Construction Times for Nuclear Power Plants

I had thought we were pretty much done but the kids in the basement came up with another reason why we can't use nuclear energy. Nukes take too long to build! they whined, their lips quivering. It took ten years to build them before. Jim Hansen at NASA says we only have ten years to stop all the greenhouse gases or we're gonna die! We'll build windmills instead.

Now, how could anyone know all that? The estimable Dr. Hansen actually says

"The Energy Department says that we're going to continue to put more and more CO2 in the atmosphere each year--not just additional CO2 but more than we put in the year before. If we do follow that path, even for another ten years, it guarantees that we will have dramatic climate changes that produce what I would call a different planet--one without sea ice in the Arctic; with worldwide, repeated coastal tragedies associated with storms and a continuously rising sea level; and with regional disruptions due to freshwater shortages and shifting climatic zones."

Dr. Hansen goes on to make concrete suggestions, such as a cessation of building new fossil-fired power plants. He offers the possibility that in ten years it might be necessary to bulldoze all the existing ones.

Actually, a better plan would be to replace all the boilers with nuclear reactors. Most of the construction cost and effort is in the generating portion of the plant, anyway, so replacing the boilers will be much cheaper and faster than building new nuclear plants from scratch.

His other recommendations, such as financial incentives and improving the efficiency of buildings and vehicles, agree with the views of just about everyone who's looked at the issue.

But besides oversimplifying Dr. Hansen's remark, the kids are presuming to know more than they do.

We don't know how long it will take to build nuclear plants. In Japan they can build them in less than five years. If other countries got serious and cranked up their capacities for building them they could do it even faster.

The new designs are simpler than the earlier designs. The designers have incorporated features into them that make them inherently safer so that the risk of accident is lower even while the safety systems are less complex. Furthermore, manufacturing and construction technology has advanced in the last few decades. Just as office buildings can be put up faster and cheaper, so can power-plant structures. Computers and laser-guided machine tools have revolutionized the manufacture of heavy machinery. New testing techniques ensure quality control both cheaper and more thorough.

Besides, wind farms take more construction effort than nukes. Consider that a 1000 MW nuke will average over 850,000 KW. A very big (rotor-tip height ~ 450 feet!) wind turbine rated at 1.5 MW will average less than 500 KW. So 1 nuke equals more than 1700 big turbines.

As luck would have it, the British House of Lords studied the question and compiled some comparative data (and you wondered what the lords did!). A good measure of the construction effort is the energy inputs required for manufacturing and construction. What they show is that a 1000 MW nuke takes 6280 terrajoules per average GW, while a 25 MW wind farm takes 20,575, more than three times as much.

This is not an energy comparison, because that's much more complicated. We're only using these numbers to represent the manufacturing and construction effort.

What it shows is that the kids got it wrong. Again. Even if wind could provide full-time power, it still couldn't outpace nuclear in converting away from fossil fuels.