The Carrington Event: not something to worry about (2020)
yarchive.net
yarchive.net
Here's the more advanced version.[2] This discusses the Carrington event, the Hydro Quebec event of March 13, 1989, and what PJM does when there's a problem.
The effect of a geomagnetic disturbance is that DC currents appear between distant points that are earth-grounded. This is caused by the ionized solar wind flowing through the space between power lines and the ground. Current flow through a loop generates voltage, remember. That puts DC currents into AC transformers, which can cause partial magnetic saturation, which makes the coils a resistive load for part of the cycle, which causes heating. Transformer damage is possible if power levels are not reduced. That happened to two transformers in 1989. The DC currents aren't huge; 10 amps is enough to cause trouble.
So, PJM now has DC ground current meters at a few key points. These report to the system control centers (they have two, just in case), where operators can take actions to reduce power flows at the vulnerable points. In [2], there's considerable detail about this. ("The procedure is implemented when a DC measurement of 10 amps or greater is detected at either Missouri Ave or Meadowbrook. Reduce Salem 1 & 2 units to 80% power, and Hope Creek to 85% power ...")
Journalists and pundits writing on this subject should read that before writing. If you don't understand it, find and take the "PJM 101" online training course. PJM's materials are written for the people who run the system, and there's little nonsense or hype.
The reports go out to anyone who gets PJM alerts. Those have been on the web for years. Since the last time I looked at this, they've added apps for IOS and Android. Since energy traders need this info, PJM lets anybody view it.
[1] https://insidelines.pjm.com/geomagnetic-disturbance-what-is-...
[2] https://pjm.com/-/media/training/nerc-certifications/trans-e...
I've seen physicists, power workers, electricians, congressional testimony, and lay loons on all sides over and over, and been convinced in opposing ways multiple times. It feels a little bit like reading about diets!
Outside of "become a physicist", anyone have any recommendations for cutting through the noise on topics like this?
A significant carrington event, in my opinion, would cause some surprising electronic failures, but not likely widespread ones.
The basis for my assumption (and that's all it is) is that previous failures of electronic systems because of solar flares have happened only on large scale electronics systems (power grids) and those systems were older, less shielded, and had bad tolerances.
Modern electronics, at least the bulk that are in daily use are much more shielded, higher quality, and resilient enough to not bring down civilization.
I'd expect it to be something more similar Y2K - definitely hit a few people, cost some money here and there, but literally everyone on the planet heard about it nonstop for quite some time because it might have been bad.
So, I think the tension around this is that, maybe depending on how astronomically large the event is, sure, it could be bad, like, if it stripped away our magnetosphere permanently somehow, yep, that'd suck and probably be a problem for humanity, but I find it to be on the unlikely end of things.
So folks aren't crazy for thinking it might cause some trouble, maybe even widespread, maybe even apocalyptic, but I think it's much more likely it will be an annoying and costly blip for most involved.
As an aside, I bet you'd be surprised with how often modern hardware has bit flips in memory because of cosmic rays alone, and we're mostly OK with that.
Electronics will largely be unaffected unless through bad luck of being connected without adequate protection to something else that suffers a surge (mains power or wired communications).
I guess I believe we can patch up what ever does break within weeks or months and our society (the last year highlights) can survive in very extreme conditions. Temporary loss of technological comforts isn’t so big, and I believe, temporary.
So many phenomena relating to the pandemic have been disruptive more because of synchronization of normally random things.
Does the same shielding required for consumer electronics to meet FCC Class B requirements also protect them from solar events?
In 1859 the effect was to generate high enough dc voltages to cause shocks, and ruin equipment, likely due to insulation breakdown. Modern lightning protection would handle that, were long telegraph wires in use today.
The thing everyone worries is transformers. If a transformer has no power on it, the currents available would barely cause a measurable change in temperature. (lets say 100 amps across a 1 volt drop, 100 watts of heating)
Transformers are designed to maximally use the available flux handling capacity of their core. As the flux rises, eventually a saturation point is reached... at this point, the resistance to further current increase in the coil (the inductance) drops rapidly, leading to a surge in current at the top of the AC cycles, which then leads to rapid heating. As commented above, [https://news.ycombinator.com/item?id=27238108] this can then lead to damage if the saturation isn't detected.
If a Carrington event were detected in advance (we have about 12 hours warning with the satellites in the Lagrange point between earth and the sun, as the solar wind travels less than the speed of light), the prudent thing to do is turn off the grid, the second best option is to watch everything very carefully for the duration, and proactively turn off the circuits with overheating transformers.
Given the well known long term consequences of letting transformers saturate, I think prudent grid operators won't have a problem.
I was about to say the same thing about diets, before I even got to when you said:
> It feels a little bit like reading about diets!
It makes me wonder why something like that doesn't have a definitive answer, or something to point out which side (or both) is wrong, but they seem to be immune to inspection. Although, I did read a book called Diet Cults (written by just a journalist though) which contradicted all the extreme ones and suggested a very moderate approach.
I don’t know the explanations for why we hear such conflicting confident answers, but this is my suspicion:
I imagine in both cases the majority of experts in the field would provide somewhat nuanced advice, couched in cautions about how much about these systems they don’t know or fully understand. None of those experts break through.
What does break through is all you end up hearing from, which is a smaller minority on each side willing to make bold and confident predictions.
Reality is extremely complex, with many interfacing chaotic systems. Why would we expect a single answer to work in every instance?
What if there isn't a perfect diet for everyone, but each person can find the one that works best for them?
I don't think it's surprising that there isn't a definitive answer to what the effects of a carrington event might be. The analysis required to provide such an answer would have such vast complexity that it's likely not worth the effort.
All we have is our best guesses, based on informed and deliberate inquiry.
