Assuming a 40 year lifetime in an environment of rapidly declining costs (for the competition) is dubious. Note that we're already seeing some nuclear plants in the US that can't compete just based on their operating costs. TMI's remaining reactor was cash flow negative for six years before it was shut down. It still worked fine, it just didn't make sense to keep running it. Plenty of other plants are cash flow positive but aren't selling their output at a price that would justify their replacement with a new NPP.

One can view the strong position of gas-fired generating capacity in the US not just as a consequence of the low cost of natural gas, but also the low capital cost, which limits the downside risk of future competition. A combined cycle power plant might cost 10% or less the capital cost of a NPP, per unit power output.

Nuclear today is competing against cheap, dispatchable natural gas. That's not relevant to a model of a carbon-free grid, where nuclear could displace the higher costs required to get dispatchable power from wind/solar.

Nuclear's main cost is capital, so using an artificially low lifetime biases heavily against nuclear. The model accounts for operating costs separately. Whether other tech will make nuclear uncompetitive is what we're trying to find out with this model; if we start by assuming that, and limit nuclear lifetime accordingly, then we're making a circular argument.

I did notice that fixed O&M costs are expressed as a percentage of capital cost, so if I take them at face value and halve the capital cost, to be conservative perhaps I should double the O&M from 3% to 6%, which I hadn't done before. That doesn't change the German and Thai results, since for those I only changed lifetime. But in the US, it means capital cost of 3000 still results in a grid without nuclear.

But is that number correct? Not quite. Across the nuclear fleet, the average US nuclear O&M cost is $132/kWh, or 125 euros: https://www.epa.gov/sites/default/files/2019-03/documents/at...

That's only 2.1% of 6000. If we lower capital cost without lowering O&M, we get 4.2% of 3000. That puts a small amount of nuclear back in the US market.

However, they default to a discount rate of 10 for nuclear, and only 5 for wind/solar/battery. I don't see any reason to use different numbers here. Setting nuclear to 5, with the above changes, gets us back to a 100% nuclear US grid. Even if we take O&M back up to 6%, a capital cost of 3000 and discount rate of 5% means a 100% nuclear US.

Comparing overall grid cost of two examples: With no nuclear (due to 10% discount), the US average grid cost is 53.7 EUR/MWh. With 100% nuclear (due to 5% discount and 3000 capital cost), the US grid cost is 48.6 EUR/MWh.

In any case, this has changed my view somewhat. I'd thought that nuclear was a clear winner over storage, but it looks like nuclear's place is at least marginal in the US. We have copious wind and solar, and nuclear only lowers overall grid cost if we can manage the sort of nuclear costs they've achieved in South Korea. But in countries less favorable to wind/solar, nuclear dominates.

Yes. I've been saying that for nuclear to survive in the US, up front costs have to be cut at least by a factor of 3. This is going to be tough. NuScale's SMR doesn't appear to be able to do this.

If we start seeing CO2 taxes here, the way forward will be displacing gas using increased renewables, and if there are enough times where gas goes to zero then adding storage to serve that. Getting to 100% will require additional storage (and likely hydrogen), but even before that the environment will become quite hostile to new NPPs.