-months time scales. This would invalidate the Poisson model at the ground level (see the talk referenced by @gone35 above).

There are reasons (e.g., Maunder minimum, see the plot above) to believe that the counts are also not independent at the decadal time scale.

Lack of independence would be a huge problem for any Poisson model. You'd have to go up to a doubly stochastic process, for example, in which lambda(t) is itself a random process.

You are guilty of letting a lot of really messy data ruin some really nice theory!

So, with assumptions of stationary and independent increments, the Maunder minimum shown in the graph would have probability less than zip and zilch of happening in which case we would reject the assumptions.

You also are analyzing sun spots while the OP is talking about dangerous CMEs. It's been a while since I studied sun spots: You are saying that a sun spot is just a small case of a CME? Or is a dangerous CME not just a big sun spot but a different animal?

CMEs are associated with large or magnetically-complex sunspots.

Simple sunspots are bipolar, they have a N and a S region with field lines arching up connecting one to the other.

Complex sunspots are multipolar, and the field lines therefore cross in weird ways (in the 3D solar plasma). They can change configurations suddenly, releasing a lot of energy as the magnetic field lines re-arrange to a lower-energy configuration.

This magnetic reconnection is what drives CMEs (roughly speaking).

It should be fairly clear intuitively that at least roughly increments, that is, arrivals of points, are stationary in time and independent.

Far from it, insofar as the (time-dependent) solar cycle modulates the storm rate [1]. Besides there is some evidence of dependence between the flares themselves --strong enough to manifest itself in a complicated waiting time empirical distribution [1,2].

Be careful with the common --but nefarious-- habit of adducing stationarities and 'nice' distributions unwarrantedly. In the meantime, y-you are not allowed statistics anymore, sorry.

[1] http://www.physics.usyd.edu.au/~wheat/talks/fstats_agu_2005....

[2] http://researchcommons.waikato.ac.nz/handle/10289/1382

> Far from it,

What? While I'm reading your references, let's be clear on what you are saying: On the one hand, in the context of the OP, we're talking about flairs, as essentially discrete events, 'points' in a 'point process', over time intervals from the present back to at least 1889 or whenever it was, that is, over 100 years, and, if we could get more data, thousands or millions of years and (2) what happens in detail with the sun spots, the 11 year sun spot cycle, the magnetic field of the sun, etc.

So, sure, on short time scales, say, within one 'point' there can be a lot of dependence, that is, lack of 'independent increments'. But for these 'points' over years, centuries, or millennia? You are saying that there is a dependence between a flair today and 11 years from now?

Assuming that the 'dangerous flairs' are more common when sun spots are in the 11 year cycle, for modeling the stochastic arrival process over several decades, centuries, etc. we should, first-cut, and I saw a lot of first-cutting when I was in physics class, 'smooth out' the 11 year cycle and continue on. Then, tough to believe that you are claiming dependence between points separated by months or years.

E.g., I see in your [2] in Figure 2 on page 463 a distribution of waiting times with the horizontal axis a log scale up to 10,000 hours but clearly nearly all the probability mass less than 100 hours, here being sure to note the log scale on the vertical axis. So, in the context of the OP, that 100 hours is within just one particular 'point' in the stochastic point process and essentially irrelevant when considering years, decades, centuries, etc.

So, the independent increments assumption seems to continue to hold. And with the smearing out, so does the stationarity. Then, from Cinlar, presto, a Poisson process.

Oops: Apparently the Maunder minimum, as in the graph below in this thread, kills off the stationarity assumption, even with the 'smoothing'.

So, apparently until we can understand things like the Maunder minimum, we're really short on a good stochastic point process model of sun spots and maybe also dangerous CMEs.

Then maybe the best we have is just observe that both humans and life on earth have been around a long time and in some intuitive sense the sun seems fairly 'stable' in some meaningful sense so that we can guess we will still be here a long time. Math fails again!

"Now it should be fairly clear intuitively that at least roughly increments, that is, arrivals of points, are stationary in time and independent."

