Load shedding why?
October 17, 2024 10:51 AM Subscribe
I feel silly asking this, but how and why does load-shedding work?
Load shedding helps the power grid avoid blackouts by reducing power demand, but I don't know how. If you turn off the power and then turn it on 15 minutes later, don't you just go back to using the same amount of power? I feel like my household actually goes to using more power, because we immediately turn on the kettle and plug in our phones as soon as it goes back on. Also, why is it different from rolling blackouts? Just a different name, or what?
There's something with letting the power generating infrastructure catch up, but it doesn't make sense-- wouldn't you just need to reduce demand or increase supply at some point? Why does turning it off and then on again help? There has to be something I am missing here....please help me out with your understanding and/or some resources that describe exactly what is going on with the electricity grid.
Load shedding helps the power grid avoid blackouts by reducing power demand, but I don't know how. If you turn off the power and then turn it on 15 minutes later, don't you just go back to using the same amount of power? I feel like my household actually goes to using more power, because we immediately turn on the kettle and plug in our phones as soon as it goes back on. Also, why is it different from rolling blackouts? Just a different name, or what?
There's something with letting the power generating infrastructure catch up, but it doesn't make sense-- wouldn't you just need to reduce demand or increase supply at some point? Why does turning it off and then on again help? There has to be something I am missing here....please help me out with your understanding and/or some resources that describe exactly what is going on with the electricity grid.
I believe you're correct that load shedding and rolling blackouts refer to the same thing. This wikipedia page on Demand Response has plenty of information that might be relevant.
posted by number9dream at 11:06 AM on October 17
posted by number9dream at 11:06 AM on October 17
Sorry, posted too soon. Another reason for load shedding/rolling blackouts, in addition to what AndrewInDC noted above, is that some types of power plants take longer than others to ramp up. So the grid operator might start load-shedding until additional power can be brought online.
posted by number9dream at 11:08 AM on October 17 [2 favorites]
posted by number9dream at 11:08 AM on October 17 [2 favorites]
Demand varies quite a lot and can change quite quickly. Supply can also vary, either planned to meet predicted changes in demand, or unexpectedly due to technical issues. If demand is unexpectedly high or some supply falls over then the system may need to lose some demand temporarily. This may mean initiating a planned shutdown for some areas. Some technical issues can mean back up is delayed. If it runs longer then you can move the areas about so everyone gets a turn without. The delay may only be the length of getting another power station up to speed, but this is also a variable. Hydro is less than a minute, gas is fast, coal less so, nuclear slow.
You can also use demand side management to get demand down by basically asking for incentivised volunteers to shut off some demand and meeting your unexpected shortfall from what you save. You may see more of this in future.
You also see load shedding in places without enough power to bring online in which you might get rolling blackouts regularly. This is pretty common in some developing countries.
posted by biffa at 11:17 AM on October 17 [1 favorite]
You can also use demand side management to get demand down by basically asking for incentivised volunteers to shut off some demand and meeting your unexpected shortfall from what you save. You may see more of this in future.
You also see load shedding in places without enough power to bring online in which you might get rolling blackouts regularly. This is pretty common in some developing countries.
posted by biffa at 11:17 AM on October 17 [1 favorite]
I don't know much about commercial-scale power, but on a warship this was sometimes a thing as well. On a much smaller scale...
I feel like my household actually goes to using more power, because we immediately turn on the kettle and plug in our phones as soon as it goes back on.
Not just that, but big motors have a higher starting current than running current, and you get all that starting current at once when power is restored. Then lots of things have to play "catch-up", like air conditioning bringing hot buildings back down, etc.
But as a drastic measure, the total system load IS lower during the "off" time, making more power available to the vital areas that weren't shut off. And a lot of controllers aren't auto-restart. Someone has to go press the button or reset the switch to restart the load. So even a momentary "off-on" can end up with a lower load (temporarily), and then you can try to manage which loads can get restarted when.
posted by ctmf at 11:19 AM on October 17 [1 favorite]
I feel like my household actually goes to using more power, because we immediately turn on the kettle and plug in our phones as soon as it goes back on.
