News stories are saying that had the utilities not cut off demand during the emergency, long-lasting damage to the infrastructure would have been much worse.
Why is that? If users demand more power than is in the system, what physical part of the network is damaged?
It’s a good question, I remember reading that and being puzzled by it. A supplementary question is why there aren’t emergency automatic shutoff procedures to prevent something that would cause damage that would take months to fix, and why this apparently depends on a human making some kind of could-go-both-ways judgment call.
It boils down to frequency. As people may know, our AC power generation works on 60Hz (+/- 0.5Hz). So, there’s a ‘pulse’ of sorts that beats 60 times a sec. And everything that’s interconnected needs to be synchronized to the same ‘beat’ essentially. If that frequency drops outside tolerance - even as low as 59.4, equipment (at both ends) can be damaged and power plants knocked offline. In a bad enough situation, plants get knocked offline in cascade. Out of sync power is quite a problem for anything with a motor (like large parts of our power system and industrial units).
The problem in Texas was that the system frequency was dropping. It got down to 59.3Hz and there was an imminent danger of system collapse.
But ok, let’s say the system collapses. Why does it take so long to recover? That’s because when elements are added back into the grid, they have to be in-sync as they’re added. If voltage and phase are out of sync, that can lead to massive current spikes arcing and further damage. If the grid is fully operational, adding a single generator back online takes a bit of work but isn’t too big a hassle in the grand scheme. But if major parts of the grid are down, they have to be added back in and sync’ed in a slow, careful process almost one at a time to avoid just having to shut the whole thing down again.
This is where home solar generation that feeds into the grid is sometimes an issue. There’s actually a bit of work that has to be done to make sure that power is sync’ed to the system as well.
Why aren’t shutdowns automatic? It’s a question of what do you shutdown. If the frequency drops, the solution is to shed load from the grid. So, who loses power first? Who makes that call? And how many people lose power in freezing weather? I suppose those choices can be automated, but that’s also a big mess - more political/humanitarian than technical.
As an aside, in other countries, especially in Europe, this may be 50Hz. Japan is a weird example where the eastern half the country is 50 and the western half 60 for historic reasons - two of the original generators in the country were purchased from Germany and the US. This makes it difficult to interconnect power generation between halves of the country, which was a problem in the aftermath of the Fukushima meltdown.
Also any equipment that is damaged has to be manufactured from scratch, which is a months-long process. There’s no warehouse of large substation transformers and switch gear just sitting around waiting because most of it is custom made to order. https://fas.org/sgp/crs/homesec/R43604.pdf
My thanks to the Great Antibob.
How does the frequency change with the loss of capacity? And (don’t know if this is even possible) could there be automated shut-offs to basically wall off a region with power if the grid frequency drops too far?
It seems to me that even if you had a series of disconnected islands of power surrounding operating generators, that it would be easier to reconnect them than it would be to let the whole thing crash and try and bring it up from cold.
It’s more that the generation can’t keep the frequency at 60Hz if there’s too heavy a load. So, drop people off the grid, and it can maintain 60Hz.
I suppose so. But considering the way people react to school districting and flood zones, do you want to know you live in an “automatic shutoff” island? Would your company want to be located there if there’s a chance sensitive equipment will be shut off? That’s probably less expensive for the power companies - they don’t have to spend on updates and can even reduce capacity knowing that some people will just be kicked off during high usage days without risk to the system.
It’s probably better overall if the Texas grid was like the rest of the country and not at risk of these situations in the first place.
May I restate the concept to see if I clearly understand?
Build a scale model of the Texas grid in your back yard. Now fire it up. If you light off the whole thing at once your would blow a gasket. Instead you turn on this power plant and attach it to that load. You wait for that configuration to stabilize then you add more load. Next you turn on another section and let that settle in. After that you can very carefully try to connect these first two areas.
If you have to do that across the entire interconnection it would be both tricky and time-consuming.
Is that the idea?
Think of generators like huge motors (that’s actually basically what they are). How fast they spin determines the frequency of the alternating current they generate. A generator with no load on it spins almost freely, you just have to overcome friction and inertia. It’s basically a flywheel. Once you attach an electrical load to the generator however, the resistance goes up precipitously, so it gets much harder to turn. So if the load on a generator goes up because of increased demand, lack of additional generators on the grid, or whatever, it either needs more mechanical input power (steam for the turbines usually) to keep up the speed/frequency, or it starts to slow down.
Adding to this, and elaborating on other material upthread:
When you want to power up a new power plant and connect it to the grid, the process goes something like this:
Light the burner/nuke, make heat and start making steam.
