Hacker News Reader

Comment by m2fkxy | Hacker News Reader

> In conclusion, (...) nuclear reactors are simply not going to be built because they are not competitive

What do you mean? 50+ reactors are under construction as we write [1], and almost twice as much are planned [2] (with even more proposed).

Furthermore:

> A nuclear reactor cannot be shipped

Marine nuclear reactors (of which 200-ish are in active service) and Chukotka's floating nuclear power plants are notable counter-examples. Especially with marine reactors being the baseline for the philosophy behind SMRs.

[1] https://pris.iaea.org/pris/worldstatistics/underconstruction...

[2] https://world-nuclear.org/information-library/current-and-fu...

View on HN

That table doesn't support your thesis, specially when you think of the one reactor being built in France.

Something being built means little when the time from "start" and "end" is meassured in decades. A better table would be "increase in nuclear output". E.g. fully self driving vehicles are being built, but little or none see the street, much less have a positive ROI.

Planned and proposed is even worse. A lot of them would never leave that status. Only energy super eager countries like China or India are effectively increasing nuclear capacity, and in a way that nobody knows if they are financially sound investments.

>Something being built means little when the time from "start" and "end" is meassured in decades.

If you'd go to check for the average build time of a nuclear reactor or the new reactors that come into service you'd notice they're often/typically connected and running 5-6 years after they start building. A lot of that is helped by china and the likes not having horror stories like Finland.

It's also helped by China starting the construction clock after concrete and steel are already in the ground. You can get good numbers by cheating.

The capacity of plants built now is about the same as it's being retired. This will only accelerate in the future as old plants get more and more decommissioned. I'm comparison there is almost 300GW of capacity in renewables added every year. All nuclear capacity being built now for the next decades is less than what's added in renewables every year, and that's even considering the capacity factor.

So they are not being built in any significant way. Percentage of nuclear went down from 20 to 10 percent and it will continue to go down, until it becomes economical or we figure out we can't get renewables to work. I personally wouldn't bet on either of those, but that's just my opinion. There is a third option of course that is very realistic, and that is that the nuclear lobby buys itself a business model through FUD and other means. Currently they realistically don't have one.

If renewables can't do it, they can at least reach a state where the residual demand is utterly unsuitable to nuclear. If renewables are maxed out there will be long periods where they are supplying all demand; the residue has zero base load.

Plants planned or under construction will be cancelled anywhere they are not massively subsidized by government coercion of tax- or ratepayers, as energy prices continue downward. They will be recognized as unable to compete.

Existing plants will be mothballed as energy prices drop below the level needed to continue operating, again except where coerced.

Nuclear power is the cheapest option:

https://www.oecd-nea.org/lcoe/

And it is far more reliable than renewables.

With enough thumbs on the scale, you can support any conclusion. In this case, neglecting all public subsidies and neglecting CAPEX.

It is furthermore a snapshot. Policy for construction needs to consider the future, and renewables cost is still falling fast.

That's for "LTO": extending the lifespan of existing nuclear plants. The cost of power from new construction nuclear would be much higher.

The LTO is the estimated lifespan of the plant. The LTOs provided in that chart are well below the real lifespan of these plants.

The capital cost of building the plants is also provided in the chart. That means this accounts for "the cost of power from new construction nuclear" you refer to.

LTO is referring to refurbishment efforts to extend the lifespan of an existing plant.

Look at the first line in that chart. It's for 20 years of operation of a 1000 MW reactor. The capital cost is listed as $5.93/MWh. If that reactor is operating at 90% capacity factor for 20 years, the total capital cost comes in at less than $1B. That cannot possibly be the cost of building a new nuclear plant. Instead, it's the capital cost of the refurbishment needed to allow an existing reactor to operate for another 20 years. This can surely be a nice way to get more from many existing NPPs (although some of these refurbishment efforts have gone disastrously wrong; see Crystal River 3 in Florida for example), but it has little relevance to the cost of building a new NPP.

Yes, you're correct. Your interpretation is the only one where the numbers make sense.

Can you share sources? I'm surprised that so many reactors are being built.

Yes, sorry for not having done so. Edited my comment.

On reactors under construction, IAEA's PRIS database. https://pris.iaea.org/pris/worldstatistics/underconstruction...

On planned reactors, World Nuclear Association. https://world-nuclear.org/information-library/current-and-fu...

EDIT: wrong link for IAEA's db.

What matters is how many will be completed, not cancelled. That number is rather smaller.

Reply on news.ycombinator.com