That's not intuitive at all. They are all produced by the same source, the sun, according to some internal cycle. They are definitely NOT independent.

Tell you what: I'll give you the history of the solar storms, mass ejections, up-chucks, whatever, and you tell me when the next up-chuck is.

If it can't be done, then independence has to hold. If in practice we don't have a clue, then intuitively we have to guess that we have a good first-cut approximation to independence. Here I'm omitting the little derivation in terms of conditional probability -- I've done it too often in the past and am in a hurry this morning.

"The same source" doesn't have to mean much. We can flip the same coin and get independent results.

The internals of the sun are complicated with, no doubt, a lot of randomness, i.e., roiling and boiling. For a good first cut in practice, at least over time intervals longer than, say, a week, we assume independence. Or, I've got a pot of soup overheated and boiling on the stove, and occasionally it splatters and puts a mess on the stove top. When's the next splatter? Who knows. So, first-cut, independence.

Sounds like there's a bit of "no one ever got fired/pilloried for claiming that things were going to be worse than they turned out to be!" going on. I remember a lot of that going around in 1995-99. Study the issue, advocate for mitigation strategies, by all means. But let's dial back the "WE'RE ALL GOING TO DIE LIKE LEMMINGS" a bit, hmmm?

This is true, it's also reflected in weather reports where there is a bias to forecast rain. Nobody is mad that they brought an umbrella.

As the event lasted only 4 hours, does that mean it would have had different levels of effect on different regions of the planet? Countries on the far side would be better shielded.

I don't see mention of 4 hours, but presumably that's how long it took to either erupt or transit through empty space.

If it hit the earth it would have slowed down and stayed longer.

The charged particles interacting with the magnetic field of the earth would have produced a planet wide EMP (larger than even a nuke could do). The charged particles would have induced electric current in any large wire and would have acted on the whole planet.

> I don't see mention of 4 hours, but presumably that's how long it took to either erupt or transit through empty space.

It took only 4 hours to get from the Sun to Earth, but then it 'slows down' and hangs about?

4 hours was approximately the duration of the CME according to the times under one of the videos.

I was thinking of the 1989 CME, which had localised effects in Canada. But that was much smaller

“This double-CME traveled through a region of space that had been cleared out by yet another CME four days earlier,” NASA says. ” As a result, the storm clouds were not decelerated as much as usual by their transit through the interplanetary medium.”

Just curious, what exactly was cleared out? I thought the interplanetary medium is just empty space.

There are no absolutes in physics (well rarely any naturally curring anyway), space still has stuff in it. Even if it's just a small rock every kilo meter, that's still a lot of rocks. 149,597,870 rocks to be exact. (Assuming the number isn't wrong.)

What was exactly cleared out I don't know though.

It isn't empty, although it has a very low particle density.

I assumed that 'clearing out' means the opposite, that the area now had a higher than normal charged particle density, and thus it was easier to pass through?

I'm also curious if anyone can give a thorough explanation.

If the power was out for a period of months, if the internet, cable tv, phones, and all electronic communication were down so on day 2, you only knew that the power was out and you couldn't reach anybody, what would you need to do to survive?

Wait it out seems somewhat optimistic. No refrigeration means any fresh food would spoil pretty quickly. Better hope that garden can be scaled up.

Clean water would become an issue first though. I don't know where the nearest lake or stream is. Though I imagine several days away on foot.

Assuming you made it that long protection would become an issue. You'd want to be sure you had good relationships with your with your neighbors.

It really doesn't take much to get pretty far down the doomsdayer path I guess.

Lesson learned... optimize for happiness everyday.

> what would you need to do to survive?

Guns. Lots of guns.

After the fresh food and water runs out in cities, you'll have masses of people abandoning urban areas in desperate hope to find food. Any settlement nearby will be pillaged.

So if you have a garden, better be prepared to defend it.

It's still possible to talk to people in person

> Clean water would become an issue first though. I don't know where the nearest lake or stream is

Under your feet. :)

You are so right, it amazes me that people are focusing on getting guns rather than teaching people to dig a well (few days effort, less if others are helping) or plant crops.

