I'm curious how the numbers stack up of completed plants - but I am not very good at math and don't have a great understanding of electricity units, especially at the grid scale and big numbers.

Any chance you could help compare the construction cost of this nuclear plant to another recently constructed solar or wind farm measured against... I guess capacity?

Given the intermittent nature of solar/wind, does capacity even make sense to compare in a context without supporting batteries?

I'll give it a shot but I am probably super wrong.

** Nuclear:

I'll use the plant from the article https://en.wikipedia.org/wiki/Vogtle_Electric_Generating_Pla...

Construction costs 18b USD (does that include loans?)

Nameplate capacity of 2302MW

Used capacity is 91% so 2094MW

$18b / nameplate capacity = $7.8 USD per rated W

$18b / used capacity = $8.60 per realized W

** Solar (excluding batteries):

I picked a relatively large, recent, US based solar farm from the list of plants in wikipedia

Agua Caliente Solar Project (2016) https://en.wikipedia.org/wiki/Agua_Caliente_Solar_Project

Construction costs 1.8b USD

Nameplate capacity of 290 MW

Used capacity is 28% so 81 MW

$1.8b / nameplate capacity = $6.2 USD per rated W

$1.8b / used capacity = $22 USD per realized W (that can't be right?)

** Note:

I don't know if my math is right, I don't know if the costs factor in loans, also the nameplate capacity for the nuclear plant is MWe and the solar plant is MWac so I am unsure how that works out.

> I picked a relatively large, recent, US based solar farm from the list of plants in wikipedia

> Agua Caliente Solar Project (2016)

Note that solar being cheaper than nuclear is a more recent phenomenon than 2016. The solar panel prices went from $0.63 to $0.26 in the time span between 2016 and 2022.

https://ourworldindata.org/grapher/solar-pv-prices?time=2002...

A better example is Spotsylvania Solar/Highlander Solar: https://www.sheppardmullin.com/assets/htmldocuments/PFI%2020...

Construction costs: $905m USD

Nameplate capacity: 618 MW

I couldn't find used capacity factors but Yuma is one of the sunniest counties in the USA while Spotsylvania county is further north and also has less sunny days. With an assumed capacity factor of 18%, one gets 111 MW.

$905m / nameplate capacity = $1.48 USD per rated W

$905m / used capacity = $8.24 USD per realized W

This year alone prices for solar panels have dropped by 30-40%. The biggest reason for prices not dropping for solar farms/large scale solar at the moment is that the supply of other equipment needed like transformers etc can't meet the demand with delivery times having reached 2+ years. So the prices for solar will likely drop another 40%-50% in the next 2-3 years at least in the rest of the world maybe not in the US because of trade wars/restrictions.

Using Solar Star as another datapoint [1]. 579 MWac x 32.8% capacity factor ≈ 190.

I found mention of a bond issuance and someone purchasing the project here [2]. If it’s $1b, then it’s $5.26. If $2b then $10.53.

So they’re in the same ballpark. But one type of plant runs 20-30 years and the other for 50-80 years @ 90% capacity factor. The CANDU reactors are especially cool in that they can use natural uranium and refueled without a shutdown [3].

—- [1] - https://en.m.wikipedia.org/wiki/Solar_Star [2] - https://www.sustainablebusiness.com/2013/06/1-billion-bond-o... [3] - https://energyeducation.ca/encyclopedia/On-line_refueling_of...

One thing you didn't include: Opex, a solar farm is pretty low maintenance and requires zero fuel. A NPP is the opposite.

That's fair, I wonder if there are numbers on running costs.

I'm certain nuclear running costs would dwarf solar - I think solar just needs fresh water, cleaning and hardware maintenance (replacing inverters, and such).

Would be interesting to work the running costs into the "$ per realized W" calculation.

I'd also like to see how battery-backed solar compares. I assume the objective would be to solve the intermittency issue, but I am hopeful it would increase the capacity factor as well.

Another thing that's interesting to consider is multi-purpose energy utilization you get with nuclear - like desalination and hydrogen generation - though the latter is uneconomical because hydrogen produced from fossil fuels is much cheaper.

Had to dig in my comments, but there you go:

https://news.ycombinator.com/item?id=38303819

Self citing:

Better not use biased opinion pieces, when there numbers from government sources (US, but eho cares):

LCOE (total, incl. CAPEX, in USD per MWh):

coal 82.6, combined cycle 39.9, advanced nuclear 81.7, geothermal 37.6, biomass 90.1, onshore wind 40, offshore wind (that one was a surprise, since offshore wind should be quite cheap, mainly driven by capital cost of 104 USD per MWh) 105, solar 33.8, solar hybrid 49 and hydro 64.

Variable cost (same as above):

coal 23.7, combined cycle 27.7, adv. nuclear 10.3, geothermal 1.2, biomass 30, onshore wind 0, offshore wind 0, solar 0, solar hybrid 0, hydro 4.1

All number from here:

https://www.eia.gov/outlooks/aeo/pdf/electricity_generation...., page 9.

Does the solar figure include the batteries?

It is very sad that nuclear is actually cheaper than coal.

Hey thanks for adding links to formulas. Great use of wolfram alpha imo.

>That's missing the huge and expensive nuclear power plant around that kilogram of uranium.

Whereas I think you are missing the huge and expensive battery array for solar to be useful outside of peak times? And the fact that the best sites for solar are far away from the transmission network?

You dropped a decimal point. The 1 kg of sunlight at 15% efficiency produces 13.5 petajoules of electricity.