Slashdot Mirror
The Truth About Solar Storms
StartsWithABang (3485481) writes On Wednesday, The Washington Post ran a story about a very large solar flare two years ago that missed Earth, but not by too much. From a scientific point of view, what is it that happens when a solar flare interacts with Earth, and what are the potential dangers to both humans and humanities infrastructure? A very good overview, complete with what you can do — as both an individual and a power company — to minimize the risk and the damage when the big one comes. Unlike asteroids, these events happen every few centuries, and in our age of electronics, would now create a legitimate disaster.
If the humanities infrastructure suffers, no doubt there'll be fewer English majors, and more CS majors, so it'll be a good thing, right?
Or did someone mean "humanity's infrastructure"? Yes, I know, "my people don't do editing"....
"I do not agree with what you say, but I will defend to the death your right to say it"
I saw a great documentary on this by Lucasfilms called "Howard the Duck" and I am prepared. Sure, they phonied it up a little bit, but the basics work for solar storms too. My Quackfu is second to no man!
Time Bomber the Book coming soon.
High induced votlages in open wires are a problem, but they're not the big one.
The biggie is common-mode currents in long high-voltage transmission lines adding a strong DC component to the current in the substation transformer windings - high enough that when the same-direction peak of the AC's cycle adds to it, the core saturates. Then the inductance of the transformer drops to the air-core value and no longer substantially impeeds the current.
The current skyrockets. The resistive heating of the windings (and the force on the wires from the magnetic fields) goes up with the SQUARE of the current. The windings quickly soften, distort, form shorted turns, melt, open, short out to the frame, etc. The transformer is destroyed, or committed to a self-destructive progressive failure, in just a handful of such cycles - too fast for the circuit breakers to save them (even if they DO manage to extinguish the arcs with the substantial DC component to the current.) Even if the transformer doesn't explode and throw molten metal, gigawatt sustained arcs, and burning oil (or burning-hot oil replacement) all over the substation area, it's still dead.
This happens to MANY of the giant transformers in the power grid. Each set of three transformers that has one or more failed members means a high-voltage transmission line that is shut down until the transformer is replaced.
There are essentially no spares - these are built to order. Building one takes weeks, and there are few "production lines" so little parallelism is available. What is destroyed overnight will take years to replace, while each intercity power transmission line is not functioning until the transformers at its end ARE replaced.
The current occurs because the transformers are organized in a "Y" arrangement, and the center of the Y is grounded at each end (to prevent OTHER problems). The transformers have enough extra current handling capacity to avoid saturation from the DC through that center connection to/from ground from ordinary electrical and solar storms - just not a giant one like we get every couple centuries.
The solution is to put a resistor in that ground connection, to limit the DC in the lines (and dissipate the energy it represents). Indeed, a few lines have such resistors already.
But a suitable resistor is a box about the size of one of the transformers. It's very expensive. And it only makes a substantial difference to the operation of the lines in such a once-in-centuries event. So most executives don't spend the money (and get dinged for costing the company millions) to put them in, to prevent a failure mode that hasn't happened in the generations since Tesla and Westinghouse invented the three-phase long-line power grid.
Or at least they don't until the regulators or their stockholders require it. Which means said decision-makers need a little educational push to decide it's worth the cost and get it done.
Thus articles like this. B-)
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
... the induced DC from a solar storm isn't as instantaneous as a lightning strike. It takes minutes to develop, which leaves time to disconnect the lines and affected transformers if they are properly monitored.
But ARE they monitored for DC? It's not a usual problem.
Warnings on the order of minutes might be useful if the transmission line were the only one invoved. Unfortunately, the power grid is a GRID. Lots of multiple, parallel, transmission lines, and many, many, more going elsewhere and often creating loops.
Redundancy is a good thing in most situations. But when you have to drop a high line, and don't drop all the others simultaneously, you shift the load onto those that are still connected. When you're cutting off because you're near the limit - either due to heavy load at the time or because of the DC issue - you can drive the others beyond their limits (or throw things out of sync and add a bunch of "reactive current" to the load) and create a cascading failure. (Indeed, this is how the first Great Northeast Blackout occurred: Three of a set of four high-lines crossing the St. Lawrence Seaway near Niagra tripped out, and the redistributed load put one after another generator above its limits, blowing its protective breakers and making it progressively harder on those remaining.)
Gracefully shutting down the grid is not something you do on a couple minutes' notice, even if you have a plan in place.
As I understand, the induced DC is something on the order of hundreds of volts, which is much less than the tens of thousands of volts transmitted across ordinary high voltage transmission lines; disconnecting them should not result in arcing problems across the switches.
First, the problem with the induced near-DC is not the voltage, but the current. Transformers and transmission lines have as little resistance as possible, because it's pure loss of valuable energy. The magnetizing alternating current (i.e. the part of the AC that's there all the time, not just when there's a load) is also limited by the inductance of the transformers, but that doesn't impede the direct current at all. A couple hundred "DC" (very low frequency - fractional cycle per minute) volts, induced for minutes around the loop, can drive a hysterical amount of current.
Once the transformer is saturated, most of the damage comes, not from the direct current, but from the line power, which ends up dissipating lots of energy in the transformer. Meanwhile, at these voltages and currents, the switches that interrupt the AC are largely dependent on the momentary off time as the cycle reverses to quench the arc. If, say, the event happened when the line was running at about half its rated load, the direct current will be higher than the alternating current, so there will be no off time. This can keep the current flowing even through an open breaker (while dissipating megawats IN the breaker). Interrupting DC is MUCH harder than interrupting AC.
Heck, at these voltages even interrupting AC is hard. (The video is of an interrupter where the jet of arc-suppressing gas failed for one leg.)
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way