Has Minecraft taught this generation nothing? ;-)

> These plasma clouds, known as coronal mass ejections (CMEs), compromised a solar storm thought to be the most powerful in at least 150 years.

Tsk, Washington Post. I think you mean "comprised."

How predictable are these storm events?

There are different factors that make some things predictable and some not. Many models use the solar rotation as a baseline, so some structure that was there in one rotation will likely be there in the next one (this is similar to persistence modeling in climate).

Sudden, large CMEs are not necessarily predictable, but they may be to some degree, most of the solar wind modeling efforts are aimed at determining the time of arrival of a structure on the Sun at Earth (or whether it will arrive) rather than determining if such a structure will happen or not, though as I said, there are some ways to make educated guesses.

Once you think a CME will be Earth directed, the important factors are how strong it is (magnetic field and proton density) and how soon it will arrive. Steady solar wind takes about four days to make the journey but huge solar flares may take less than 24 hours, all told, this is still plenty of time for most satellite operators to prepare their satellites by baffling them or taking other precautionary measures.

Could some explain technically what happens to transformers with a solar storm? Can they not be repaired in some manner?

It's not 100% clear yet and there might be other factors (eg gas leakages or harmonic-induced damages to relays), but it seems the main effect is rapid internal heating caused by stray magnetic flux leakages from the saturated core [1,2]. Hot spots beyond a certain threshold end up severely damaging internal components, rendering the transformer inoperable --or at best sharply reducing its operating lifetime [3]. Repairs in these cases pretty much entail retrofitting the whole damn thing, so the months-long disruption is all the same either way.

[1] http://www.ferc.gov/industries/electric/indus-act/reliabilit...

[2] http://fas.org/irp/agency/dod/jason/spaceweather.pdf

[3] http://www.lloyds.com/~/media/lloyds/reports/emerging%20risk...

My E&M is rusty but if I recall correctly, the electric field from the steam of charged particles will induce a voltage change in the transformer's wiring as the stream flows past/through the transformer, and as the transformer works by using a magnetic core to induce voltage changes from one coil to another coil, the voltage output is higher than the components on the output side can handle and the components get fried. I think.

What happens to everyone's net worth if power goes away for months or years? If I have, say, $1M in the bank (I don't), and the power returns after 2-3 years, do I still have money?

A big steel box full of batteries and solar panels kept mothballed in the basement sounds like it might be a good idea.

people interested in regular updated on solar flare can look into this youtube channel : https://www.youtube.com/channel/UCTiL1q9YbrVam5nP2xzFTWQ

Would the solar storm affect only the side of the Earth facing the Sun or the entire Earth?

Thanks in advance.

Could such an event be powerful enough to flip the Earth's polarity?

It was in 2012. Too close for comfort.

Warning: Click-bait title.

I absolutely hate these exaggerated headlines!!!

Bad title jamessun.

Some men just want to watch the world burn.

bullshit headline

I guess "destroyed life as we know it" is awfully literal. It means that "life as we know it" is "electricty costs 22 cents a kilowatt hour and GPS exists".

Yes, the power grid could have blown up. We would have then paid more for electricity. (It's unclear to me if the effect is global, or Earth shields the nighttime side from the impact.)

Satellites would probably stop working. So air traffic would have to use VOR/NDB/DME instead of GPS. Ships would use LORAN. Weather forecasters would use radar instead of satellite (a problem during hurricane season).

This headline makes me angry. It's not like we would have been instantly vaporized. Some stuff would be less convenient until we rebuilt it. That's it.

> We would have then paid more for electricity

With the current power grid structure in most developed countries, this would not be an option: some components (such as large transformers) in the current grid would be destroyed by a large CME event and it would take years to replace enough of them to restore the grid.

> Ships would use LORAN.

That would be difficult since LORAN doesn't exist any more; it was taken off line years ago when GPS became widely available.

> Some stuff would be less convenient until we rebuilt it.