Use the steam to spin the turbine and alternator (the actual device that converts mechanical shaft power into AC electrical power). The alternator is not connected to the grid at this time.
Get the alternator up to its operating speed, at which it’s producing 60 Hz on the line.
Observe the phase of the alternator’s output, relative to the phase of what you’re observing on the electrical grid (see synchroscope). They’re probably out of phase, unless you’re extraordinarily lucky. If the alternator is lagging behind the grid, speed it up just a smidge; if it’s leading the grid, slow it down just a smidge.
When the alternator and the grid are exactly in phase, you you can connect it.
Now that your alternator is connected to the grid, which one is driving which depends on how much shaft power you’re supplying to your alternator. If your turbine stops driving your alternator, the grid will continue to drive it; your alternator will still spin at 60 Hz, but it will lag in phase behind the grid, consuming power from it and reducing grid voltage. If you run your turbine at full steam, the grid will hold back your alternator’s speed so that it still spins at 60 Hz - but your alternator will lead the grid, putting out power and raising grid voltage.
Your job as a plant operator is to supply the right amount of turbine power so that your plant sustains grid voltage within a narrow range. Meanwhile, that’s what all the other power plant operators are doing. If you want to bring them all up together, it’s a slow, incremental balancing act to avoid systemic voltage excursions and excess currents happening in expensive places.
This analogy maybe misleading. Motors and Generators above a certain power (MW range) are synchronous Motors/Generators
Their speed doesn’t change with the power they produce. A synchronous motor/generator spins at about the same speed at 100% capacity or 50% capacity.
@jjakucyk is correct in that their speed determines their frequency. You are also correct in that as synchronous alternators, once they are connected to the grid, their speed doesn’t change WRT the power they produce; their power production/consumption only affects whether their output frequency leads the grid or lags behind it.
The complexity of bringing up a bunch of power plants from a dead stop might be better understood by understanding how many plants are involved: Google says there are about 650 power generating plants in Texas. Trip a bunch of them out, and you have to coordinate between all of them to get them all back up in a controlled manner. This includes connecting/disconnecting parts of the grid in a carefully planned way to avoid drawing so much power from any isolated group of plant(s) so that you actually cause those plants to slow down, which would cause grid frequency and voltage to fall out of tolerance - at which point you’d have to trip those plants out and start over.
Huh. I always figured there was a fairly large layer of electrical equipment that somewhat buffered the generator itself from that kind of thing and kept the frequency steady, etc…
How quickly can the electrical grid react to a sudden spike in demand? For example, what would happen in a “Christmas Vacation” scenario, where a homeowner flicked on so many lights that it caused outages and the power company has to bring on auxiliary power?
Any kind of buffer would itself be a load and/or generating layer to match to on both sides. Such a buffer would have to be powered, after all. It’s just turtles all the way down at that point.
Mechanical inertia is useful here, and you don’t have to power it. I mean once, sure, to get the rotating assembly spinning at its target RPM, but after that the inertia acts as reserve of mechanical energy to mitigate the slowdown of your power plant’s alternator when there’s a sudden surge in demand.
So say you and ten of your neighbors all plug in your Teslas at the same time. That’s a whole megawatt of new demand. The power plant’s turbine is still making X amount of power, but now the grid is dragging down the plant’s RPM. The plant’s control system will make more steam to make more shaft power to get the alternator RPM back up to its target speed. It’s going to add a megawatt of turbine mechanical power output to supply the new demand of all those Teslas, plus some extra get the plant’s alternator back up to target RPM. Inertia in this system reduces how far the RPM deviates from target before the control system can increase turbine output and get RPM back up where it’s supposed to be.
Seriously. I nominate @Great_Antibob for “Ignorance Fighter of the Week.” @Machine_Elf a close second. Thanks, both, for explaining the mysteries of power grids such that someone like me can understand it.
Yeah, it is pretty much impossible to isolate one section of the electrical power distribution system from another section and be able to reintegrate them rapidly enough to prevent disruption. The general public does not have an appreciation for just how delicate the electrical distribution is. And while renewables were not responsible for the failure of the ERCOT grid, the time-varying nature of wind and solar does present additional challenges which argues for the creation of an integrated nationwide HVDC grid.
Not directly related to the failure of the ERCOT grid due to a lack of preparation for freezing weather, but regarding the fragility of the electrical power grid and potential impact, here is the JASON Group Report: “Impacts of Severe Space Weather on the Electric Grid”.
Stranger
How could Germany sell generators to Japan after the Russians had carted all of theirs off?
The two different frequency energy grids were established in the late-1800s.