Not just that, but big motors have a higher starting current than running current, and you get all that starting current at once when power is restored. Then lots of things have to play "catch-up", like air conditioning bringing hot buildings back down, etc.
But as a drastic measure, the total system load IS lower during the "off" time, making more power available to the vital areas that weren't shut off. And a lot of controllers aren't auto-restart. Someone has to go press the button or reset the switch to restart the load. So even a momentary "off-on" can end up with a lower load (temporarily), and then you can try to manage which loads can get restarted when.
posted by ctmf at 11:19 AM on October 17 [1 favorite]
Rolling blackouts are one kind of load shedding, though there are other ways to do it. It's basically any kind of controlled blackout.
I think one of the challenges of our modern energy grid system is that for various technical reasons I'm not qualified to explain, power plants actually can't sit around pumping out extra power when demand is low, waiting for demand to increase. Supply has to be managed very carefully, both undersupply and oversupply. Power plants do have basically scheduled times to turn off based on prediction of loads. But if something happens to suddenly drop energy production or transmission, they can't snap their fingers and turn another plant back on.
posted by muddgirl at 11:27 AM on October 17
I think one of the challenges of our modern energy grid system is that for various technical reasons I'm not qualified to explain, power plants actually can't sit around pumping out extra power when demand is low, waiting for demand to increase. Supply has to be managed very carefully, both undersupply and oversupply. Power plants do have basically scheduled times to turn off based on prediction of loads. But if something happens to suddenly drop energy production or transmission, they can't snap their fingers and turn another plant back on.
posted by muddgirl at 11:27 AM on October 17
if you black out 1/3rd of your customers 1/3rd of the time, switching which third every so often, you only need 2/3rds as much power generation at any one time. everyone gets power for 2/3rds of the day, though it will be a different 2/3rds in different areas
posted by BungaDunga at 11:31 AM on October 17
posted by BungaDunga at 11:31 AM on October 17
One potential driver for a load shed action is to prevent a mismatch between the load and the supply that, if of a certain degree, can cause the AC frequency to drop to a problematic level. In North America and a few other places, all elements of the electrical system expect a frequency of 60Hz. Some can stand variation in that frequency, but others cannot. A utility will proactively shed load to prevent problems that can arise in the system if the frequency drops too greatly.
Under Frequency Load Shed, or UFLS, was discussed at length after the troubles in Texas a few years back.
An excerpt from a writeup:
Around 01:33 hours the system’s frequency moved to center stage. Too little generation vs. load, and the system frequency decreases, which was exactly what was happening. For the record, large shifts in system frequency can damage generators and other equipment. There is also an issue with the generators’ and transformers’ volt-per-hertz limits, but I’m not going into that. Let’s just say exceeding these limits isn’t good.
All of this means that when ERCOT’s normal system frequency was starting to drop below the 60 Hz point it got everyone’s attention. By 01:45 the system frequency was nearing 59.86 Hz and more load shedding was ordered. Unfortunately, ERCOT continued to experience generation outages and the system frequency continued to decline. A major low system frequency event took place about 01:50 hours, with the system frequency dropping to 59.4 Hz.
Hitting a system frequency of 59.4 Hz or going below is an important milestone because of the protection scheme known as generator under frequency ride-through. This protection scheme has several trigger points, each with a progressively shorter countdown. But for the Feb. 15 events, ERCOT’s operators were able to keep the system frequency between 59.4 Hz and 58.4 Hz bandwidth. This is important because when the system entered that band, a nine-minute countdown begins.
With the countdown running, successive generation tripping events caused the system frequency to drop to its lowest point of 59.302 Hz (01:52). At that point, another of ERCOT’s safety protocols, the UFLS (under frequency load sheading) scheme kick in. The UFLS system dropped about 6,500 MWs of load over the next couple of minutes.
posted by jerome powell buys his sweatbands in bulk only at 12:24 PM on October 17
Under Frequency Load Shed, or UFLS, was discussed at length after the troubles in Texas a few years back.
