This matches observed data, eg. Areva EPR plant cost estimates have (amid major media controversy) risen from ~3 to ~9 billion € as construction has progressed.

There's one ongoing nuclear power plant construction project in Finland, and the current estimate is at around €8.5 billion.

https://en.wikipedia.org/wiki/Olkiluoto_Nuclear_Power_Plant#...

I wouldn't want a plant from the 60s, and I would wonder how costs were hidden back then, and how much subsequent safety overhauls cost.

> I wouldn't want a plant from the 60s

Me neither. Unfortunately, that's all we have, since we're not building any more :(

There are 4 new AP1000 reactors under construction in the United States. Unfortunately, they're over budget and behind schedule. UCS is if anything charitable to estimate $8 billion for a new reactor.

http://www.enr.com/articles/41134-turnaround-eludes-nuclear-...

Last spring, at the Plant Vogtle construction site near Waynesboro, Ga., executives representing Georgia Power, its lead contractors and trades groups gathered to commit publicly to turning around the $16-billion nuclear power plant expansion. At least three years behind schedule and billions of dollars over its original budget, the massive effort—the first U.S. project of its kind in decades—is a closely watched bellwether of the country’s nuclear plans, which could fizzle if this project fails to deliver, many believe.

Executives with Westinghouse and Fluor, which represent the contracting team that took over on Jan. 1, publicly vowed to do what they already had contractually agreed to do: deliver the project according to schedule by achieving the in-service target dates of June 2019 and June 2020 for Vogtle Units 3 and 4, respectively.

Now, nearly a year into Westinghouse and Fluor’s Vogtle project takeover from former contractor CB&I, the site’s workforce has swelled, but productivity and schedule have actually slipped, say state construction monitors. In joint testimony, PSC analyst Stephen Roetger and construction monitor William Jacobs noted, “The contractor has failed to achieve the project’s critical milestones in accordance with the January 2016 Integrated Project Schedule, which reconfirmed the June 2019 and June 2020 Commercial Operation Dates.” Jacobs, the state’s construction monitor, holds a doctorate in nuclear engineering and has more than 12 years of nuclear-plant construction and start-up experience.

AP1000s are also behind schedule in China. Whatever problems are causing the schedule/budget misses, they seem to be coming from inside the nuclear industry itself.

Meanwhile in Finland the cost overruns are several billion and the contractors are suing each other for billions. https://en.m.wikipedia.org/wiki/Olkiluoto_Nuclear_Power_Plan...

That design seemed so promising. The first plant, in China, was supposed to come on line in 2013. Current date is mid-2017. The original structural design was revised for better earthquake resistance. Then there was some big problem with the coolant pump manufacturer.

I am a tradesman in the metal fabrication industry, and all these cost overruns and missed milestones makes me wonder...

Have we lost something? Have tradespeople lost the ability to deliver these massive projects on time and on budget? Their managers? Which layer(s) is/are dropping the ball.

The structural steel fabrication has become increasingly complex. Our workshop has become dependant on robots (a laser cutter and a 20 meter drill-line machine, and a small fleet of button clicking humans to do all the CAD modelling) that a breakdown can bring fabrication to a rapid grinding halt.

It makes me shudder a bit to imagine the technical complexity of the design and fabrication of a modern mega-project.

At least wind turbines are relatively simple in comparison.

I think the tolerance for imperfection has gone down considerably since, for instance, the 1930s and 40s, when we were throwing up skyscrapers in a few years and churning out entire Liberty ships in a week. We've traded some safety for an enormous increase in complexity of construction. Meanwhile, those old steel-frame skyscrapers are still standing...

I worked a few summers in a small, wood-burning power plant. A couple times a year, they would shut down for one or two weeks, and do some heavy maintenance on different systems. One of the times I was there, they replaced all of the tubes for making steam in the boiler. Every single weld joining together the panels of tubes had to be inspected, and xrayed to ensure that there were no gaps and sufficient penetration. Finding people that can weld to that standard is a more difficult task, because it's not easy, and takes skill, and training, and experience. Moreover, those people who have those skills can command higher rates.

With nuclear plants, the tolerance for imperfection is very low. You usually can't fix anything inside the radioactive zone for the life of the plant. Everything in the reactor vessel has to last for many decades in a very hostile environment. Minor problems can cause billions of dollars in costs.

Ft. St. Vrain was a sad example. That was a nice gas-cooled reactor. But they had some leakage in the helium circulator, which was supposed to keep water and helium separate. This resulted in corrosion and eventual reactor shutdown and decommissioning.

The Three Mile Island meltdown was due to a valve problem. Nobody was hurt, but cleaning up the mess cost billions.

That's part of the problem with all these new reactor designs. A good design with 20 years of operating experience is preferable to a new design, where you don't know what's going to give trouble.

Hyman Rickover on reactor design: "An academic reactor or reactor plant almost always has the following basic characteristics: (1) It is simple. (2) It is small. (3) It is cheap. (4) It is light. (5) It can be built very quickly. (6) It is very flexible in purpose. (7) Very little development will be required. It will use off-the-shelf components. (8) The reactor is in the study phase. It is not being built now.

On the other hand a practical reactor can be distinguished by the following characteristics: (1) It is being built now. (2) It is behind schedule. (3) It requires an immense amount of development on apparently trivial items. (4) It is very expensive. (5) It takes a long time to build because of its engineering development problems. (6) It is large. (7) It is heavy. (8) It is complicated."

Realistically, even 60s-era plants are expensive enough to construct that the existing plants are going to continue to operate for as long as possible.

That's exactly how Fukushima happened: "these plants will be so expensive to decommission, we can't afford NOT to keep extending their lifetime!"

Or, "We planned to decommission these plants, but construction on the replacement plants was blocked."

Either way, a major safety decision basically came down to a political decision over finances. And because nuclear power was involved, the downside became huge. Even when nuclear power is designed to be safer, it comes down to the whims of human politics. Why risk making the downside of a poor political decision so massive?

I feel like you're missing why I said this. I was responding to this sentence in the message above mine: "Actual, real-life plants (like from the 60s) have much lower costs than recent estimates have claimed (adjusted for inflation)."

Why were those plants cheap? First a) they probably actually weren't, thanks to hidden costs, and b) they probably weren't nearly as safe.

And "not safe" = "increased likelihood of some really big unexpected hidden costs".