5 ms·

This goes along with the coronal mass ejection risk story that was on the front page a day or two ago. Is a nationwide electric grid even a good idea anymore? Y

This goes along with the coronal mass ejection risk story that was on the front page a day or two ago. Is a nationwide electric grid even a good idea anymore? Yes, it's more efficient, but it's also more fragile, in that a failure in critical infrastructure can cause nationwide blackouts. Imagine if most places had a microreactor, a small fossil fuel generator, renewable + storage, or some combination of these, and each small to medium sized community of homes had their own. Things would go wrong all the time on the small scale, but there would be no such thing as a nationwide blackout.

notJim 5y ago
This is interesting, because usually I see nationwide grids being touted as MORE resilient, not less. The idea is if power is knocked out in one area, you can still get it from elsewhere. This came up a lot during the Texas situation.
- rthomas6 5y ago
  Texas is what I would still consider a "nationwide" power grid. I was thinking of something very very decentralized, like individual neighborhoods. It wouldn't completely solve what happened in Texas, but blackouts would be contained to small areas for hopefully shorter amounts of time, as failure in one area would have no effect on other areas, and individual neighborhoods could probably coordinate and conserve power usage as needed better than an entire state.
- godelski 5y ago
  I think there's a balance. You want a slight surplus in your local area and connected so you can pump it somewhere else when that somewhere else goes down. Though IIRC Texas has far more problems than just not being connected (also far more than just renewables either, which was a big talking point).
wmf 5y ago
Smaller grids like Texas are more susceptible to weather events though.
thereare5lights 5y ago
Are there examples of how a failure at a power plant could affect the national grid? As mentioned by another comment, one advantage of a national power grid is the ability to transfer power from other places during power failures. But I'm not sure how a power failure would affect the rest of the grid.
- akiselev 5y ago
  > Are there examples of how a failure at a power plant could affect the national grid? A failure at any single power plant is very unlikely to cause cascading failures in the rest of the grid because one power plant usually can't generate enough power relative to its grid. Gross oversimplification but all of the power plants in a national grid are essentially coupled together [1] as a giant circuit made up of a ton of electromagnets - the generators converting coal/natgas powered steam or water/wind to electricity. These electromagnets are all operating at the same frequency in phase with each other which literally means that the generators spin in sync at ~60hz. These generators are often (usually?) heavy turbines with a lot of angular momentum so normally when you connect a power plant, it stays in phase just due to the sheer inertia of the rest of the grid because power follows phase [2]. When the phase changes suddenly - like when connecting a generator that is out of phase with the grid - all that inertia of the generator essentially slams into the electromagnetic force generated by the incoming current. If the rest of the grid (of hundreds/thousands of other generators) produces significantly more power, it'll just destroy the connecting one as if you brought a turbine powered jet engine to an instantaneous dead stop - the momentum of the blades will literally rip them off of the shaft. However, if the newly connected generators are producing a significant amount of power relative to the rest of the grid, like when a corona mass ejection hits or a large islanded subgrid is poorly reconnected, the resulting differential can generate a wave (again, gross simplification) that travels through the grid with enough power to destroy every single generator connected to it. A single power plant just can't generate enough power to do damage to the rest of the grid [3] but if a natural disaster severed the connection between two regions of a grid and they drift out of phase, they must be carefully reconnected because they might have enough power to seriously damage other power plants. [1] Some parts of a grid might be separated by HVDC which in effect isolates the parts [2] Power flows from the faster generators to the slower ones, speeding them up until they hit equilibrium. [3] There are exceptions depending on grid topology, but generally grids are designed to avoid them
epistasis 5y ago
A better national grid for renewables will be hugely redundant, with few single points of failure, because renewables are distributed throughout geography. If something takes out that sort of infrastructure, it's probably going to take out the local micro reactor too. Existing solutions for micro grids are getting super cheap too, if we didn't already have tons of transmission built out, starting a grid from scratch today would probably include not only far less transmission, but far more storage, which also hugely increases resiliency. Having tons of storage and solar right behind the meter, by which I mean 10%-40% of buildings, would hugely increase the resiliency of the grid, if utilities would embrace the Information Age build bottom-up infrastructure in addition to top-down.
Terr_ 5y ago
IANAElectrician, but IIRC the biggest problem isn't the grid size per se, but how grid operation requires always matching the amount of power coming from all producers and the amount being used by all consumers. Like a tightly-sealed water-pipe system where the pressure is uniform throughout, and too much pressure pops things, while too little won't go anywhere. So it's a coordination problem, and ideally you'd have a big-grid which can be arbitrarily partitioned (and recombined) to ensure producer-consumer matching as you get more and more of each involved again. It's one of the few cases where the phrase "smart home" doesn't make me cringe: Imagine a neighborhood of houses in a brownout situation all coordinating to dim the lights and space out their high-energy usage, like running refrigerator compressors.

