Washington is pulling out all the stops to restore the United States’ advantage in civil nuclear technology vis-à-vis Russia and China. The private sector is also investing heavily in the revival of nuclear energy in hopes of meeting the growing power demand of data centers, particularly those used in artificial intelligence. Though substantial, these efforts have underprioritized critical capacity-building in nuclear heavy manufacturing. In doing so, they risk repeating the mistakes that helped short-circuit previous hopes for a “nuclear renaissance” more than a decade ago.
In March 2017, the leading U.S.-based nuclear engineering company, Westinghouse Electric, declared bankruptcy under the weight of exorbitant delays and cost overruns deriving from the construction of its latest AP1000 reactors at the Vogtle Nuclear Power Plant in Georgia. The AP1000 technology was viable, as was demonstrated by the startup of four Chinese AP1000 reactors the following year. What led the Vogtle project into trouble, among other failings, was the near-complete degradation of U.S. nuclear manufacturing capacity after domestic U.S nuclear construction essentially ended in the 1980s.
The Vogtle project reflected the atrophy of the U.S. civil nuclear supply chain. The failure rates of components rose to 40-80 percent over the course of the project, and the design for certain components was discovered to be incomplete or impossible to manufacture, leading to further delays. The Shaw Group—the main contractor Westinghouse selected to make components for the Vogtle units in 2008 and with which it was financially intertwined—was a metal fabricator with no nuclear industry experience apart from its acquisition of a bankrupt nuclear company that had lost all its nuclear assets. The Shaw Group’s new heavy manufacturing facility in Louisiana failed to deliver the Westinghouse project’s nuclear modules as planned and, in subsequent years, transitioned to the fabrication of oil and gas equipment instead.