That's more or less how it works. Knowing what 50hz is and how out of sync a generator is is easy. AFAIK, the problem is one of inertia, both electrical and mechanical. A generator is a spinning mass, the more load on the generator, then the slower it turns. How fast it turns decides the frequency of power it generates. Say you get your generators running at 50hz at night, and then during the day it gets hot and thousands of AC units get turned on. Now your generators are running at 40hz.
It's not about getting the generator to spin at 50hz, it's about getting it to spin at 50hz under dynamic load. Most generators are mostly fixed speed I think for the sake of efficiency, so the mechanism the grid has to regulate the frequency of power is largely by controlling the number of generators running at any given time, and how much of the grid is connected to them. Ramping up and down based on the current observed frequency of the grid is exactly how this all works. However it's more complicated because what if two or more power stations see a low frequency grid and decide to ramp up more generators? Well now the grid frequency is too high. If they them respond to that by ramping down, it'll go too low. Rinse and repeat and you get undesirable frequency oscillation, so there needs to be more communication across the grid than just responding to the locally observed frequency.
Connect a stopped, or sufficiently out of phase generator to a full scale power grid and the puny thousands of pounds of metal that make up a generator are going to get ripped apart by the sheer inertia of the power grid forcing it to match speed and phase in an instant. This is also a problem when attempting to connect two isolated grids, of they're too out of phase with each other, catastrophic physical damage will occur.
Thanks for the reply. In the 50Mhz-modulated-50Hz scenario, however, the modulated 50Hz signal doesn't change under load; it has zero inertia (or infinite inertia maybe) because it's electronically generated.
Any individual station that gets out of sync either works hard to catch up, or gets disconnected.
The problem, as I see it, is that the load signal is equal and equivalent to the sync signal and maybe they could be separated. (e.g. the FM modulated signal wouldn't have to be a sine wave, nor would it even have to be 50Hz)
> what if two or more power stations see a low frequency grid and decide to ramp up more generators?
Seems to me it would solve this problem - they both have a single invariant sync signal.
> Connect a stopped, or sufficiently out of phase generator to a full scale power grid and the puny thousands of pounds of metal that make up a generator are going to get ripped apart
Seems to me this problem is solvable with electronics, but I don't know enough to say how much out-of-phase is problematic. What if, as you say, some heavy load is switched on, the local nuclear turbines slow by a fraction for a few 10s of seconds, and so a home PV array starts leading by 0.01Hz (e.g. home PV at 50.00Hz, local power station at 49.99Hz, so after 10s the PV is fully 36ΒΊ or 0.6rad ahead). Does this result in the inverter (or the small local windfarm) exploding? Is that a realistic scenario?
> there needs to be more communication across the grid than just responding to the locally observed frequency
Precisely - a grid-wide sync signal, modulated at high frequency.
You need to consider that a master clock signal transmitted through air will arrive at a different time than an AC power wave transmitted through a wire.
A millisecond latency will cause a significant out of phase shift to the AC signal too.
The AC data you want is already codified in the power signal
> A millisecond latency will cause a significant out of phase shift to the AC signal too.
Ooh, good point. But then, we have I think micro- or even nano-second clock sync over public internet now. Or use (reliable) GPS timestamps codified in the sync signal.
> The AC data you want is already codified in the power signal
You're right, but:
> there needs to be more communication across the grid than just responding to the locally observed frequency
Maybe large stations need a grid-wide sync signal that is independent of load, but small (e.g. domestic, community) suppliers just latch on the observed AC (which is easy for an inverter).
It's not about getting the generator to spin at 50hz, it's about getting it to spin at 50hz under dynamic load. Most generators are mostly fixed speed I think for the sake of efficiency, so the mechanism the grid has to regulate the frequency of power is largely by controlling the number of generators running at any given time, and how much of the grid is connected to them. Ramping up and down based on the current observed frequency of the grid is exactly how this all works. However it's more complicated because what if two or more power stations see a low frequency grid and decide to ramp up more generators? Well now the grid frequency is too high. If they them respond to that by ramping down, it'll go too low. Rinse and repeat and you get undesirable frequency oscillation, so there needs to be more communication across the grid than just responding to the locally observed frequency.
Connect a stopped, or sufficiently out of phase generator to a full scale power grid and the puny thousands of pounds of metal that make up a generator are going to get ripped apart by the sheer inertia of the power grid forcing it to match speed and phase in an instant. This is also a problem when attempting to connect two isolated grids, of they're too out of phase with each other, catastrophic physical damage will occur.