I don't really know the physics of this, but it seems CMEs only really effect power supply, not electronics in general. So we are taking massive damage to power infrastructure, but that can be repaired, and there are already mitigation plans in place.
No, I think Avalaxy is confusing CMEs and EMPs (he's concerned about tractors, trucks, sundry battery-powered computers, and hugely inductive industrial machinery) and he's extrapolating wildly from his own inability to think outside the box (e.g. how checks / offline card processing work, how computer experts might be useful in a world where a good chunk of computing infrastructure suddenly gets fried, etc).
A big CME could be hella destructive and having a backup stash of supplies is a fine idea. That $2T damage estimate doesn't sound crazy if we assume that the heads-up from astronomers gets ignored and real-time distribution management isn't capable of shutting things down in time (both possible AFAIK). But even in that scenario...
I'm confused how an EMP and a CME differ. Aren't they both magnetic disturbances that affect electronics due to the fact that electromagnetism is one and the same?
Perhaps there's a difference in intensity and duration.
Wikipedia indicates[0] that an EMP could generate 50kV/m electric fields lasting ~1μs, a second component similar to lightning lasting ~1s (electric field intensity not listed), and a third, long-term component that lasts hundreds of seconds.
CMEs, it appears, only generate the third, long-lasting type of electric disturbance. According to NASA[1], these have electric fields up to 26V/km, which is a much lower intensity, but for a significantly longer duration.
[someone please check my numbers; they seem wrong but I'm posting them anyway; my guess is misinterpreting electric field potential as actual induced voltage]
To estimate how much power would be dissipated in transmission lines, let's consider a disconnected 1km wire with a DC resistance of 30 milliohms per kilometer (based on [2]) and ignore all other losses. The so-called E1 pulse at 50kV/m would generate a voltage of 50MV across the wire. Let's assume that lasts 200ns. Power is V^2/R, or (50MV)^2/30mohms=83.33petawatts [3].
That sounds enormous, but it's for a short time. Energy is Pt, or 83.33PW200ns=16.7GJ. To find out how much the wire heats up, let's assume a perfectly straight copper conductor 2cm in diameter. That's a volume of 1kmpi(1cm)^2=1.257m^3 or a mass of 2815kg[4]. Taking energy/(specific heat * mass), we get a temperature rise of 15410K[5]. Copper melts at 1357K, so if the math is right that transmission line would be toast. These numbers seem way too large.
An E3-type pulse from a CME lasting 1000 seconds would generate up to 26kV across the length of wire. That's (26kV)^2/30mohms=22.5GW. Over 1000 seconds that generates 22.5terajoules[6]. I'm not even going to bother calculating a temperature rise for that (okay, it's 20.8 million K[7]). Again, this seems several orders of magnitude off.
I've ignored capacitance, inductance, increased resistance with temperature (up to 170milliohms if I used [8] correctly), and no doubt a lot of other things that would significantly affect the current flow, so keep in mind that these numbers are probably way off. I've also ignored heat dissipation to the air, which would mitigate the energy input a bit but also keep the resistance lower. I welcome any corrections.
Assuming these numbers are completely bogus, it seems to me that an EMP would be more damaging to small electronics, while the CME would be more dangerous for transmission infrastructure.
Claims a 1km copper wire with 10mm has a resistance of ~0.2Ohms. But I don't think it's ok to use normal resistance for pulses that last on the order of microseconds. I imagine capacitance and inductance play a very significant role.
> CMEs, it appears, only generate the third, long-lasting type of electric disturbance. According to NASA[1], these have electric fields up to 26V/km, which is a much lower intensity, but for a significantly longer duration.
And on a much bigger scale distance wise. An EMP would not engulf half the planet in one go.
It looks like nitrogen (the user, not the gas) has saved me the trouble of doing the math, but he (she?) has omitted an explanation of the physics so I'll leave one here. Sorry if you already know it, but if you already know it and didn't apply it that's your own damn fault :P
Google keyword: Faraday's Law of Induction
Kirchoff's Voltage Law (sum of voltages around a loop is zero) breaks when there's a changing magnetic field poking through the loop. The amount it breaks by is V=A*dB/dt where V is "extra" voltage, A is area of the loop, and dB/dt is change in magnetic field with time.
An EMP creates a shockwave of enormous dB/dt. Even loops with small A can induct enough voltage to fry sensitive electronics.
A CME has dB/dT of 5mT/min (linked article). Fridge magnets are ~5mT (wikipedia) so you could make a local field with similar dB/dt by walking very slowly towards your computer while carrying a fridge magnet. Of course, unlike fridge magnet vs computer, CMEs are damaging because A is massive (distribution infrastructure is a big loop) so even though dB/dt is tiny they can fry power lines.
That is, unless real-time distribution management notices voltages starting to go haywire and shuts them down because a CME isn't instantaneous and hard to anticipate like an EMP. Or circuit breakers do their job because voltage / km is relatively small. Or inductive loads (anything with a motor or transformer in the circuit) protect themselves from what essentially looks like just another voltage spike (they happen all the time because every time you shut off a poorly designed inductive load it kicks a high voltage back onto the line). Or TVS diodes do the same for sensitive electronics somehow connected without a transformer...
tl;dr a CME is not like an EMP. Same physics equation but with very different numbers.
I'm probably missing something?