3 ms·
Electricity demand isn't that simple, it's not like Sheetz dropping the price of gas to $1.776/gallon on the 4th of July and having to bag the pumps within the
by saltminer 3y ago
Electricity demand isn't that simple, it's not like Sheetz dropping the price of gas to $1.776/gallon on the 4th of July and having to bag the pumps within the hour [0].
Electricity is quite interesting as a market because it's truly the logical endpoint of just-in-time manufacturing: the time between generation and consumption is measured not on the order of months, weeks, days, or even hours, but in milliseconds. It travels at 300,000 km per second/186,000 miles per second [1], which is incomprehensibly fast (it's fast enough to cross the widest span of the continental US, 4,799 km/2,892 miles, over 60 times in a single second).
As such, in order to maintain 50/60 Hz at nominal voltages and amperages (and that frequency is very important - in a 60 Hz system, 59.4 Hz is "we have 5 minutes before the entire grid blows up and plunges us into a 3+ month blackout" level bad [2]), utility companies aren't in the power generation business so much as they are in the predictions business. They have to take everything into account, from the weather to consumer purchasing habits, to determine exactly how much power needs to be generated at any given moment or the whole grid will collapse. That wide area synchronous grids like the UCTE (continental Europe) and Eastern Interconnection (eastern US) are able to operate is a testament to human ingenuity, as these really are quite fragile machines.
All this to say that you can't just "make power free when there's excess and people will magically use it up". Even if we can reliably predict excesses, people won't really be able to take advantage of it unless their schedule lends itself to it (e.g. being able to plug their car into a charger at work that only turns on when there's an excess), but even then, when done on a large scale, this only makes the predictions even harder for utility companies to plan around (think: thundering herd problem but on a national scale). At a macro level, it just doesn't work.
[0]: https://www.cnn.com/2023/07/04/business/sheetz-july-fourth-gas/index.html https://www.cnn.com/2023/07/04/business/sheetz-july-fourth-g...
[1]: https://en.wikipedia.org/wiki/Speed_of_electricity?useskin=vector#Electromagnetic_waves_in_circuits https://en.wikipedia.org/wiki/Speed_of_electricity?useskin=v...
[2]: https://www.houstonpublicmedia.org/articles/news/energy-environment/2021/02/24/392290/texas-power-grid-was-4-minutes-and-37-seconds-away-from-collapsing-heres-how-it-happened/ https://www.houstonpublicmedia.org/articles/news/energy-envi...
- alex_young 3y agoCurtailment from solar is a real thing that happens predictably: https://pv-magazine-usa.com/2023/10/31/california-is-curtailing-more-solar-power-than-ever-before/ https://pv-magazine-usa.com/2023/10/31/california-is-curtail...
- _ph_ 3y agoWith electric cars, demand is that simple. They can be hooked up to the grid and charge whenever the prices drop down to zero. This is already done. And then it is an individual decisions to either just use the charge to drive or even put part of it back into the grid or for you own home usage, when the grid prices are high. And more and more houses start to have battery storage.