I don't think labor costs account for much of the difference. You can pay the salaries of 10,000 US construction workers for 7 years with a single billion. (I seriously doubt they need to hire that many people for one reactor.) And, due to cheap shipping these days, the cost of materials is pretty much the same everywhere.
How about the costs to manufacture the various components? I imagine that even with a standardized design, many components get sourced locally (like pipe and such) and that labor cost factors into the overall price too.
Also, did I mess up my math, or are you saying that hiring an American construction worker costs only $100,000 for 7 years? That's (barely) below minimum wage, and I don't imagine that construction workers earn minimum wage either.
The estimate for the two reactors involves 3500 on-site construction workers [1] with a completion date of 2016 or 2017 [2]. This means 1750 construction workers employed for 4-5 years at each reactor. The average annual wage for a construction laborer is 34,040 [3]. So we're looking at labor costs of about 0.3 billion / reactor. Even if the US workers get paid double the average, and the Chinese close to nothing, labor costs only explain about 10% of the price difference.
Regarding components, anything that's not built on the spot (including pipe) can and will be shipped internationally from the cheapest supplier (particularly if it's expensive). Again, we're looking at price differentials of around 10% rather than 2500%.
Presumably nuclear reactor construction requires more skill than building houses or whatever. And this is for people with absolutely zero prior experience, so not counting experienced people, not counting management, not counting various overhead (payroll taxes, HR, etc., which is often significant).
I would not be surprised if components had a similar story to the workers: nuclear means extremely specialized, and that means the normal rules for commodity resources may not apply. For such a political project, it also would not surprise me if certain things were required to be purchased from US suppliers even if they aren't the lowest bidders.
I don't know exactly where all of the difference comes from, but the situation seems more complicated than simply hiring standard construction workers at average salary and buying commodity components from international suppliers.
Also, you probably want to double the salary figure to get an idea of the total compensation cost to the company doing the hiring. A $70k worker can easily have a compensation package that costs $140k to the company doing the hiring, depending on the exact benefits and what sort of discounts the employer can wring from insurance companies, if most employees have families, if there's relocation assistance, etc. Since there aren't a lot of qualified people to work in the nuclear industry I'd expect the packages to be pretty decent.
The wage of construction workers is $34,000, but what is the cost of employing a construction worker? I'm surprised if it does not double the total cost per worker. That said, your point is still taken that it does not explain the full price difference.
It's $100k per year per construction worker to build a nuclear power plant. Their pay will be $50k just for the base for an average worker.
The security issues (for various stages it requires a lot of background checks), the liability issues (requiring massive insurance outlays), the actual pay, the engineering requirements that the workers must qualify for (just finding the right engineers), any union tangles you run into, and so on. There is an extraordinarily long list of costs per employee to build a nuke plant.
You're looking at upwards of $1 billion per year with 10,000 employees. It's very easy to see where it ends up costing $7 billion to deploy a new plant (it's going to cost 50% more with cost overruns). Even if you were to amazingly slice the worker cost to $50k per year flat fee, with zero added costs per employee, you're still talking $500m per year with 10k workers, and over the build term that's going to be $3.5 billion to $5 billion (7 to 10 years).
Upon closer inspection, they only use 1750 construction workers for 4 years per reactor. By your figures, that's 0.7 billion total. Let's assume the Chinese pay absolutely nothing for nuclear-reactor-qualified workers. Where do the other 4.3 billion per reactor go?
The actual construction cost I recall being cited for this project is closer to $9.5 billion. The remaining $4.5 billion are the interest you have to pay on a somewhat-high-risk $9 billion loan. Oh, and you have to start paying interest when you borrow the money, but don't start having revenues until 4+ years later.
Cost of capital is very different in China, depending on who's funding the project and how. It's hard to compare exactly what the construction cost component of the Chinese numbers is, or whether they're even including the cost of capital at all. I strongly suspect it's not, since the article quoted is talking about the revenue Westinghouse/Shaw will receive, not the expenditures of the purchaser.
So we're really comparing about $4.75 billion in the US to $2 billion in China. If the actual structures are somewhat different, land in the US somewhat more expensive, a lot less in the way of earthmoving needs to be done in one of the locations (say because people aren't worried about environmental impacts), then I can see a 2x difference in the price tag...
