Western industry has progressively lost their ability to efficiently manage many types of heavy industry and large construction projects. This is particularly true for nuclear, which requires the application of disparate and highly specialized techniques and skillsets that are quickly lost. Though not to the same extent as the U.S., Japan, France, and South Korea have lost many of the same skills even though only a couple of decades ago they could churn out new nuclear plants with breakneck speed. (The only thing more astonishing than the efficiency of the French and Korean nuclear industries was how quickly they died.)

Reviving the nuclear industry will be painful and costly, but the same is true for reviving and rehabilitating other heavy industries that have withered away, if and when these countries ever choose to do so. The justifications for resolving ourselves to their death are hypocritical. Both solar and wind power manufacturing industries, not to mention the automobile industry, would be non-existent in the U.S. without heavy subsidies. For example, special trade zones that incentive partial manufacturing of certain wind generator components in the U.S., or in the case of automobiles flat-out quotas. But even then the future of these industries in the U.S. remain tenuous. Fortunately for consumers (but less so for semi-skilled and skilled factory labor), solar panels, wind generators, and cars are easily imported; nuclear plants not so much, not unless the modular reactor designs take off.[1]

[1] Interestingly, the skillsets, including project management techniques, for these industries can't be imported, either. Toshiba bought Westinghouse because they thought they could use their superior heavy industry skills to rehabilitate Westinghouse and the American nuclear industry. But heavy industry skillsets are especially tuned to local industrial supply chains and labor pools, as well as domestic regulatory frameworks. Toshiba couldn't do squat with Westinghouse; it was deadweight from day 1.

To some extent, it turns out that the problem is France never actually had those specialized nuclear techniques and skillsets. For example, they had huge problems with defective welding and large forgings that weren't up to spec and had faked QA testing documentation which caused big delays. Then the regulators went back and looked at the old plants from the original wave of nuclear building, and they had the same problems. It's just that French nuclear wasn't so well scrutinized back then, possibly for political reasons.

I suspect this is true of many industries, including aircraft and car manufacturing. And I suspect the quality issues in the French nuclear industry were, very broadly speaking (trying not to equivocate), similar to issues in nuclear industries elsewhere.[1] As quality control and analysis become better, you're often going to belatedly realize that earlier claims fell short. As a rule of thumb I always add a fudge factor to any sales pitch. For example, whatever the 2020 nuclear industry claims today, let's just assume what they can deliver is what they were selling in 2000. And I'm comfortable with what they were selling circa 2000 in terms of safety, as I suspect many others were, too. Nuclear wasn't "unsafe" in 2000, neither in reality nor on paper. With the benefit of hindsight it's clear it was less safe than claimed, an effect hindsight has on most things. But the evidence regarding nuclear has been clear for quite awhile: safety is very good, with large and ever growing margins for error.

[1] Fukushima being an obvious example. I'm neither an engineer nor nuclear cynic, but I was always skeptical of the claims nuclear engineers were making on the news in their self-defensive and, ultimately, self-defeating attempts to mitigate panic. They were far too credulous of engineering tolerances and quality control despite the evidence piling up in front of them regarding the probability of the supposedly improbable. Anyone remotely familiar with any kind of construction project knows things always fall short in the field.

Westinghouse was all rotten from the beginning, they sold reactors like cigarettes packs and disregarded scientists' safety concerns. Just look at the guy at 15:21.

https://youtu.be/EviEN0ScOwg

So how do you turn a whole unsafe industry into a safe one? That's a lot of organizational DNA to change. Might as well convert a dinosaur into a gazelle.

Dunno, probably act responsibly like the AECL (Canadian company responsible for CANDU reactors) has. Usually it's a management problem. Start by changing the CEO? The nuclear industry, just like aerospace, should be very risk adverse. It's no place for SV or tobacco company mentalities.

I am hopeful about the new crop of companies currently designing SMRs. This was probably the way it should have been done from the onset, rather than pushing single unit power to hundreds of gigawatts.

Interesting stuff you are mentioning here which I don’t normally see brought up. Do you have any links summarizing this information?

My comments are a synthesis of stuff I've read over the years. This might have been one of the many articles I've read: https://www.vox.com/2016/2/29/11132930/nuclear-power-costs-u...

I only began recording in 2018 particularly noteworthy reporting and research pieces (regarding any topic) I stumble upon. I just added the above to my personal bibliography for good measure.

It's always astonishing how hydro and nuclear are so much better handled by SoEs like EDF or HydroQuebec that manage to turn a profit too, while private companies can never do it well.

EDF is the "private company" which is building Hinkley Point C and Sizewell C.

For Hinkley, EDF have assumed the construction risk in return for an exceptionally high, guaranteed "strike price" for the electricity it produces.

But for Sizewell, it sounds like a different finance arrangement will be in place, likely where the UK government/public assumes more of the risk in return for a cheaper per-MWh price.

Private company mostly owned by the French Government.

https://www.edf.fr/en/the-edf-group/dedicated-sections/inves...

The main advantage that EDF has that allows it to work well isn't that effective out of France. They still do good work, but they act much more like a private company than an SoE, which is why it's so much more successful in France.

Well, EDF have not been so successful with their most recent project in France, Flamanville 3.

Construction started in 2007, with estimated completion in 2012 at a cost of €3.3 billion.

As of 2020, the plant is still not complete and the cost has spiralled to €19.1 billion. It is now tentatively planned to go online at the end of 2022.

We saw what happened to Areva building reactors on a fixed cost contract. Today's economic realities do not quite let one plan a 10 year project on a fixed cost. 3.3bn do not equate to 2007 3.3bn in 2020. The Chinese are able to do it because they not a free market economy, they can force people to work at gunpoint.

So, these things are not viable in a market economy? Why should they be pursued then?

They are viable but I'm quite skeptical about accurately estimating the costs for a 10 year project. Concrete, steel, fuel, workforce costs fluctuate from a year to another, let alone 10 years. We'll just have to accept that there are going to be cost overruns.

Comparatively to other nuclear companies and in aggregate, EDF has the best safety record, the cheapest energy when subsidies are taken out, and is solidly profitable. That one or two constructions went overbudget does not matter if the result is still head and shoulders ahead of the fray.

This is basically about the cost of capital for the build and inability to insure for liability and future unforeseen costs.

But nuclear plants are not insured against major accidents. Nobody insures that, for some reason.