[AnthonyMouse]

The problem with operating a natural gas plant only 0.1% of the time is that you have to cover its fixed costs over whatever time it is run, and if that's only 0.1% of the time then the fixed contribution per kWh becomes enormous. Worse, people like to point out that a high proportion of the cost of a natural gas plant is fuel, but the fuel cost also includes fixed costs. If you're using only 0.1% as much natural gas at scale then you have to recover the costs of all the pipelines and other infrastructure over 0.1% as much sales volume.

You end up paying a significant fraction of the cost of having the generating plants producing power 100% of the time, but only get power 0.1% of the time.

The main advantage of not running them all the time is that then you're not emitting CO2, but nuclear plants have that advantage even when you do run them all the time.

[notTooFarGone]

How is "have to run nuclear all the time" a benefit with solar?

If solar blasts through the day you are unprofitable and have to deal with extra excess power.

Maintaining gas power plants is something that can be shared by the grid and is 100% cheaper than building new nuclear plants.

[AnthonyMouse]

If you build so much solar that rates are consistently zero during the day then solar is unprofitable. But if rates are normally non-zero during the day then generating power generates revenue -- and then your alternative generation method gives you the additional benefit of needing less solar, which isn't free even if it's cheap.

And, of course, the main benefit is at night, because solar is cheap but solar + storage is significantly more expensive, so you get to generate all night -- and get the higher rates from generating at night -- without emitting CO2.

[notTooFarGone]

The business plan then says something like "we run nuclear 24/7 and in the pasts it's been ok from a cost perspective. Now we actually only get money half of the time but trust me this is economically viable"?

With batteries there is no "too much solar", only too few batteries because they compliment each other so good and batteries are now at a price where adding it to solar is a no brainer economically. Building nuclear now would take 10+ years and then we proabably don't need it anymore. Why take the risk if the new state of the art works?

[AnthonyMouse]

You're assuming they would only get half as much money, but you're not considering how generation volatility affects prices.

Suppose you have 1000 MW (constant) of nuclear and 1000 MW (daytime average) of solar. Therefore on the average day you're getting 2 GW total. The daytime price on that day is $0.04/kWh, which is just enough to make solar viable. If every day is like this, solar is doing fine.

But then there's a day when it's extra sunny. Solar is generating not 1000 MW but 1800 MW. Is that good for solar's profitability? No, it's bad, because that means there is oversupply and the price per kWh is zero. Nobody is making any money that day. Solar generated 1800 MW for 12 hours and got zero return. Solar's average is now down to $0.014/kWh. That's below sustainability. Oops. Nuclear also got zero return that day, but only generated at 1000 MW, so its average wasn't negatively affected by as much.

Then, another day, it's extra cloudy. Solar only generates at 300 MW. It's a supply emergency and the wholesale price per kWh rises to $0.28/kWh. Finally everyone is given an opportunity to bring up their average. So solar generates 300 MW for 12 hours and nuclear generates 1000 MW for the same 12 hours and they each get $0.28/kWh. At this point solar's average is back up to $0.04/kWh, which is its breakeven. Meanwhile the nuclear plant's average -- during only the daytime hours -- is $0.1067/kWh.

Then you have the nighttime hours. To play here solar needs storage. Storage is something like $0.14-$0.50/kWh by itself. If you charge it with solar, you're up to $0.18+/kWh. But that's on average again. On the day it was extra sunny, the batteries were already completely full, so that night the price didn't stay at zero and the nuclear plant made some money. On the day it was extra cloudy, the batteries got low, and then the nighttime price wasn't just $0.18/kWh, it was much higher.

Then, once in a while, it's cloudy for a whole week. The batteries aren't just low but completely drained, even before sunset. Solar + batteries can't address this case at all because an entire week's worth of battery storage is prohibitively expensive for something that only happens once a year or so. The nighttime price that week -- because of the volatility created by solar -- goes through the roof. Nuclear plants gets all of that money while the solar plants get none of it, because they're the only available source of electricity.

The economics aren't the same as before because now the prices fluctuate all over the place. But that only means that a generation source that supplies power all the time can make up for the times when rates are low because there is oversupply during the times when the rates are much higher because there is undersupply. Whereas the power source causing the supply fluctuations can't, because its ability to supply power inversely correlates with the price.

[guerby]

"Storage is something like $0.14-$0.50/kWh by itself"

LFP at the cell level is below $100/kWh, 5000 cycles means $0.02/kWh out of the cell (maximum since after 5000 cycles battery is still at 80% capacity so the real price is even lower).

Current LFP cell price look more like $50/kWh these days and still going down...

Also about near $0 or below $0 prices, they exist only because we don't currently signal those prices to consumers. Anyone with electric car (parked 95% of the time) or batteries at home or business would charge them at $0.01-0.05/kWh no question asked and price would never go to zero or negative.

[AnthonyMouse]

> LFP at the cell level is below $100/kWh, 5000 cycles means $0.02/kWh out of the cell (maximum since after 5000 cycles battery is still at 80% capacity so the real price is even lower).

The cells are by far not the only cost. You also have land, installation, operations, taxes, insurance, etc. Most of these costs also have to be paid upfront even though you won't actually hit 5000 cycles for more than a decade, so then you have to add time value of money.

It's like saying "it's really cheap to own a home, just get lumber in bulk for like $100/unit and it'll last for decades."

> Also about near $0 or below $0 prices, they exist only because we don't currently signal those prices to consumers. Anyone with electric car (parked 95% of the time) or batteries at home or business would charge them at $0.01-0.05/kWh no question asked and price would never go to zero or negative.

That's what happens for the first few hours of peak generation, but then the batteries are all fully charged and you're still generating in excess of demand because it's still a week with clear skies in the summertime.

Also, a lot of reason for the zero or negative prices is that you have more generation at that time than there is transmission capacity in the grid. You can't sell it to someone because you don't have the capacity to deliver it, and expanding the grid for something that only happens a fraction of the time (and is only relevant when the price is low) isn't cost effective.

And actual zero isn't really the point anyway. If the wholesale price is $0.01/kWh instead of actual zero on the day when solar is at peak generation, it's still selling below cost on the days when it has the most to sell, and only getting a fraction of the premium during supply shortfalls because the shortfalls are caused by solar being unable to supply more power then.