ELI5, how can an electrical grid be “minutes away” from month long blackouts? What would’ve happened that devoted employees avoided?

223 views

[ad_1]

I’ve seen lots of posts lately on Texas being “minutes and seconds away” from months long blackouts. What could’ve happened, what was avoided that caused that?

In: Engineering
[ad_2]

Operators shut down switching equipment that was very close to overloading. Well, “seconds” may have been hyperbolic, but this equipment is extremely specialized and there are not huge stockpiles of unused equipment waiting for someone to want it. Losing too much switching equipment might lead to months of rolling blackouts as replacement equipment was manufactured.

Most world power grids are susceptible to this sort of thing, nobody has a spare power grid to use for fallback. A large solar storm or coronal mass ejection could fry enough transformers that it would take months for worldwide power to return to normal.

An electric grid is a balance of push and pull. Most grids right now have little or no storage, which means the grid has to maintain enough flow to feed the demand without exceeding the maximum amount the grid can handle. Because it is a network there is some capacity to buffer by opening and closing distribution nodes and varying the amount of production coming from power plants. You want to do this within a fairly short period of changes to demand– seconds to minutes, no more.

If you have a smaller network like Texas that normally has some number of plants running (let’s say ten for the sake of discussion) and one of those plants stops its production, in this case valves and other machinery that moved the gas (which drove the plant’s generators) iced up. Now the grid is operating at 90% output.

Once this happens, either a percentage of the remaining nine need to pick up the slack by increasing their output, or… you shut down a percentage of the grid equivalent to that plant’s output. Either up the output or decrease the demand.

Long story short, this sequence of balancing can set off a cascade of ups and downs that ripple through the network as the system tries to maintain the narrow window of balance mentioned up top. If two plants still online try to compensate for the decrease of the plant that went down and overshoot, then cut back and drop under. Then plant five may ice up and things get more complicated. The push and pull becomes the equivalent of a truck that starts fishtailing– the more the driver struggles to correct things, the bigger the oscillations grow.

At some point this push and pull results in a power dump so big that you are in danger of overloading some major piece of equipment, and you have to cut off production entirely and start turning things back on one-by-one.

It must be stated that these type of situations are actually somewhat common, but MOST of the time the electricity can be shunted around the network enough to allow the operators to get the situation under control without cutting anyone off. There are exceptions, some very notable, but those are unusual.

The other possibility is that so much production went down that the remaining plants tried to take up the demand and ended up overloading their capacity, which would also fry equipment.

​

Either way, once the problem moves from the category of “evolving, use the network’s buffer volume now to correct” to “overload imminent, you’re f’d” the only good option is to literally throw the switch and completely knock out the whole thing in order to prevent equipment loss. Then you can start turning things back on in a controlled fashion, though in this case the gas lines feeding the power plants are iced up and can’t get the gas they need to turn their turbines, so there is no turning on that can be done.

​

[https://en.wikipedia.org/wiki/Cascading_failure](https://en.wikipedia.org/wiki/Cascading_failure)

From the [Houston Chronicle](https://www.texastribune.org/2021/02/18/texas-power-outages-ercot/):

>”As natural gas fired plants, utility scale wind power and coal plants [tripped offline due to the extreme cold brought](https://www.texastribune.org/2021/02/16/natural-gas-power-storm/) by the winter storm, the amount of power supplied to the grid to be distributed across the state fell rapidly. At the same time, demand was increasing as consumers and businesses turned up the heat and stayed inside to avoid the weather.
>
> …
>
>The worst case scenario: Demand for power outstrips the supply of power generation available on the grid, causing equipment to catch fire, substations to blow and power lines to go down.
>
>If the grid had gone totally offline, the physical damage to power infrastructure from overwhelming the grid could have taken months to repair, said Bernadette Johnson, senior vice president of power and renewables at Enverus, an oil and gas software and information company headquartered in Austin.

It’s like if you were powering a big bulb with a bunch of batteries… but the batteries started getting disconnected one by one. As you lose power, but demand stays — it puts too much strain on the remaining batteries and catches stuff on fire.

Batteries = power plants

Bulb = customers

Large utilities have somewhat complicated algorithms called “Remedial Action Schemes” (RAS). A RAS is “armed” by system operators when conditions are nominal. The RAS equipment then monitors a set of transmission lines and generation in a geographical area. If a given combination of lines and/or generation goes offline that would result in a severely overloaded line, the scheme automatically trips remaining lines or generation offline to reduce the possibility of a line overheating. All lines have their own protection to interrupt currents that exceed a maximum determined by the line characteristics (wire type, wire size, ambient temperature, sag of the line between towers or poles, etc.) However, the RAS actuates much more quickly than normal protection, averting an overheating or ‘annealing’ of the line which could cost millions to repair or replace. Most lines are not designed to operate above 275 degrees C for any length of time.

Not saying with certainty that this is what happened in TX, but it is one way that lots of resources can be tripped off line in seconds.

They don’t. It’s bullshit.

There have been multiple failures like that, none have taken out the grid for months.

Just going down the list of [cascading power failures](https://en.wikipedia.org/wiki/Cascading_failure) in wikipedia:

Blackout in Northeast America in 1965 – 13 hours
Blackout in Southern Brazil in 1999 – 4 hours
Blackout in Northeast America in 2003 – 4 days in some places
Blackout in Italy in 2003 – 12 hours
Blackout in London in 2003 – 2 hours
European Blackout in 2006 – 2 hours
Blackout in Northern India in 2012 – 2 days
Blackout in South Australia in 2016 – 1 day
Blackout in southeast South America in 2019 – 2 days

Imagine one of those [rides at the fair](https://en.wikipedia.org/wiki/Kamikaze_(ride)) where you have a car full of people swinging back and forth. That’s the electrical grid, and the people on the ride are the load on the electrical grid – if you add more people it’s harder to keep it swinging, just like if you use more electricity for heating, pumping water, running appliances, it’s harder to keep the electrical grid running.

