Nuclear power plants are most profitable when run at full capacity. They have massive sunk costs, but the marginal cost is low. At a grid level you certainly can change output based on demand. But the markets may need to be adjusted to keep them profitable through mechanisms like a capacity market. Or by finding new uses for the power such as making hydrogen.

With peaker plants fuel costs are a factor, but also the cost of the machine. An old fashioned reciprocating engine maybe the cheapest as a last resort that is only used when other peakers are already utilized. Fuel efficiency becomes more important as the plant is more utilized.

The markets already adjust. I work for a company that melts iron in one factory. This uses a ton of energy, but we work with the power company and so we get discounts for only melting iron at night (when power is cheap). I'm pretty sure that we shut down for regular maintenance based on input from the power company (either they are shutdown for maintenance themselves, or they are expecting a big load)

> If you want to replace gas peaker plants with nuclear then you need to build up enough nuclear capacity to replace all other intermittent electricity producers.

Why even build all those intermittent electricity producers in the first place then? It sounds like problem searching for a solution.

>Why even build all those intermittent electricity producers in the first place then?

Because they're the cheapest. All else being equal, you want most of your grid to be dirt-cheap, and the last 10% can be crazy expensive and you'll still end up ahead.

Because we have no choice. We ideally need better storage solutions or something like cheaper fission to keep the price of energy down, but we need to avoid a climate disaster even more to keep the price of living down.

> Why even build all those intermittent electricity producers in the first place then?

Cost a new nuclear power plant in the US: $10,000/kW

Cost a combined cycle gas turbine power plant: $1,000/kW

Cost a simple cycle gas turbine power plant: $400/kW

Ignoring the economics for a second, the persepctive in play when someone says "replace gas with nuclear" should be:

1) If solar is deployed, gas/batteries/something has to be deployed as well to deal with nighttime.

2) So the 'replacement' of nuclear is steady power that replaces some solar panels, some gas for a net improvement, technically speaking.

Of course, the main argument against that is whether the economics plays out. I find it humorous that I'm taking the side of technical excellence and the pro-solar people are saying "who cares, lets just follow the economics". There was a decade where those positions were mostly reversed when it was coal v solar.

Also solar is made primarily in China. It's cheap currently but when the inevitable cold war happens with China and the USA those solar panels may not be nearly as cheap or reliably sourced. Sure we can make our own but it would take a decade or more to ramp up.

This run on the assumption that demand is static. Renewables are great at providing cheap energy to consumers that can match their usage when the price is low.

One such use case happen to also be one of the biggest energy consumer, communal heating. If those invested into bigger capacity they could heat the water when prices is low and operate much like a battery for the system at large. The demand for base load would then be reduced and nuclear would have a easier job without displacing the use of cheaper renewables.

Idea: excessive nuclear but we have big datacenters that ramp up power usage during periods of otherwise low usage, and ramp down during peak hours. Ideally shared compute nodes for job batches that can wait up to a day. Plus EV charging can help reduce variance.

Gas plants are also best run all the time in advanced combined cycle plants. Nuclear often in the US already has backing of very large pumped hydro storage facilities. So no, you don’t need enough nuclear to replace all peakers.