Comment by Manuel_D | Hacker News Reader

Comment by Manuel_D | Hacker News Reader

Parent Full thread

Nuclear power is cheaper when built in serial production. That's the basic conclusion of any analysis of the history of nuclear power construction. There's nuance as to why, but the basic pattern is that putting an order of 40 steam generators [1] is cheaper than a run of just 4 of them. Similar deal with specialized pumps and other nuclear power components. Suppliers can re-use infrastructure and expertise, and get better deals on input materials by guaranteeing a stable demand.

France's nuclear program in the 70s was much more effective than current nuclear projects because of that economy of scale. They build ~50 reactors of only a few types. The US is similar: many of its plants built in the late 1960 and 1970s delivered power at $2-3 billon USD per GW (adjusted for inflation), and some of them under $2 billion per GW. And these aren't equal to other forms of generation: nuclear power's capacity factor is among the highest: https://en.wikipedia.org/wiki/Capacity_factor

1. https://en.wikipedia.org/wiki/Steam_generator_(nuclear_power...

> Nuclear power is cheaper when built in serial production. That's the basic conclusion of any analysis of the history of nuclear power construction. There's nuance as to why, but the basic pattern is that putting an order of 40 steam generators is cheaper than a run of just 4 of them. Similar deal with specialized pumps and other nuclear power components. Suppliers can re-use infrastructure and expertise, and get better deals on input materials by guaranteeing a stable demand.

Unintuitively though, this is incorrect - both for France[1] and the US[2]. Building subsequent versions of the same reactor design increases the cost - instead of it staying the same or going down.

[1] https://www.sciencedirect.com/science/article/abs/pii/S03014...

[2] https://news.mit.edu/2020/reasons-nuclear-overruns-1118

Not coincidentally, that study only decided on 1976 for its analysis. If I order 6 plants of the same design, then drop the next batch down to 4 plants for the next run, then 2 plants for the third run, then I should expect the cost to go up as scale goes down.

Here's a publication that points this out. In fact, it calls out the article you linked specifically: https://www.sciencedirect.com/science/article/pii/S030142151...

See that cluster in the late 60s and early 70s: https://ars.els-cdn.com/content/image/1-s2.0-S03014215163001...

That's the benefits of scale. The slower pace of construction after that is accompanied by higher costs.

Saw this interesting chart recently on the French case. Shows generally decreasing costs between first to last of a generation but then stepwise cost increases when moving to the next generation.

https://twitter.com/autommen/status/1559097869940006916?s=21...

I have worked on large infrastructure projects and this doesn't surprise me at all.

* Repeat projects bring a new set of biases. People see it as an opportunity to "get it right this time".

* People are excited about the new project (more bias).

* The presence of expertise and experience will make contractors want to charge more rather than less.

* New contractors may bid for the project at a low price but lack that hard won experience.

* Efficient manufacturing is a difficult and expensive field in its own right. It will be more complex than building the initial run. And the scale of production may never justify that.

* Many contractors will bring expertise and experience from other fields. But you will also compete for their time with other projects. The guys doing the massive civils contract are sought after by the tunnelling project or the roads project.

* Civil engineering is very dependant on the site and can be very expensive even if it is technically trivial.

* You work in an economy with a high cost of living. Salaries grow.

From the article:

So many orders came in that “[w]ith only two companies building plants, a rapid increase in orders escalated costs for major components and strained the limited supply of qualified labor.”

Plants built in the 1960s and 1970s did exceptionally well on a cost to energy ratio. They were also built in greater numbers than in the 1980s and onwards: https://ars.els-cdn.com/content/image/1-s2.0-S03014215163001...

The source given for that sentence is: Steve Isser, Electricity Restructuring in the United States: Markets and Policy from the 1978 Energy Act to the Present. So whatever that sentence is talking about, it happened after the period of rapid construction and cheap costs.

Reply on news.ycombinator.com