As things stand currently, some stuff would not exist until we rebuilt it.

I certainly think we could get to a point where a large CME event would only be an inconvenience. But doing that will require redesigning significant parts of our infrastructure to be much more decentralized, and having plans in place to replace key components quickly. We are not at that point now.

I assume we would just use generators in the meantime. Yes, this is expensive. Very expensive.

It's not just expensive, it's impossible. There isn't enough fuel for all those generators, even if enough of them could be built.

The point is that the price will move to reduce demand to some level that's sustainable. That probably means critical infrastructure.

Which means most people will not be "paying more for electricity"; they will be without electricity, because generators are only affordable (which really means "available", since the steep price rise is just the market's way of rationing a very scarce resource, as you say) for critical infrastructure.

Yes, the power grid could have blown up. We would have then paid more for electricity.

Many people wouldn't be able to get electricity at all. Sure, you can buy portable generators, so having the power grid collapse isn't automatically a problem -- but there aren't enough portable generators in the world to power everything, even if we had the necessary quantities of fuel on hand.

So air traffic would have to use VOR/NDB/DME instead of GPS

Assuming VOR/NDB/DME stations had power and hadn't been set on fire by induced currents during the magnetic storm, sure...

To think of primary scenarios is a wrong way to think about it. We may not get blown to ash. But production of much of critical stuff[Food/Medicine/Sanitation] requires electricity these days. There is absolutely no way you are going mass produce and sustain food production for the level of population we have on earth, by using old stone age techniques. And this is first secondary effect. Food needs to transported, processed etc. How will all this happen without electricity?

Image what would happen if there is no electricity in the military. Which is such critical functions like that will get bulk of the available supply, leaving very little for everything else.

We may not go extinct, but given general shortage of food, struggle for resources and break down of law and order. The world will be a very different place.

It would be a disaster no doubt, but people would use generators and older vehicles to temporarily fill critical functions. There might be an economic collapse, but there would also be a lot of new jobs to rebuild stuff.

> but people would use generators and older vehicles to temporarily fill critical functions

Only if they remained calm and organized. When your (electricity-dependent) job is gone your first worry will be securing access to food and medicine. So depending on how much damage the incident does to power grid and how quickly governments react, you may see anything from short-term economical crisis to riots and collapse of the society.

The idea that rebuilding from disaster generates jobs doesn't require a mention of the Broken Window Fallacy. No one is creating the disaster with the aim be 'generating jobs' or attempting to argue that the disaster is a good things because it creates jobs.

I'm not sure how that's relevant.

The links for up voting and down voting are notoriously too small for mobile. For this post, I made sure to zoom to max so I could make 100% sure to upvote rather than risk the alternative.

"Life as we know it" is a valuable idiomatic phrase and we should protect it as such.

I think you're pretty woefully underestimating the consequences of subjecting our frighteningly fragile electrical grid to a CME of this magnitude.

I really need to get busy building that Faraday cage in the basement.

I know you are joking, but to take you seriously for a moment, a Faraday cage would not help at all.

A Faraday cage only shields against high frequency fields, but a CME is a very low frequency event. You need mu-metal, and huge amounts of it.

Plus it would knock out the power grid, so all the gear in your basement wouldn't be all that useful.

Can you explain the frequency dependence of the Faraday cage? The only time I can think of them being ineffective is against waves which are above the plasma frequency of the metal the cage is made of.

If the wavelength of the EMF is larger than the size of the entire cage it will go through the cage like it wasn't there.

(Although it would also have a hard time affecting the small items inside.)

A Faraday cage also can not shield against a static, or slowly changing magnetic field.

So basically the Mayans would have been right?

"21 December 2012" was the proposed date of the Mayan Apocalypse. So the end of the world would have been a few month early.

Should've used NTP or PTP.

The state of California spent $500 million on the bike path for the new eastern span of the SF bay bridge. And probably significantly less preparing for threats like this one.

Yes, that $500M could have been spent encasing every power line, transformer, and appliance in SF in 2 inches of lead.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact