Anyone who’s had to build a power system rapidly learns that electricity is not as simple as “electrons move, work gets done”. Real electrical systems have to deal with issues of reactance and other exciting math-heavy constructs designed to drive you into some other field of study.
Power Grids experience this on an epic scale. They have to concern themselves with a few needs simultaneously:
- ensuring electrical potential doesn’t sag under load (maintaining voltage)
- ensuring the integrity of the AC waveform (maintaining frequency)
- ensuring the system doesn’t lose too much energy to fighting its own electromagnetic behavior (controlling the power factor)
That last one is the part that is profoundly nonintuitive. Capacitance and inductance inherent to the system create a sort of inertia in the system that must be fought to provide those other two guarantees. Together they work to create what’s called “reactance”.
Long range lines and the equipment they connect to can have a lot of reactance. High voltages make it even weirder.
One of the strange things that you don’t experience at lower voltages is “corona discharge”. Very high electrical potentials cause the air around the conductors to become ionized. When sufficiently ionized, this creates discharges.
You can see static examples of this natural phenomenon in the form of “Saint Elmo’s Fire”. This often precedes a lightning strike if the potential difference is extreme enough.
But power transmission systems are not static. They fluctuate dynamically with the AC waveform. This causes situations where discharges or perturbation in fields that create them act as a new component of the reactance of the system.
Modeling this is very complicated and very important. Real power transmission systems have active components that work to provide the above guarantees. Most of the time, they are modeled well and tune things to keep the voltage and frequency where it should be with a minimum of losses due to reactance.
But these corona discharges are unlike the other two components of reactance, because they are affected by the environment outside of the system. Consider… what affects the voltages at which ionized air molecules might cause a dishcarge? Temperature and humidity do.
And now we finally reach the part where “induced atomospheric vibrations” starts to make sense. When things get hotter or drier, discharges happen more readily. And when they happen, they tend to happen at a given frequency.
When that happens, the active components of the transmission system try to reduce the reactance; but they can’t. They aren’t built to deal with this kind of reactance and the models that drive them make incorrect assumptions (e.g. some increase in capacitance means we need to adjust the phase angle).
This would work fine in a system without these discharges. With the discharges, it’s ineffective. And, since both ends of the connection may be actively doing things that make the problem worse, they can get themselves so far out of sync that they’re basically burning power competing with each other (instead of working together like normal).
This is what took down the grid. Unexpected components of reactance appeared due to corona discharges because of atmospheric conditions. From there, automatic safeguards designed to keep the grid in sync were ineffective or even counterproductive. The grid decohered and everything shut down as safeguards tripped in a cascade.
As the grid gets more modern, it also exacerbates these problems. When big spinning generators were driving the grid, they could only act so fast. This created a kind of inertia that damped oscillations created by responses to these issues.
With inverter-based power storage, generation, and transmission, the grid can now react incredibly quickly. This is good when they do the right thing, but can be very bad when they do the wrong thing.
As the weather grows hotter, the grid also sees more of these previously unusual conditions, too. Today they formed a kind of “perfect storm”. In the future, we will likely see this more often until they find ways to counteract, mitigate, or model these issues.
You can read more about this phenomenon here: https://electrical-engineering-portal.com/7-bad-effects-corona-transmission-lines
English (US) ·