by Andrew Reimers
I recently had a conversation with Jessica Lovering about her article, “Historical Construction Costs of Global Nuclear Power Reactors,” and the subsequently published replies to that article written by various authors, including Professors Jonathan Koomey and Benjamin Sovacool.
AR: What was your motivation for writing this article?
JL: We had heard a lot of arguments about nuclear power being intrinsically costly and increasing in cost, and we wanted to investigate whether those arguments were true. We put together comprehensive cost histories for seven countries, the U.S., France, Germany, Japan, South Korea, Canada, and India, which significantly expanded the size of the data that was available and added a lot of nuance to our understanding of the cost of nuclear power.
AR: You argue that it is somewhat problematic that most of the data sets for nuclear power costs only include plants constructed in the U.S. and France.
JL: They were the two major early adopters of the technology, and they suffered a lot of cost disadvantages from having to learn about what worked and didn’t work as they went along. The U.S. suffered especially because they never settled on a standard reactor design. Every plant was basically a one-off. After the incident at Three Mile Island (TMI), the U.S. had to jerry-rig solutions to different safety problems because every plant was different. In France, there were a lot of reactors under construction when the meltdown at Chernobyl happened, but because it was a standard design, they were able to make do with a smaller set of solutions and apply them at every plant. The result was that costs rose more quickly in the U.S. after TMI than in France after Chernobyl.
AR: In the response to your article, you were accused of cherry-picking data despite having used a more comprehensive data set than your critics used in their research.
JL: One of the responders to our paper said that we should have looked at their data set, which only had one plant from the U.K. – the most expensive plant in the U.K. by far. How is that not cherry-picking, but us using every single number we can verify is cherry-picking? They say we’re just choosing countries that have good stories, but we show the U.S., Germany, Canada, and Japan, and they all have rising costs.
AR: You use overnight cost and construction time as the primary metrics for describing how nuclear power plant construction costs increased over time. Your critics argue that these metrics are misleading because they don’t account for the cost of financing.
JL: We wanted to compare the trend in nuclear plant construction costs across different time periods and countries, and there are a lot of things you need to know or make assumptions about when you try to calculate the cost of financing. For example, financing is very different in South Korea than it is in Germany, and we didn’t want those kinds of differences to obscure the information we were trying to tease out. The cost of financing is ultimately a function of how much the project costs and how long it takes to build. In the U.S., overnight costs increase over time even if you correct for the increase in construction time, and so we wanted to tease out the different drivers of that increase in cost rather than lump everything together.
ANS member Andrew Reimers is a mechanical engineering Ph.D. student in the Webber Energy Group at the University of Texas at Austin. His research focuses on thermodynamic and economic analysis of power generation and water treatment systems. Andrew blogs about current events related to energy and water at andrewreimersblog.com.