The ride is swinging back and forth once every second (the electricity in the electrical grid swings back and forth at 60 Hz, or _sixty_ times per second, but people move more slowly than electricity, so let’s make it sixty times per _minute_ instead). And, instead of having a big motor swinging the ride back and forth, you have a bunch of carnival workers. When the rides is at one side they all push, and when it’s at the opposite side, they pull, just like pumping on a swing. The carnival workers are the generators tied into the electrical grid, each one contributing a little bit in unison.

Normally, this works fine. You have more carnival workers than you really need to keep the ride moving, so no one of them has to push as hard as they can, and you can have a few more people get on the ride without needing to add more workers. This is like how the electrical grid has more power available than is needed in normal times.

Now, suppose that you have a cold snap that shuts down the other outdoor events, so people want to ride the rides that block out the wind. More people get on the ride, just like more people start turning up the heat. At the same time, the workers start to get cold – some of them may have bad backs or knees, and can’t push when it gets that cold, so they drop off to take a break, and now you _barely_ have enough workers to keep the ride running, pushing and pulling as hard as they can. This is where the ride (and the electrical grid) start to fail – the generators on the electrical grid are running as hard as they can, but people keep needing more and more power.

Next, the generators start to fail – some of them just _break_, like an aging carnival worker who tries to push and pull as hard as they did a decade ago and throws out their back, but it’s more likely they just start to get a bit tired – nobody can work _as hard as they possibly can_ for very long.

Here’s where things start to get “out of phase”. The workers are trying to push and pull once per second, just like the electrical grid is trying to run at 60 Hz. But as they get tired, they slow down a little bit. A tiny bit is okay, because they’re still pushing and pulling _about_ when they need to, at _almost_ the same time as everyone else, but it’s not great, just like a generator running at 59 Hz – it’s still helping, but not as much as it could.

But since they aren’t in sync with everyone else, they aren’t helping as much as they could. Things get even harder, because there are still the same number of people on the ride. If things get _too_ bad, if a worker gets too out of sync with the other workers, then they’ll start pushing when they should be pulling, or pulling when they should be pushing. The ride is so much bigger than them that, well, it’s going to be moving back and forth regardless. So if they don’t stay in sync with it, then it’s either going to smash into them, yank them off their feet, or (best case…) pick them up and start carrying them with it.

As soon as one of those things happen – just like a generator turning into a motor or being damaged and dropping off – the load gets that much worse for the remaining ones, which escalates the problem further until _all_ of the workers (or generators) have been damaged or knocked off, and the ride slows to a halt. This would have been the blackout, and it would have taken months to get everything back online to start the ride moving again – repair the generators, heal the workers, find new ones, etc.

This can happen quickly, because once it starts it goes faster and faster. If you can’t bring on more workers or generators, the only solution is to lighten the load – to kick riders off so that the remaining workers can manage it. That’s what happened in Texas – they were _that_ close to the electrical grid smashing its own generators like a carnival ride hitting the workers who were running it, and they could only stop that by shutting things down.

California once toyed with the idea of a smart-grid, instead of building more power plants.

Their biggest issue was air conditioning in the summer during noon to 6PM. No matter how expensive electricity got, some people would pay whatever it took to keep the house/business cool.

Customers who signed up to get a state-controlled breaker box were given a break on the price of electricity.

If the system neared an overload, certain strategic homes would have their A/C turned off for an hour, then after an hour it would go back on, and a different group would have their A/C cut.

I dont know how that worked out, but they dont ser to be doing it for some reason.

Rolling brown-outs can be useful. Because once the power is back on, people who care about the grid staying up should only turn items back on that are the most necessary. For Texas, that would be the heaters. Even if they use gas for heat, the electricity must be on for the required safety features to operate.

The management of a Power Grid is a balancing of demand and supply, there was a story from England how the Power Plant Operators would watch the soccer games as at intermission so many people would turn on their kettles it would destabilize their grid.

So what happens when your demand exceeds your supply, well the power grid functions at a set voltage and 60 cycles per second and as load exceeds supply all the motors connected slow down, but as they try to keep doing their job the amount of current required increases. All switch gear, transformers and transmission lines have a max ampere rating.

You couple this with the heating load increase due to cold temps and now you are in a situation where if you do not disconnect part of the load to maintain the balance of available generation and load you will overheat and melt switch gear, transformers, transmission lines, you will damage any equipment that is supposed to be protected by disconnects if they melt solid.

You can think of your own home if you plug a kettle and waffle iron and a toaster into the same outlet, the 110 volt breaker for the kitchen would trip as the amperage draw for these three would exceed 12 amperes as designed.

Now your know it all brother in-law puts a 20 amp breaker in the panel and plugs in all 3 appliances but this over-loads the wires in your house and sets it on fire.

Now you don’t have a home. It would have been simpler to unplug 1 or 2 items and manage the load.

The power is mostly restored already; it has been days. The idea of months long blackouts is unrealistic nonsense. There was nothing wrong with the electrical grid its self, just that too many of it’s various diverse types of power supply sources failed to work in the extreme cold. This is an extremely politically charged topic, so such things as projecting “month long blackouts”, is likely just dishonest political rhetoric.

Which statement sounds better:

“We purposefully did not winterize any of our equipment, despite warnings 10 years ago that this could occur. We apologize for causing this mess and hope not too many people needlessly die.”

OR

“We saved – by mere seconds – the complete collapse of the electrical grid. We saved everyone months and months of being without power. Aren’t we awesome!?!”