Hacker News Reader

Comment by Swizec | Hacker News Reader

If more nuclear takes 20+ years to build while we can get wind farms online next month, is it really faster?

I love nuclear but have to admit it’s dead until we figure out how to build these things faster. You can get 10MW of wind built in 2 months. A nuclear plant gives you 1000MW (1GW). You need to build that in less than 50 months (4 years) to compete with wind. And wind can already deliver incremental value for most of those 50 months so lead times aren’t a problem.

When’s the last time anyone built a nuclear plant in 4 years?

View on HN

> When’s the last time anyone built a nuclear plant in 4 years?

South Korea is one of the world leaders in commercial nuclear power, and as of 20 years ago they were able to build plants in under 5 years[0]. Seems like recent construction has slowed down there as well, for reasons unknown to me.

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

Just comparing MW when wind is blowing is never going to be useful or result in a productive discussion. You need compare nuclear with other useful, base load generation methods in terms of build cost/time/CO2 per MW.

I believe the numbers I quoted are averages, not only when the wind is blowing. Source: https://www.ewea.org/wind-energy-basics/faq/

Here’s a better number: https://www.usgs.gov/faqs/how-many-homes-can-average-wind-tu...

> At a 42% capacity factor (i.e., the average among recently built wind turbines in the United States, per the 2021 edition of the U.S. Department of Energy’s Land-Based Wind Market Report), that average turbine would generate over 843,000 kWh (843MWh) per month—enough for more than 940 average U.S. homes

For comparison, an actual nuclear reactor in practical use produced 4,697,675 MWh in 2017. That’s the equivalent of 5572 of those wind turbines.

That’s a lot of wind turbines. But if we can build 278 wind turbines per year and it takes 20 years to build a new reactor. They’re even.

USA currently builds 3000 turbines per year. https://www.usgs.gov/faqs/how-many-wind-turbines-are-install...

So I think even with 42% capacity, wind is winning. We’d need to build 10 reactors per 20 years (1 every 2 years) for nuclear to beat wind.

It doesn’t make much sense to compare the average outputs of wind vs nuclear unless you budget for a bunch of storage too.

> It doesn’t make much sense to compare the average outputs of wind vs nuclear unless you budget for a bunch of storage too.

It depends. If you're adding wind or nuclear to an already existing grid (which is usually the case), the already existing power plants can usually take over whenever the new power plants are offline or generating less power, and you don't need to budget for any extra storage. In these cases, the average output is the most relevant metric; it shows how much fuel (for a mostly fossil fuel grid) or water (for a mostly hydroelectric grid) it will save.

Possibly, but the discussion is about price per MW of nuclear vs wind, and I'm saying that's not a great comparison between generation types if you want power delivered all the time. 50% capacity is worth much less than 50% of the cost of 100% capacity.

A study was undertaken for the Australian grid.

https://reneweconomy.com.au/a-near-100pct-renewable-grid-for...

It basically found that you can get close to 100% relatively easily with minimal storage (5 hours average demand).

The issue with getting over the mid-90s and to handle 100% is that you need to cater for the edge-cases (conditions that may occur 1-2 times per year max).

Either way from memory, LCOE (in $/MWH) for grid-scale solar is around 50-60, adding storage brings costs up to 80-120. Nuclear is in the range of 200-300+, so even with storage the economics of renewables are better with costs trending downward. Especially when you can build them and scale them up quickly (compared to 10-15 years of nuclear builds).

> so even with storage the economics of renewables are better with costs trending downward

Possibly, but nuclear TCO is also extremely well understood, and it's possible that the TCO for these new systems is not. I'd say the error bars are likely to be much longer on the renewables for the next 20 years or so, as we flush out the details.

Thanks for the more in-depth analysis. It's interesting. But my point - which I didn't really say clearly enough - is: you can't only compare MW, which is what a lot of these calculations tend to do, even if you scale provides for capacity. The fact that nuclear (or any base load provider) works all the time makes it much, much more valuable than something that works intermittently.

> The fact that nuclear (or any base load provider) works all the time

It doesn't; for instance, the nuclear reactors near where I live have to power down for a whole month once every year, for refueling and maintenance. That's around a 92% capacity factor.

That is planned in advance, though.

Wind turbines have additional maintenance downtime too.

Yes and nuclear is much more compact. Wind farms take up lots of space.

It’s a shame we aren’t building more smaller reactors quicker. I think it would solve a lot of the problems that come from treating every plant as a unique one off megaproject.

> If more nuclear takes 20+ years to build while we can get wind farms online next month, is it really faster?

Building (roughly) plants 1 through 3 will be the hardest, take the longest, and cost the most, but after that—assuming you use the same design throughout—you can build quite quickly. Japan averaged ~5 years between beginning of construction and commercial operations:

* https://en.wikipedia.org/wiki/List_of_commercial_nuclear_rea...

France build 58 reactors in 25 years:

* https://en.wikipedia.org/wiki/History_of_France%27s_civil_nu...

It helps if you don't start and stop projects/plants: many cost over-runs are because of wishy-washy decision making.

> When’s the last time anyone built a nuclear plant in 4 years?

2004 to 2009 is five, Tomari 3 in Japan:

* https://en.wikipedia.org/wiki/Tomari_Nuclear_Power_Plant

It's not dead. It shares skills and industrial capacity with the military. For countries that have nukes or want an option to build some in a hurry, they want nukes.

The sheer expense just means it's vastly too expensive to do anything about climate change.

The military industrial complex will still try to sell it as better or at least equivalent to solar and wind though - and people will buy it.

Yep, agreed, it's pretty much dead, unless the SMR people can figure out the low-end. There's a reason why our anti-nuclear friend directed his vitriol there, lol.

That said, I hear that China is building nuclear. We might be able to import the know-how, or develop the political will to spin ours back up. A bit of national jealously goes a long way.

Reply on news.ycombinator.com