In theory that works as long as you're willing to let the price reflect actual supply and demand even when the difference is very large, e.g. it has been cloudy and still for a couple weeks so the batteries are low and then you get a hot summer day or cold winter night with a lot of demand. No problem, we'll just set the price to "high enough to get people to stop cooling/heating their buildings" and the market will clear. But people aren't going to like that.
It’s not something that’s likely to happen at the retail level, but at the industrial level. Battery farms buying power when the price is low or negative (due to too much wind/solar) and selling when the price is higher (early evening). Aluminum smelters curtailing. Etc.
There is something interesting happening in the retail space, though, called a “virtual power plant.” Worth googling if you’re curious.
Aluminum smelters are something like 4% of global electricity consumption and 60% of them are in China. In general industrial is less than a third of electricity consumption in the US, not even all of that can be curtailed, and that number is only going to go down if we electrify heating and transportation. It's pretty hard to curtail heating and cooling by more than a minor amount. It's easy to do with transportation over the course of hours or days, but not weeks or months.
Batteries work great to let you generate power at noon and use it at dusk. They're not so great at letting you generate power on days with a surplus and then use it later in the year when there is a multi-week or seasonal shortfall.
The demand can be more elastic than you envision. If power is expensive on a given day, electric cars can wait to charge, or even discharge if they aren't going to be used. People can wait to run laundry dryers.
The market will incentivize actors to smooth out before those kinds of restrictions are necessary.
People might not like changing their habits to follow the energy, but they'll probably be pretty happy when the end result is both good for the environment and cheaper overall. At least in my corner of the Midwest, either the sun is shining or the wind is blowing, and often both.
> If power is expensive on a given day, electric cars can wait to charge, or even discharge if they aren't going to be used. People can wait to run laundry dryers.
That buys you days, not weeks.
The smoothing out things also have kind of an ugly failure mode. People set their cars to sell power into the grid if the price is X% above normal, but that prevents it from getting to be 2X% above normal on the first day, and then fewer people choose not to run their dryers. The batteries get exhausted sooner because their own existence prevented the price from going up very much at first, but that's the profit-maximizing strategy because nobody knows exactly how long the shortfall is going to be and the shorter ones are more common. Then the batteries get depleted quickly and when the shortfall lasts for more than a couple of days, you're not only low on battery storage, you now have more people whose cars have a charge gauge pointing to E and they need to get to work in the morning.
> The market will incentivize actors to smooth out before those kinds of restrictions are necessary.
It isn't a regulatory restriction. It's, where are you setting your thermostat if electricity hits $5/kWh today?
> At least in my corner of the Midwest, either the sun is shining or the wind is blowing, and often both.
The problem is that it's occasionally neither and that doesn't have to happen very often to cause a lot of trouble.
> The problem is that it's occasionally neither and that doesn't have to happen very often to cause a lot of trouble.
The odds of it being neither everywhere (grid) for an extended period of time (storage) is astronomically low. You don't build solar plants and windfarms where prolonged periods of non-production are to be expected.
The once in a century black swan event where distributed power production across a continent all goes down at the same time is basically the same as a blackout from damage to the power grid, which you need to be ready to deal with on those timescales anyways.
> The odds of it being neither everywhere (grid) for an extended period of time (storage) is astronomically low.
The problem isn't that production is literally zero, it's that production is non-trivially below average for an extended period of time.
Continent-spanning grids mitigate that to a certain extent, but they're also a) expensive (purposely-idle high capacity very long distance transmission lines), and b) are essentially a scheme to trade frequent localized shortages for less frequent continent-spanning ones.
Making the grid larger to make it more reliable is also inconsistent with incentives. You install the long-distance transmission line because you're supposed to be using it to get power from far away in an emergency but then it turns out that one end of the transmission line has power which is less expensive, e.g. it gets more sunlight or has more favorable regulations or subsidies. Soon your generation is concentrated in the place where it's cheaper to build and the line gets used to average out long-term prices instead of short-term prices, leaving you with the short-term volatility but now it will affect even more people.
Then you're in more trouble. You get a month of unfavorable weather in the place where generation was concentrated, or an extended period of elevated demand at the same time as total supply is on the low side of average, and instead of a shortfall in Massachusetts you now have one in the Eastern Grid containing everything on the New York side of the Mississippi river.
And the US is one of only a few countries that spans something on the scale of a continent to begin with. Is the UK inclined to be reliant on power from the EU? Is Japan supposed to rely on China? How about India and Pakistan or Israel and any of its neighbors?
> You don't build solar plants and windfarms where prolonged periods of non-production are to be expected.
Prolonged periods of low production happen as a result of weather, e.g. Florida is suitable for solar because it's at a favorable latitude, but it also gets a lot of rain, and sometimes the rain continues for several weeks at a time.
Moreover, suppose there are places that are expected to experience prolonged periods of non-production from solar and wind. What are they supposed to do for power?
> The once in a century black swan event where distributed power production across a continent all goes down at the same time is basically the same as a blackout from damage to the power grid, which you need to be ready to deal with on those timescales anyways.
Having the same weather system affect most of a continent at the same time isn't really that uncommon. You also get nasty correlations like cloudy and still weather in the heart of the summer, causing simultaneous low generation and high electricity demand for air conditioning over wide areas.
Compare the likelihood of a prolonged period of low production versus our current system, with a prolonged period of low oil/gas availability.
We already experience exactly the same shortages. Look at how countries are responding to the oil shock right now.
You indicate the problem with localized generation, but the extraction and refinement of our current energy system is far more concentrated than renewables already. Look at https://openinframap.org/#5.62/28.611/114.086/A,B,L,P,S
Solar is being generated all over the world, and your concern is that it's production will be concentrated and subject to supply shocks?
Yes, intermittency will always be a problem, but that is true of all energy systems. Wind/solar/batteries has BETTER intermittency than oil/gas, and the added benefit of not cooking our entire biosphere.
> The problem isn't that production is literally zero, it's that production is non-trivially below average for an extended period of time.
That isn't a problem though for precisely the same reason.
> Continent-spanning grids mitigate that to a certain extent, but they're also a) expensive (purposely-idle high capacity very long distance transmission lines), and b) are essentially a scheme to trade frequent localized shortages for less frequent continent-spanning ones.
We already have the grid. And again continent spanning shortages don't happen.
> Prolonged periods of low production happen as a result of weather, e.g. Florida is suitable for solar because it's at a favorable latitude, but it also gets a lot of rain, and sometimes the rain continues for several weeks at a time.
Suitable means satisfies all requirements, not favorable conditions for some of them. If you're building solar in Florida it is with the expectation that there will be some rainy days.
> Moreover, suppose there are places that are expected to experience prolonged periods of non-production from solar and wind. What are they supposed to do for power?
GRID
> Having the same weather system affect most of a continent at the same time isn't really that uncommon. You also get nasty correlations like cloudy and still weather in the heart of the summer, causing simultaneous low generation and high electricity demand for air conditioning over wide areas.
It is extremely uncommon. There's a difference between a low pressure front moving across a few states and "the wind has stopped everywhere." You might get cloudy conditions in summer, Phoenix doesn't.
Which is why we have ancillary markets. For all parts of the grid that the energy only market does not solve well enough, but we as a society deem necessary to solve.
When these edge cases are found, if they even exist, design a technology neutral market solving it and let companies bid.
