Cost Escalation in France's Nuclear Reactors: A Statistical Examination

[bananas]

http://www.komanoff.net/nuclear_power/

NEW (January 2010): Cost Escalation in France's Nuclear Reactors: A Statistical Examination (pdf, 12 pp)

<snip>

[madokie]

1. recommend

[joshcryer]

2. K&R

I'll be called a nukenut for noting this, but he notes that concurrently built nuclear plants are significantly cheaper and there needs to be more data to model it:

Improve upon my modeling of the savings from concurrent reactor construction (“factor 2” in Table 2). As noted, this variable, denoting the number of reactors begun in a given year, was included in the regression model to help isolate the true annual cost-escalation rate. However, the phenomenon of “construction start proximity” is worth investigating in its own right. It doubtless can be specified more incisively than I have done here — perhaps with a construction-start-date “radius” (e.g., “radius” of months around the start of construction in which X other reactors were also started), or with a different functional form, or by a different approach altogether. In any case, the pattern of reduced costs through building many units at once merits attention, not just through finer statistical analysis but with informed interpretation.