I don't think we can reasonably expect anything more than that.
Most "news" stories of Carrington Event-type catastrophes are not reasonable but sensationalist. Pure speculation to get ad revenue. Comments on reddit or various internet fora are about as good as Amazon reviews. So your first order of business, save becoming an expert yourself, is to discard all commentary on Carrington events not produced by those with sufficient expertise to illuminate beyond speculation - electrical engineers, space weather physicists, etc. That will eliminate 99% of page hits on Google, probably.
From what remains you'll find a much smaller range of answers. Usually much more boring.
And reading further down, the article claims the change in magnetic field is slow. It's not. It's very fast.
The article does not get the basic physics correct. The one thing that is correct is that it is unlikely to damage your cell phone. What will happen is that massive currents will be induced in transmission lines blowing up transformers and possibly damaging anything plugged into mains.
The average speed of a CME is 500 km/s. I think they've been seen to go as fast as 3000 km/s. At 3000 km/s, that would take about 4 seconds to traverse the diameter of the Earth, and much longer to traverse the magnetosphere.
I would call that slow, especially compared to the timescale of the EMP from a nuclear weapon.
But no, it is not like an EMP. That is the wrong way to think about it.
There have been historical political efforts to have off line back up transformers in place ready in the event of an EMP. also fuses built in to protect the transformers from overload. These have not got as far as legislation.
The toilet paper was obviously going to be very temporary, the total demand for the year would be unchanged. It also didn't cause significant problems. Very few people will die from lack of toilet paper, so governments aren't going to give out emergency subsidies. That makes it very uneconomical to scale production up.
The power grid being down is an absolute disaster in most highly developed nations that depend on power, so warlike resources would be dedicated to addressing the problem. However, what I often don't see considered: Lead times are months in a country where everything works. If the transformer factory is without power, without a phone to reach the suppliers, and the suppliers are also without power, it could be significantly longer, and that ignores the general collapse of society that would happen without a power grid.
Anyways, just transport from factory to destination usually takes weeks. Even if only a few 100 kilometers.
I am curious what sort of fuse you think would prevent EMP-like damage though.
My understanding is that an emp introduces electrical currents in the lines. Protecting the transformer from electrical overloads should protect it from emp. No?
The effects of EMPs vs geomagnetics are very different.
This sort of ignorance is just sad.
No, there is no way "that production could be sped up considerably if need be". None.
And lead times are actually counted in years.
We have about two dozen schnabel cars in the entire USA and they can only roll about 10 MPH.
Luckily they only roll at 10 MPH or so, because RR personnel have to do crazy stuff to move oversize loads like remove and reinstall traffic lights and occasionally, level crossing "arms". Sometimes they have to cut down and replace telephone poles and crazy stuff like that.
Logistics has friction and obviously if your entire national grid is down you can't haul using electric RR engines but USA is mostly diesel anyway. Another friction is you need 10 cars for a local event in Florida but they're all over the country and the bearings can't roll faster than 10 MPH and it takes a long time to cross the USA at 10 MPH as the pioneers learned. Another friction is they actually use those cars for stuff other than carrington event transformer replacements, so taking all those cars for replacements for a decade or so means no nuclear reactor transport no steel mill roller transport etc.
Of course "in reality" it would be easier to massively decentralize large transformer production and ship the personnel and raw copper on 747s to the site rather than now where we centralize production and then ship out. Of course we have no experience doing that, but its more likely to succeed than trying to ship via rail.
In practice I think the problem with replacing a large number of large transformers is we MIGHT be able to do it under ideal conditions including full electrical power and normal levels of law and order, which we will not have after a carrington event.
But for a specific substantial claim, even if we could 3-d print or star trek replicator large transformers on demand, or even if we had secret government warehouses full of them in Nevada, it would take months to years to transport all of them to specific sites. SOME work like clearing rights-of-way to pass the equipment could be done ahead of time. We have SOME ability at heavy mfgr plants like our world leading mining equipment plants to build new schnabel cars with a lead time of maybe months each. But it would be rough.
Anyway you wanted a specific substantiated claim so here it is, we can't logistically ship a large number of large transformers.
https://www.gao.gov/assets/gao-19-98.pdf
my point wasn't about 'speeding up prduction' it was about previous political efforts to legislate to have transformers offline ready to replace blown ones, GIC resistors and most importantly early warning tech to detect incoming solar flares and take transformers offline to protect them.
> most importantly early warning tech to detect incoming solar flares and take transformers offline to protect them
This is politically impossible. If one takes the entire grid offline, people suffer and lash out. If the grid is instead taken out by a solar flare, aliens, you name it - it's much easier to shift the blame.
'The fastest Earth-directed CMEs can reach our planet in as little as 15-18 hours. Slower CMEs can take several days to arrive. They expand in size as they propagate away from the Sun and larger CMEs can reach a size comprising nearly a quarter of the space between Earth and the Sun by the time it reaches our planet'. https://www.swpc.noaa.gov/phenomena/coronal-mass-ejections
http://docs.house.gov/meetings/SY/SY18/20180426/108217/HHRG-...
> Well, if you have an electrical loop whose size is measured in miles, or (better) hundreds of miles, you can pick up enough current to matter. The telegraph stations were like that.
There's new research coming out in a few months on this question -- how such an event may take out transoceanic Internet connectivity:
https://mobile.twitter.com/sangeetha_a_j/status/138930329505...
I don't know why such an event would affect only part of one continent, unless (say) the Earth's field touched down only there and in unpopulated areas.
Anytime we look back into history to try to find an event that wasn't noticed at the time we're going to be dealing with spotty records. It's hard enough finding evidence of tsunamis, and this is much less noticeable.