This actually isn’t true: nuclear power was already becoming cheaper than coal and so on, and improvements have been available. The problem is actually regulatory: Starting at around 1970 various reasons have caused the law to make even the same tech to become MUCH more expensive. This was avoidable and some other countries like France managed to make it keep going cheaper than alternative sources. This talks about it in detail. Here’s a graph from the article:
I have seen that post and others like it, and that incorrect view is what I’m arguing against here. The plot you’re posting, cost vs cumulative capacity, shows that learning with cumulative capacity built happened for a while and then stopped being relevant; that doesn’t contradict my point.
Nuclear power in France, when including subsidies, wasn’t exceptionally cheap, and was driven largely by the desire for nuclear weapons to deter a Soviet invasion using small nukes and lots of tanks in Europe and the threat of ICBMs to prevent the US from participating. Which is a thing that was planned.
BTW, the French nuclear utility EDF is building Hinkley Point C, and it’s not particularly cheap.
Inflation doesn’t explain a tenfold increase between ~1967 and ~1975. Regardless, this graph is in 2010$ so inflation doesn’t explain anything. I agree with @Andrew Vlahos that inflation doesn’t explain the trend.
That graph is by construction start. It would be more appropriate to use construction end date, or some activity-weighted year average.
Regardless, this graph is in 2010$ so inflation doesn’t explain anything.
Did you read my post? My main point was that inflation is an average across things that includes cost-reducing technological progress. If you have less such progress than the average, the correct rate is higher.
I did read your post, but how high an inflation do you need to match this graph? If you argued that the inflation for nuclear plants is let’s say 1% or 2% higher than the inflation in general, why not. But a crude estimates gives an inflation of 30% per year to account for the increase shown in this graph. That’s unheard of.
You mention several times a 5% inflation figure. For a 5% inflation figure to explain this graph, it would take more than 40 years when the date of construction start is separated by only 8 years. So unless, within 8 years, the construction duration was increased by 30 years (which would already be concerning, and fuel the “this is a regulation problem” narrative), your explanation doesn’t match the data.
unless, within 8 years, the construction duration was increased by 30 years
Yes, that’s approximately what I’m saying, but the more-delayed plants had extra costs from the delays. It can take a while sometimes.
which would already be concerning, and fuel the “this is a regulation problem” narrative
No, things can get delayed because costs increased, and then costs increase because of the delays. Sometimes, eventually more government/utility money comes in, old investors get wiped out, new investors come in, and things get finished. But more often, the project just gets cancelled.
You are cherry-picking. From the wikipedia page list of commercial nuclear reactors, the duration between 1960 and 1980 ranged from 3 to 15 years. Only one plant (the one you cite) took 40 years to build, and only one other as far as I can tell took more than 30 years, out of the 144 entries of the table.
That chart relies somewhat on outliers, too. But OK, if you’re making a point about costs increasing after Three Mile Island, I agree: new regulations introduced during construction increased costs a lot.
I think those regulations were worthwhile, maybe you disagree, but either way, once the regulations are introduced, builders of later plants know about them and don’t have to make changes during construction. Once compliant plants are built, people know how to deal with them. The actual regulations were about things like redundant control wiring, not things that directly impacted costs nearly as much as changing designs in the middle of construction did.
I don’t care about cost increases during specific past periods where new (reasonable) regulations were introduced, I care about cost increases from then to now—and costs have continued to increase, in multiple countries.
It’s worth noting that the South Korean strategy to build cheap reactors was essentially about bribing the regulators. After they started charging people with corruption they didn’t build cheap reactors anymore.
I wouldn’t be surprised in India used similar methods to build their cheap nuclear plants.
This actually isn’t true: nuclear power was already becoming cheaper than coal and so on, and improvements have been available. The problem is actually regulatory: Starting at around 1970 various reasons have caused the law to make even the same tech to become MUCH more expensive. This was avoidable and some other countries like France managed to make it keep going cheaper than alternative sources. This talks about it in detail. Here’s a graph from the article:
I have seen that post and others like it, and that incorrect view is what I’m arguing against here. The plot you’re posting, cost vs cumulative capacity, shows that learning with cumulative capacity built happened for a while and then stopped being relevant; that doesn’t contradict my point.
Nuclear power in France, when including subsidies, wasn’t exceptionally cheap, and was driven largely by the desire for nuclear weapons to deter a Soviet invasion using small nukes and lots of tanks in Europe and the threat of ICBMs to prevent the US from participating. Which is a thing that was planned.
BTW, the French nuclear utility EDF is building Hinkley Point C, and it’s not particularly cheap.
From the same paper:
That graph is by construction start. It would be more appropriate to use construction end date, or some activity-weighted year average.
Did you read my post? My main point was that inflation is an average across things that includes cost-reducing technological progress. If you have less such progress than the average, the correct rate is higher.
I did read your post, but how high an inflation do you need to match this graph? If you argued that the inflation for nuclear plants is let’s say 1% or 2% higher than the inflation in general, why not. But a crude estimates gives an inflation of 30% per year to account for the increase shown in this graph. That’s unheard of.
You mention several times a 5% inflation figure. For a 5% inflation figure to explain this graph, it would take more than 40 years when the date of construction start is separated by only 8 years. So unless, within 8 years, the construction duration was increased by 30 years (which would already be concerning, and fuel the “this is a regulation problem” narrative), your explanation doesn’t match the data.
Yes, that’s approximately what I’m saying, but the more-delayed plants had extra costs from the delays. It can take a while sometimes.
No, things can get delayed because costs increased, and then costs increase because of the delays. Sometimes, eventually more government/utility money comes in, old investors get wiped out, new investors come in, and things get finished. But more often, the project just gets cancelled.
You are cherry-picking. From the wikipedia page list of commercial nuclear reactors, the duration between 1960 and 1980 ranged from 3 to 15 years. Only one plant (the one you cite) took 40 years to build, and only one other as far as I can tell took more than 30 years, out of the 144 entries of the table.
That chart relies somewhat on outliers, too. But OK, if you’re making a point about costs increasing after Three Mile Island, I agree: new regulations introduced during construction increased costs a lot.
I think those regulations were worthwhile, maybe you disagree, but either way, once the regulations are introduced, builders of later plants know about them and don’t have to make changes during construction. Once compliant plants are built, people know how to deal with them. The actual regulations were about things like redundant control wiring, not things that directly impacted costs nearly as much as changing designs in the middle of construction did.
I don’t care about cost increases during specific past periods where new (reasonable) regulations were introduced, I care about cost increases from then to now—and costs have continued to increase, in multiple countries.
It’s worth noting that the South Korean strategy to build cheap reactors was essentially about bribing the regulators. After they started charging people with corruption they didn’t build cheap reactors anymore.
I wouldn’t be surprised in India used similar methods to build their cheap nuclear plants.