4 ms·

Turning the power off would work, but is that feasible? Even if we’d be able to spot the CME, would then have enough time left to shut down the entire power gri

Turning the power off would work, but is that feasible? Even if we’d be able to spot the CME, would then have enough time left to shut down the entire power grid? The time between detection and the CME hitting us would probably be measured in minutes. I don’t think it’s possible to shut down the grid in that timeframe.

javajosh 3y ago
I say this only half in jest, but could we not request that an adversary trigger their malware to shut down our grid? When life gives you lemons...
- yifanl 3y ago
  Wouldn't such an adversary always choose the option that causes us more damage?
digging 3y ago
Yeah I question if it's something we even could do. But even if is, it's going to kill and hurt a lot of people. Probably less than setting every electrical device on (half of) the planet on fire or whatever, but good luck convincing people of that when their dad is on dialysis or all their food spoils and they can't get to the store.
onenukecourse 3y ago
We can, and have, shut off large portions of the grid in seconds. Take the 2003 blackout. Yes, the whole shut down took 15 minutes (?). But thats because it was a cascading effect that had to travel down the lines. Once the fault was detected by a particular segment of the grid, the relays responded in milliseconds. They have since the 1920s? Add in an "incoming solar flare" fault condition and we can trip the whole grid in seconds and send a start signal to the diesel generators to warm up to bring her back up. Pretty nifty trick. Question is why would we? The grid has been undergoing a lot of strengthening against EMPs and flares for decades. Its not obvious to me that a flare can take it out, especially if we shed dumb loads (partial blackout, say data centers) before it hits to give the conductors and transformers head space.
- spdustin 3y ago
  If we had done enough to mitigate EMPs, the nuclear powers of the world wouldn't have space-based nuclear EMPs as the first step of their attack plan. We still do, and so does Russia. Geomagnetically-induced current is different from the plain-vanilla EMPs anyway—GIC can last for hours.
  - roenxi 3y ago
    We might easily be unprepared, but that the military tries things that might not work. Military attack is all about trying things that might cripple the enemy and/or increase the cost of an effective defence. So an EMP isn't necessarily because it is expected to do horrific damage. It is just part of a thorough test of an adversaries preparations, making it harder to protect their infrastructure.
  - AnarchismIsCool 3y ago
    I fail to understand why we don't do more to make equipment robust to this kind of thing. There's a whole range of problems that this solves looong before we get to general nuclear exchange. If stuff was shielded, isolated, and grounded better, everything from your phone to your WiFi would work a lot better and have longer range. Wind slapping power lines together wouldn't destroy everything plugged in inside your house and solar flares wouldn't be more than a passing concern. The design changes to affect all of this aren't remotely expensive or difficult, we just don't.
    - ianburrell 3y ago
      There is no need to shield electronics. The induced currents only cause damage to long conductors, to the electrical grid and to long fiber optic cables.
      - AnarchismIsCool 3y ago
        That's a complicated one, it's still the electronics but they need a particular circuit that can stop short rise time transients. Also, any larger devices probably want to have shielding and ethernet devices will need some extra hardening, we're talking 25+kv/m events here so unless your computer can handle the monitor being at a 25kv differential from the tower you're gonna have a bad time.
btilly 3y ago