notJim 5y ago

This is interesting, because usually I see nationwide grids being touted as MORE resilient, not less. The idea is if power is knocked out in one area, you can still get it from elsewhere. This came up a lot during the Texas situation.

rthomas6 5y ago
Texas is what I would still consider a "nationwide" power grid. I was thinking of something very very decentralized, like individual neighborhoods. It wouldn't completely solve what happened in Texas, but blackouts would be contained to small areas for hopefully shorter amounts of time, as failure in one area would have no effect on other areas, and individual neighborhoods could probably coordinate and conserve power usage as needed better than an entire state.
godelski 5y ago
I think there's a balance. You want a slight surplus in your local area and connected so you can pump it somewhere else when that somewhere else goes down. Though IIRC Texas has far more problems than just not being connected (also far more than just renewables either, which was a big talking point).

wmf 5y ago

Smaller grids like Texas are more susceptible to weather events though.

thereare5lights 5y ago

Are there examples of how a failure at a power plant could affect the national grid? As mentioned by another comment, one advantage of a national power grid is the ability to transfer power from other places during power failures. But I'm not sure how a power failure would affect the rest of the grid.

akiselev 5y ago
> Are there examples of how a failure at a power plant could affect the national grid? A failure at any single power plant is very unlikely to cause cascading failures in the rest of the grid because one power plant usually can't generate enough power relative to its grid. Gross oversimplification but all of the power plants in a national grid are essentially coupled together [1] as a giant circuit made up of a ton of electromagnets - the generators converting coal/natgas powered steam or water/wind to electricity. These electromagnets are all operating at the same frequency in phase with each other which literally means that the generators spin in sync at ~60hz. These generators are often (usually?) heavy turbines with a lot of angular momentum so normally when you connect a power plant, it stays in phase just due to the sheer inertia of the rest of the grid because power follows phase [2]. When the phase changes suddenly - like when connecting a generator that is out of phase with the grid - all that inertia of the generator essentially slams into the electromagnetic force generated by the incoming current. If the rest of the grid (of hundreds/thousands of other generators) produces significantly more power, it'll just destroy the connecting one as if you brought a turbine powered jet engine to an instantaneous dead stop - the momentum of the blades will literally rip them off of the shaft. However, if the newly connected generators are producing a significant amount of power relative to the rest of the grid, like when a corona mass ejection hits or a large islanded subgrid is poorly reconnected, the resulting differential can generate a wave (again, gross simplification) that travels through the grid with enough power to destroy every single generator connected to it. A single power plant just can't generate enough power to do damage to the rest of the grid [3] but if a natural disaster severed the connection between two regions of a grid and they drift out of phase, they must be carefully reconnected because they might have enough power to seriously damage other power plants. [1] Some parts of a grid might be separated by HVDC which in effect isolates the parts [2] Power flows from the faster generators to the slower ones, speeding them up until they hit equilibrium. [3] There are exceptions depending on grid topology, but generally grids are designed to avoid them

epistasis 5y ago

A better national grid for renewables will be hugely redundant, with few single points of failure, because renewables are distributed throughout geography. If something takes out that sort of infrastructure, it's probably going to take out the local micro reactor too. Existing solutions for micro grids are getting super cheap too, if we didn't already have tons of transmission built out, starting a grid from scratch today would probably include not only far less transmission, but far more storage, which also hugely increases resiliency. Having tons of storage and solar right behind the meter, by which I mean 10%-40% of buildings, would hugely increase the resiliency of the grid, if utilities would embrace the Information Age build bottom-up infrastructure in addition to top-down.

Terr_ 5y ago

IANAElectrician, but IIRC the biggest problem isn't the grid size per se, but how grid operation requires always matching the amount of power coming from all producers and the amount being used by all consumers. Like a tightly-sealed water-pipe system where the pressure is uniform throughout, and too much pressure pops things, while too little won't go anywhere. So it's a coordination problem, and ideally you'd have a big-grid which can be arbitrarily partitioned (and recombined) to ensure producer-consumer matching as you get more and more of each involved again. It's one of the few cases where the phrase "smart home" doesn't make me cringe: Imagine a neighborhood of houses in a brownout situation all coordinating to dim the lights and space out their high-energy usage, like running refrigerator compressors.