I think this is vastly misunderstood in the (artificial) renewables vs nuclear conversation. We keep building effectively one-off complex machines, and then flushing all that knowledge down the drain by saying it cost too much and took too long. Like yes, the first one always takes longer and costs more...
There's no reason to assume that building the same nuclear reactor design multiple time will maintain the same cost or decrease. Even with the supposedly successful French nuclear program, costs increased over time, there was negative learning:
There's huge huge risk in choosing a particular design and building even one of it, because we don't know if it will be constructible the first time, and we don't know if future builds of the first design will be more or less expensive.
When each build is a $10B roll of the dice with variance of 2-3x of initial estimates, it's a bit difficult to find rational financial backers. Especially when there's not that much profit to be had from even a successful build. The risk reward is completely out of whack compared to the other options for carbon neutral energy.
Of course there is, it's happened exactly the way I said in South Korea, Japan, and American naval reactors. These projects take a long time to complete and there have been relatively few of them. It therefore stands to reason that the cycle of learning from them and making their construction more predictable would take longer than for e.g. cars.
Far too many people are generalizing from the French and American nuclear programs, both of which built lots of reactors in a comparatively short time and then were fear-mongered into a standstill by the fossil fuel lobby.
"""In the third era of nuclear power construction in Japan, from 1980 to 2007, costs remain between ¥250,000/kW and ¥400,000/kW, representing an annual change of −1% to 1%. This period experienced relatively stable costs over 27 years."""
The negative learning rate is a strong signal of interference by the regulators. More than anything else it shows how excessive safety regulations are strangled the industry.
1970s nuclear safety standards, despite it all, were still better than the energy strategy the world adopted from 1970-2020. Killing off nuclear in search of a perfect power system was a stupid strategy, and failed. The only unfortunate point of karmic justice is that Europe ended up reliant on Russian gas and in an energy crisis as a reward for their stubbornness against making the technically obvious choice.
Well done Finland for even managing to get a reactor built in the face of all that.
What did France change about their regulations that increases cost?
Maybe you have some actual knowledge, but I have never found somebody who says that regulations are the problem but has any concrete suggestions for changing regulations. It's just a vague gut feeling. And in the case of France I doubt it applies at all.
Construction is not like manufacturing, it does not see continual productivity improvements like manufacturing does:
As an energy source whose costs are primarily construction related, we would not expect to see it falling in cost over time. We would expect that energy sources whose costs are dominated by manufacturing to outcompete nuclear as time passes.
Ironically, the evidence to date is that'd probably get more people killed due to energy security issues. The last group who tried to apply that logic in defiance of market forces - the Germans - are currently staring at each other wondering what to do about their reliance on Russian gas.
If they'd acted rationally, acknowledged Fukushima and moved on with trying to make their nuclear power cheaper instead letting ideology determine that part of their energy policy then they wouldn't be in the mess they are in. Instead they pushed ahead with renewables, ignored technical problems like reliability and have ended up looking like fools while literally paying a high price for the privilege. France continues to quietly plod away as a reliable European energy success story.
In France the 'nuclear success story' led to a state law (2015-992, from 2015, the "loi relative à la transition énergétique pour la croissance verte") stating that the part of nuke-produced electricity must fall to less than 50% in 2025, from 72% then, and that renewable sources must replace it.
In France nuke-power is backed by gas (which produced 7.5% of the gridpower in 2020).
The sole reactor currently planned (Flamanville-3) is a complete disaster, more than 12 years behind schedule, it will cost at least 19.4 billion € (initial budget: 3.7 billion €).
The other options for carbon neutral energy that does not rely on using fossil fuel as part of the energy strategy are few and far between. The few suggested solutions tend to rely on battery solutions for wind power (at least for countries this far up north).
It would be great to see an attempt to such battery solution that would cover the same amount of capacity as this plant, that can operate for at least several months without recharging, in Finish winter, and cost less than this plant and be built faster. That would check all the boxes, and if such technology already exist, people here should really put their investment money into it.
It's not only the first one. It's at least the first ones plural. The same one, being built in France in Flamanville by the same company was scheduled to be finished in 2012, and is currently planned for 2023 (11 years delay), and with crazy over cost like the Finish one.
I don't think there is any reason to think it will be different for future ones if any. We'll see what happens for the British one (Hinkley Point C), but they already know there will be large delays and cost overrun.
It’s not too big, just too politicized. Bill Gates wanted to do it for a long time, but he can’t get the political buy in.
For me even seeing the Tesla Berlin factory delays makes me sad: in China already the second factory is being built while Germany is coming up with new arguments without looking at the cost/benefit ratio for EU of those delays.
Using someone who's been comfortably in the top 5 wealthiest people for the past two decades, with a 12-digit estimated net worth is perhaps not a good way to illustrate "not too big."
The main reason I was writing it is that there _is_ willingness from the top tech billionaires to invest in modern nuclear plants. They don’t do it because the politicians don’t let them do it, and it’s not only about regulations, they want to keep full control.
So far every attempt at small modular reactors built assembly line style has failed. Currently there's two projects with some momentum: NuScale's pilot plant in Idaho, and the recently announced Rolls Royce project.
Sadly NuScale seems to be struggling to meet timeline and cost targets, and there were some headlines last year about the sponsoring utility pulling out of the project. I'm rooting for them but it doesn't look good.
The RR project is still just on paper really, and is closer to a medium sized reactor than SMR. Maybe they've cracked the code? Maybe not, we'll see.
Every nuclear thread here people chime in with a reflexive "just build nuclear smarter" answer without actually addressing why every prior attempt at this has failed. A lot of people don't want to develop an understanding of the challenges more sophisticated than smugly blaming environmental activists, over regulation, etc.
Look at these projects from the prospective of potential private investment. Anyone doing basic due diligence is going to realize that the odds are stacked against a project hitting it's time and cost goals. Then you look at the trend lines for levelized cost of renewables + storage, and you realize that even in the best case scenario, within 2 decades you may be squeezed out of the market even if the efficiencies of scale and learning kick in.
> ...and then flushing all that knowledge down the drain by saying it cost too much and took too long. Like yes, the first one always takes longer and costs more...
Yup, exactly: And then there is a huge stink about this and no more are built for a good while, so the next one is a first one again. Over and over and over. Lather, rinse, repeat, repeat, repeat, repeat...