It is feasible. In https://en.wikipedia.org/wiki/Carrington_Event https://en.wikipedia.org/wiki/Carrington_Event it is estimated that the time from spotting the flare to the solar storm was 17.6 hours. That's plenty of time. The problem is figuring out how big the event is, and how directly it will hit us. So while we have over 17 hours to prepare, there might be some false positives due to our limited prediction skills. And, no matter the real consequences, people have limited patience for large economic disruptions over things that turned out to be nothing.
- kurthr 3y ago
  Even better, if you turn off the electricity and prevent major destruction and nothing happens (other than the power down/up) then you're the one who caused "the problem". There's no reward for fixing a problem that doesn't happen and that people don't want to believe even exists. Bonus, if other networks are damaged while yours aren't, it must be because you protected your network so you're responsible!
- bruce511 3y ago
  17 hours sounds like a long time, but it's not like there's a big red button and a guy standing by to press it. So the information chain, starting with the telescope that detects the flare, and then has to work it's way up the food chain, so that sufficient people agree, and take presumably synchronised action, well, good luck with that. You'd also ideally need a multi-hour warning, planes gotta land etc. Make no mistake, shutting it down will result in some deaths [1]. And those deaths will be on the news tomorrow (if indeed news still exists.) On the other hand not shutting down will cause more deaths and massive destruction. [1] think hospitals where the backup power failed, or didn't last long enough. Traffic intersections. Elevators. Water pumps. Airplanes. Trains. Take your pick.
tejtm 3y ago
CME are "Coronal Mass Ejection" with mass being the operative word here. Electromagnetic radiation (electrons photons) can make the trip between sun and earth in about 8 minute but anything with neutrons or protons (such as the coronal plasma) takes much longer as in, a day and a half to several days. CME are not hard to spot leaving the sun with even with tiny amateur telescopes (the sun does not require much in the way of light gathering) so even without the professional scopes (SOHO) with dedicated satellites leading and trailing earth orbit constantly viewing around the edge of the sun as seen from earth or being able to acoustically "hear" (if you can call 5 minute pressure waves sound) the far side of the sun, it is not conceivable to me we would not be warned a CME was incoming, doing something about it is another story.
ianburrell 3y ago
It takes 15-24 hours for CME to arrive at the Earth after solar flare. The particles are much slower than the radiation which arrive immediately. We have pretty good prediction if CME will hit Earth.
- dylan604 3y ago
  Does ~8 minutes qualify as immediately, or are you doing the physicists thing of rounding to 0?
  - ern 3y ago
    For practical purposes in this case it’s immediate right?
    - dylan604 3y ago
      I can do a lot in 8 mins. If you had 8 minute warning for an earthquake, what could you do? Luckily, we get more notice than that now for tornadoes, but 8 minutes is enough time to seek shelter. In 8 minutes, there's plenty of time to ctrl-s on everything, and then close apps and shut down computers. The problem is communicating to everyone when that 8 minutes starts and how much time is left.
      - jakeinspace 3y ago
        But the point is that we’ll have 15-24 hours, not 8 minutes.
        dylan604 3y ago
        That may be your point, but not the point to which this thread started.
      - thfuran 3y ago
        >If you had 8 minute warning for an earthquake, what could you do That's not how it works. There may be detectable precursors that could actually give warning, but the 8 minutes referenced is the time it takes light from the sun to reach earth. It's immediate in the sense that it is physically impossible to detect that before those 8 minutes have already elapsed and the light is hitting your detectors. You could try to move your detectors closer to the sun to detect earlier, but any signal you can possibly send back to earth goes at the same speed, so it doesn't help.

