To a great extent U.S. nuclear energy policy in the last two decades has been driven by the premise that standardization of designs and construction procedures will bring down reactor costs. France's almost all-nuclear electricity sector, built basically by a single national company, is the model. But yesterday a French nuclear energy expert called that model into question. Speaking to reporters in an event sponsored by the Physicians for Social Responsibility, Yves Marignac pointed out that the country’s recent experiences with nuclear construction have not been exemplary. In a sense that's stating the obvious. Delays in building a nuclear power plant in Normandy and a similar one Finland, and a 75 percent cost overrun on the Finnish plant, have been widely reported. But Marignac says France's record building its four previous plants was not much better: They were connected to the grid only 12.5-15.5 years after construction began—a far cry from the four or five years in which Areva claims it can build new plants.
Asked what accounts for chronic construction delays, Marignac said it often is a matter of rather rudimentary problems such as failure to pour concrete or do welding to (admittedly exacting) standards. Why is pouring concrete harder for a reactor than, say, for a bank vault? For one thing, he said, because a containment's double walls have to be able to withstand hydrogen pressure and prevent hydrogen leakage in the event of an extreme meltdown accident.
Marignac, executive director of WISE-Paris, an information service, has contributed to blue-ribbon energy reports for the French government and European parliament. His views represent a significant challenge to the U.S. nuclear industry and its promoters because, if valid, they suggest the French model doesn't work well even in a country whose institutions favor it—a country with highly centralized planning, traditions of precise administrative procedure, a single administrative entity, and a single builder and single client,.