An excerpt from a writeup:
Around 01:33 hours the system’s frequency moved to center stage. Too little generation vs. load, and the system frequency decreases, which was exactly what was happening. For the record, large shifts in system frequency can damage generators and other equipment. There is also an issue with the generators’ and transformers’ volt-per-hertz limits, but I’m not going into that. Let’s just say exceeding these limits isn’t good.
All of this means that when ERCOT’s normal system frequency was starting to drop below the 60 Hz point it got everyone’s attention. By 01:45 the system frequency was nearing 59.86 Hz and more load shedding was ordered. Unfortunately, ERCOT continued to experience generation outages and the system frequency continued to decline. A major low system frequency event took place about 01:50 hours, with the system frequency dropping to 59.4 Hz.
Hitting a system frequency of 59.4 Hz or going below is an important milestone because of the protection scheme known as generator under frequency ride-through. This protection scheme has several trigger points, each with a progressively shorter countdown. But for the Feb. 15 events, ERCOT’s operators were able to keep the system frequency between 59.4 Hz and 58.4 Hz bandwidth. This is important because when the system entered that band, a nine-minute countdown begins.
With the countdown running, successive generation tripping events caused the system frequency to drop to its lowest point of 59.302 Hz (01:52). At that point, another of ERCOT’s safety protocols, the UFLS (under frequency load sheading) scheme kick in. The UFLS system dropped about 6,500 MWs of load over the next couple of minutes.
posted by jerome powell buys his sweatbands in bulk only at 12:24 PM on October 17
If you turn off the power and then turn it on 15 minutes later, don't you just go back to using the same amount of power?
15 minutes later, either you are now turning off the power for some other group of users, so the total demand remains lower than usual, you have managed to bring more power online (increasing generation by starting up power stations that are slower to come online, the weather has changed in your favour, i.e. it is now sunnier or windier, or you have managed to buy some power from elsewhere), or demand has decreased (it's warmer out so there is less demand for electric heat, it's now night and people are sleeping, you've convinced a major industrial user to shut down, you've begged residential users to reduce demand).
But fundamentally, when you get to the point of doing load shedding, it's just that you don't have enough power for everybody and you can risk grid collapse (through frequency drop as detailed above, after which it can then take a long time to bring power generation back online as we saw in Texas a few years ago) or you can just cut off some portion of your customers. The second option is obviously better.
Plugging in your phones uses very little power, but it's probably not ideal to turn on the kettle as soon as the power comes back on if you want to be civic minded about it.
We don't have rolling blackouts or load shedding where I live in British Columbia, but the utility does have agreements with major industrial users that give them cheaper power on the condition that they have to shut down in an emergency as well as programs to reduce residential demand in peak times (either via a connected thermostat that the utility can temporarily set back a couple degrees or via incentives that they text out to people to reduce demand). They also basically always have a large hydropower turbine spinning with little load that can be spun up to balance generation needs very quickly.
posted by ssg at 12:58 PM on October 17 [1 favorite]
15 minutes later, either you are now turning off the power for some other group of users, so the total demand remains lower than usual, you have managed to bring more power online (increasing generation by starting up power stations that are slower to come online, the weather has changed in your favour, i.e. it is now sunnier or windier, or you have managed to buy some power from elsewhere), or demand has decreased (it's warmer out so there is less demand for electric heat, it's now night and people are sleeping, you've convinced a major industrial user to shut down, you've begged residential users to reduce demand).
But fundamentally, when you get to the point of doing load shedding, it's just that you don't have enough power for everybody and you can risk grid collapse (through frequency drop as detailed above, after which it can then take a long time to bring power generation back online as we saw in Texas a few years ago) or you can just cut off some portion of your customers. The second option is obviously better.