javajosh 3y ago

I say this only half in jest, but could we not request that an adversary trigger their malware to shut down our grid? When life gives you lemons...

yifanl 3y ago
Wouldn't such an adversary always choose the option that causes us more damage?

digging 3y ago

Yeah I question if it's something we even could do. But even if is, it's going to kill and hurt a lot of people. Probably less than setting every electrical device on (half of) the planet on fire or whatever, but good luck convincing people of that when their dad is on dialysis or all their food spoils and they can't get to the store.

onenukecourse 3y ago

We can, and have, shut off large portions of the grid in seconds. Take the 2003 blackout. Yes, the whole shut down took 15 minutes (?). But thats because it was a cascading effect that had to travel down the lines. Once the fault was detected by a particular segment of the grid, the relays responded in milliseconds. They have since the 1920s? Add in an "incoming solar flare" fault condition and we can trip the whole grid in seconds and send a start signal to the diesel generators to warm up to bring her back up. Pretty nifty trick. Question is why would we? The grid has been undergoing a lot of strengthening against EMPs and flares for decades. Its not obvious to me that a flare can take it out, especially if we shed dumb loads (partial blackout, say data centers) before it hits to give the conductors and transformers head space.

spdustin 3y ago
If we had done enough to mitigate EMPs, the nuclear powers of the world wouldn't have space-based nuclear EMPs as the first step of their attack plan. We still do, and so does Russia. Geomagnetically-induced current is different from the plain-vanilla EMPs anyway—GIC can last for hours.
- roenxi 3y ago
  We might easily be unprepared, but that the military tries things that might not work. Military attack is all about trying things that might cripple the enemy and/or increase the cost of an effective defence. So an EMP isn't necessarily because it is expected to do horrific damage. It is just part of a thorough test of an adversaries preparations, making it harder to protect their infrastructure.
- AnarchismIsCool 3y ago
  I fail to understand why we don't do more to make equipment robust to this kind of thing. There's a whole range of problems that this solves looong before we get to general nuclear exchange. If stuff was shielded, isolated, and grounded better, everything from your phone to your WiFi would work a lot better and have longer range. Wind slapping power lines together wouldn't destroy everything plugged in inside your house and solar flares wouldn't be more than a passing concern. The design changes to affect all of this aren't remotely expensive or difficult, we just don't.
  - ianburrell 3y ago
    There is no need to shield electronics. The induced currents only cause damage to long conductors, to the electrical grid and to long fiber optic cables.
    - AnarchismIsCool 3y ago
      That's a complicated one, it's still the electronics but they need a particular circuit that can stop short rise time transients. Also, any larger devices probably want to have shielding and ethernet devices will need some extra hardening, we're talking 25+kv/m events here so unless your computer can handle the monitor being at a 25kv differential from the tower you're gonna have a bad time.

btilly 3y ago

It is feasible. In https://en.wikipedia.org/wiki/Carrington_Event https://en.wikipedia.org/wiki/Carrington_Event it is estimated that the time from spotting the flare to the solar storm was 17.6 hours. That's plenty of time. The problem is figuring out how big the event is, and how directly it will hit us. So while we have over 17 hours to prepare, there might be some false positives due to our limited prediction skills. And, no matter the real consequences, people have limited patience for large economic disruptions over things that turned out to be nothing.

kurthr 3y ago
Even better, if you turn off the electricity and prevent major destruction and nothing happens (other than the power down/up) then you're the one who caused "the problem". There's no reward for fixing a problem that doesn't happen and that people don't want to believe even exists. Bonus, if other networks are damaged while yours aren't, it must be because you protected your network so you're responsible!
bruce511 3y ago
17 hours sounds like a long time, but it's not like there's a big red button and a guy standing by to press it. So the information chain, starting with the telescope that detects the flare, and then has to work it's way up the food chain, so that sufficient people agree, and take presumably synchronised action, well, good luck with that. You'd also ideally need a multi-hour warning, planes gotta land etc. Make no mistake, shutting it down will result in some deaths [1]. And those deaths will be on the news tomorrow (if indeed news still exists.) On the other hand not shutting down will cause more deaths and massive destruction. [1] think hospitals where the backup power failed, or didn't last long enough. Traffic intersections. Elevators. Water pumps. Airplanes. Trains. Take your pick.

tejtm 3y ago

CME are "Coronal Mass Ejection" with mass being the operative word here. Electromagnetic radiation (electrons photons) can make the trip between sun and earth in about 8 minute but anything with neutrons or protons (such as the coronal plasma) takes much longer as in, a day and a half to several days. CME are not hard to spot leaving the sun with even with tiny amateur telescopes (the sun does not require much in the way of light gathering) so even without the professional scopes (SOHO) with dedicated satellites leading and trailing earth orbit constantly viewing around the edge of the sun as seen from earth or being able to acoustically "hear" (if you can call 5 minute pressure waves sound) the far side of the sun, it is not conceivable to me we would not be warned a CME was incoming, doing something about it is another story.

ianburrell 3y ago

It takes 15-24 hours for CME to arrive at the Earth after solar flare. The particles are much slower than the radiation which arrive immediately. We have pretty good prediction if CME will hit Earth.

dylan604 3y ago
Does ~8 minutes qualify as immediately, or are you doing the physicists thing of rounding to 0?
- ern 3y ago
  For practical purposes in this case it’s immediate right?
  - dylan604 3y ago
    I can do a lot in 8 mins. If you had 8 minute warning for an earthquake, what could you do? Luckily, we get more notice than that now for tornadoes, but 8 minutes is enough time to seek shelter. In 8 minutes, there's plenty of time to ctrl-s on everything, and then close apps and shut down computers. The problem is communicating to everyone when that 8 minutes starts and how much time is left.
    - jakeinspace 3y ago
      But the point is that we’ll have 15-24 hours, not 8 minutes.
      - dylan604 3y ago
        That may be your point, but not the point to which this thread started.
    - thfuran 3y ago
      >If you had 8 minute warning for an earthquake, what could you do That's not how it works. There may be detectable precursors that could actually give warning, but the 8 minutes referenced is the time it takes light from the sun to reach earth. It's immediate in the sense that it is physically impossible to detect that before those 8 minutes have already elapsed and the light is hitting your detectors. You could try to move your detectors closer to the sun to detect earlier, but any signal you can possibly send back to earth goes at the same speed, so it doesn't help.