[msandford]

I think the idea is that 10GW of solar only produces on average as much power as a 2.2GW "normal" power plant. So during the best parts of the day you're getting all 10GW. At night you're getting 0GW. On cloudy days you're getting less than 10GW. If you average it all out over a year you get about the same amount of total energy output from the 10GW solar plant as you would a 2.2GW traditional power plant.

So once you do that it goes from $1/watt to more like $4.50/watt. And it starts to look about the same as nuclear if in fact you can spend $5b and you get 1GW.

One nice thing about nuclear is that it's super steady 1GW 24/7 vs the solar needs to have battery storage too, and those batteries aren't free.

[bobthepanda]

This project has battery included as well though.

Given that solar's cost trend has been going ever cheaper and nuclear's has done pretty much the opposite, solar will probably win out long term.

[msandford]

Oh that's my mistake then! I read the article looking for that but I didn't see it. I guess I didn't read it thoroughly enough.

I do agree that over the course of the next 10-50 years renewables + storage is likely to be cheaper.

What I'm not sure about is if "the grid" will be cheaper or if we'll end up seeing the renewables self-ownership spiral that causes folks to disconnect from the grid, driving prices up for everyone else on the grid, thus causing more people to disconnect, etc. When renewables are cheap and your house's usage is fairly well known (see years of smart meter historical data) then at some point it becomes economically feasible to be your own grid.

[bobthepanda]

This isn't the worst outcome for consumers, but it is pretty bad for established utilities.

One interesting case study is Puerto Rico, which now has a lot of microgrids post-Maria because PREPA is not trusted and suffering from mismanagement.

[Symmetry]

With today's low interest rates is nuclear actually getting more expensive? Traditionally all the interest on the loans that builds up between when the approval process starts and when construction finishes was the big driver of nuclear costs but since 2009 commercial loan rates are a lot lower than they were beforehand.

[bobthepanda]

In general extending project lead times is going to cost more money because you still need to pay some people to be working for that longer duration.

[bildung]

The Capacity Factor (essentially a plant's uptime) is important to consider, but isn't 100% for the other plant types either, though. Nuclear has 93% in the US, coal has 54%, and Solar PV 26%[0]. The 22% quoted by hokkos is probably specifically for PV plants in Australia.

[0] https://en.wikipedia.org/wiki/Capacity_factor

[erentz]

Plus the point about lifespan, the nuclear plant doesn’t need to be replaced in 25 years. You get twice the lifespan for the same cost.

[natmaka]

Nuke decommission costs and impacts aren't neglectable, and may be underestimated (the UK case showed it clearly).

[lostlogin]

Would I be right in thinking that replacing parts of a solar plant could be done without a huge impact on production? Eg replacing a reactor means it’s offline. Replacing all the panels means a percentage of the plant is offline at a given time.

[nl]

Yes, and panels are getting cheaper continually.

Once the tax effectiveness of being able to depreciate the panels gets taken into account it's a pretty nice investment.

[fomine3]

From Japan, I wish the old plant was replaced after 30 years...

[EdwardDiego]

That said, there's large portions of Australia that are very lucky to get a cloudy day...

[coverj]

thank you for the explanation!