Plugging in your phones uses very little power, but it's probably not ideal to turn on the kettle as soon as the power comes back on if you want to be civic minded about it.
We don't have rolling blackouts or load shedding where I live in British Columbia, but the utility does have agreements with major industrial users that give them cheaper power on the condition that they have to shut down in an emergency as well as programs to reduce residential demand in peak times (either via a connected thermostat that the utility can temporarily set back a couple degrees or via incentives that they text out to people to reduce demand). They also basically always have a large hydropower turbine spinning with little load that can be spun up to balance generation needs very quickly.
posted by ssg at 12:58 PM on October 17 [1 favorite]
As with almost everything to do with Texas, the way ERCOT operates is an exception ☺
Load shedding is a cheap way to keep most of the lights on when demand is too high to keep all of the lights on. If your system is transmission-constrained (not enough wires*), your system operator can't get enough power to everyone who needs it. No amount of additional power can be bought to fix this. If your system operator doesn't control enough generation capacity, they might be able to buy it from neighbouring system operators at a fairly high price. If they have no neighbours, or they've isolated themselves (see Texas/ERCOT, kinda), they have to reduce demand or bits of the system might start falling off.
Lookahead markets in regional transmission organizations are complicated. If you generate too much, you'll have to sell it to your neighbours at a loss. If you generate too little (and have enough wires), some opportunistic operator with an aeroderivative gas turbine plant (aka jet engine in a box: very fast start, hugely inefficient) would be happy to sell you power at a massive price. You want to have just enough spare generation (contingency) that your largest single generator could fail and you could still deliver power to all of your contracted buyers. When it all works, you don't notice it's there, but when it doesn't, it's August 14–16, 2003 all over again.
→
Not as often as BC Hydro and local received wisdom would have you think. BC has been known to buy some extremely grotty fossil generation from other plants in the PNW to make up short-term capacity shortfalls. It's one of the reasons that independent generators got to build wind power (like the one in Tumbler Ridge I helped build) and other renewables in BC because the dependence on imported fossil fuel was a political embarrassment. Site C is an environmental catastrophe of a mega-dam project, but it's too far away from people who matter, so it went ahead.
--
*: simplistically. The grid is highly dynamic. There might be enough wires, but someone else has paid to wheel power through them at that time, or many other weird contractual things that go on every day in the post-ENRON power markets.
posted by scruss at 1:38 PM on October 17 [3 favorites]
Load shedding is a cheap way to keep most of the lights on when demand is too high to keep all of the lights on. If your system is transmission-constrained (not enough wires*), your system operator can't get enough power to everyone who needs it. No amount of additional power can be bought to fix this. If your system operator doesn't control enough generation capacity, they might be able to buy it from neighbouring system operators at a fairly high price. If they have no neighbours, or they've isolated themselves (see Texas/ERCOT, kinda), they have to reduce demand or bits of the system might start falling off.
Lookahead markets in regional transmission organizations are complicated. If you generate too much, you'll have to sell it to your neighbours at a loss. If you generate too little (and have enough wires), some opportunistic operator with an aeroderivative gas turbine plant (aka jet engine in a box: very fast start, hugely inefficient) would be happy to sell you power at a massive price. You want to have just enough spare generation (contingency) that your largest single generator could fail and you could still deliver power to all of your contracted buyers. When it all works, you don't notice it's there, but when it doesn't, it's August 14–16, 2003 all over again.
→
British Columbia ... basically always have a large hydropower turbine spinning with little load
Not as often as BC Hydro and local received wisdom would have you think. BC has been known to buy some extremely grotty fossil generation from other plants in the PNW to make up short-term capacity shortfalls. It's one of the reasons that independent generators got to build wind power (like the one in Tumbler Ridge I helped build) and other renewables in BC because the dependence on imported fossil fuel was a political embarrassment. Site C is an environmental catastrophe of a mega-dam project, but it's too far away from people who matter, so it went ahead.
