The capex is so high and fuel cost so low it just also doesn’t make economic sense to ever turn down your reactor. Or you have to be overproducing to the extent that there is a significant negative price for electricity for quite a while before the cost makes sense.
"French utility EDF began making its nuclear plants more “maneuverable” in the 1980s, and today it says a 1,300-MW reactor can increase or decrease its output by 900 MW within about 30 minutes."
> The capex is so high and fuel cost so low it just also doesn’t make economic sense to ever turn down your reactor. Or you have to be overproducing to the extent that there is a significant negative price for electricity for quite a while before the cost makes sense.
Sure, the price of power has to temporarily go negative to be worth it in a market-based system. I don't know about "significant" though. That's not a problem for grid stability, it just means your cost/benefit calculation gets more complicated.
If you shut down or reduce power at all, you essentially have to wait two days to turn back on.
There’s a decay chain of isotopes that only stick around a couple of days which are normally burned off by neutrons when the reactor is running which build up when you turn the reactor off. They eat up neutrons making the rector much harder to control until they decay naturally. In order to restart the reactor you essentially have to double the output for a while and slowly very carefully reduce it at just the right rate… and if you make mistakes the reactor melts down.
This xenon poisoning was part of the Chernobyl disaster where operator actions and mistakes causing it and then their responses and mechanical failures led to the meltdown.
Rectors are just not designed to do this because it’s too risky and complicated to be worth it. So when you stop the reactor it just stays stopped for a couple of days. Similar for any throttling down which I understand is done carefully in some circumstances but absolutely not continually to respond to demand.