--
*: simplistically. The grid is highly dynamic. There might be enough wires, but someone else has paid to wheel power through them at that time, or many other weird contractual things that go on every day in the post-ENRON power markets.
posted by scruss at 1:38 PM on October 17 [3 favorites]
> If you turn off the power and then turn it on 15 minutes later, don't you just go back to using the same amount of power
Let's say the power company has 10% more load than they can handle. This situation won't be forever - it might be for just 10 or 15 minutes, or maybe an hour or two.
Let's say just for example this problem situation will last 30 minutes.
So you can turn off power to 10% of customers for 30 minutes and problem solved.
By that time, the load overage has resolved (either demand has fallen - it is extremely variable throughout the day - or more power plants have come online). When you turn power back on in 30 minutes, it can indeed lead to a little surge in power usage as you turn everything back on. But they have ways of dealing with that - for example, by gradually adding users back into the system so that initial surge is smoothed out.
However, losing your power for a long time is very inconvenient, so what they usually do instead is turn of 10% of customers for, say, 10 minutes, then turn those people back on and turn another different 10% off for 10 minutes, then turn those back on and turn a 3rd group of 10% off for 10 minutes.
All of this is done gradually and smoothly, as I outlined above, to avoid the load spikes as you bring groups back online.
From the utility's perspective, they have cut usage by 10% for 30 minutes (or maybe not quite 10% - maybe it's only 8% due to the factors you mention). From your perspective, you lost power for 10 minutes only.
"And as soon as I got power back I turned EVEN MORE ON BECAUSE I WAS MAD (or cold or worried, or whatever)."
Doesn't matter - they generally know how people behave when the power is off and when it is turned back on again and they just take all this into consideration as they roll the outages around as needed to keep the system up.
Altogether they know they need to reduce load by X%. So they cut off X% of users in a rolling manner. From their perspective it makes perfect sense.
From your perspective as a normal user, your power just turns off randomly for a while for no particular discernable reason and you really can't figure out why.
posted by flug at 3:29 PM on October 18
Let's say the power company has 10% more load than they can handle. This situation won't be forever - it might be for just 10 or 15 minutes, or maybe an hour or two.
Let's say just for example this problem situation will last 30 minutes.
So you can turn off power to 10% of customers for 30 minutes and problem solved.
By that time, the load overage has resolved (either demand has fallen - it is extremely variable throughout the day - or more power plants have come online). When you turn power back on in 30 minutes, it can indeed lead to a little surge in power usage as you turn everything back on. But they have ways of dealing with that - for example, by gradually adding users back into the system so that initial surge is smoothed out.
However, losing your power for a long time is very inconvenient, so what they usually do instead is turn of 10% of customers for, say, 10 minutes, then turn those people back on and turn another different 10% off for 10 minutes, then turn those back on and turn a 3rd group of 10% off for 10 minutes.
All of this is done gradually and smoothly, as I outlined above, to avoid the load spikes as you bring groups back online.
From the utility's perspective, they have cut usage by 10% for 30 minutes (or maybe not quite 10% - maybe it's only 8% due to the factors you mention). From your perspective, you lost power for 10 minutes only.
"And as soon as I got power back I turned EVEN MORE ON BECAUSE I WAS MAD (or cold or worried, or whatever)."
Doesn't matter - they generally know how people behave when the power is off and when it is turned back on again and they just take all this into consideration as they roll the outages around as needed to keep the system up.
Altogether they know they need to reduce load by X%. So they cut off X% of users in a rolling manner. From their perspective it makes perfect sense.
From your perspective as a normal user, your power just turns off randomly for a while for no particular discernable reason and you really can't figure out why.
posted by flug at 3:29 PM on October 18
You are not logged in, either login or create an account to post comments
The purpose is to provide some measure of predictability for electricity users of when they will affected by the outages.
posted by AndrewInDC at 11:04 AM on October 17 [1 favorite]