The Obama administration is promoting a bold, long-range plan for building dozens of small, factory-built reactors capable of replacing coal-fired power plants that are expected to be retired in the coming decades, a Department of Energy official said yesterday.

DOE’s effort is aimed at establishing an industry that would manufacture as many as 50 small modular reactors (SMRs) a year by 2040 or sooner, said Rebecca Smith-Kevern, the director of light water reactor technology at the department’s Office of Nuclear Energy, which oversees the licensing of tiny nuclear plants.

“We have a vision of having a whole fleet of small modular reactors produced in factories,” Smith-Kevern told a regulatory conference in Bethesda, Md. “We envision the U.S. government to be the first users.”

DOE this week announced a second wave of million-dollar cost-share grants to help the industry design and license the modular reactors, which the administration defines as factory-built plants of less than 300 megawatts that are shipped by truck, barge or rail to construction sites for assembly.

The department awarded the first grants under its $452 million cost-share program to veteran reactor designer Babcock & Wilcox, which is building two small units at the Clinch River site in Oak Ridge, Tenn.

Smith-Kevern said the agency hopes to complete the awards by the year’s end.

The administration is pinning its hopes on the establishment of a factory network to make the small plants. After supporting the licensing and design process for the plants through grants and possibly loan guarantees, the “first movers” would be built, Smith-Kevern said. A network of U.S. manufacturing plants would then be established, and vendors would begin filling orders, which she said could lower plant costs and attract investors.

By 2030, the industry would produce about 20 small modular reactors a year with the support of production tax credits and feed-in tariffs – allowing for a mature industry to develop. By 2040, the United States could produce up to 50 of the small plants annually with the potential for an import and export market, she said.

“We recognize this is … an ambitious vision,” she said. “There would be a role for new public policy to support small modular reactors.”

Smith-Kevern said the administration is also hoping to enlist the military and is looking at building modular nuclear plants on the islands of Hawaii and Guam to lessen their oil dependence.

The plan has sparked debate about the costs of building small nuclear plants and how regulators would ensure that such facilities are safe.

Smith-Kevern said modular construction of the reactors could reduce costs for the industry hindered by price tags of $12 billion or so for traditional nuclear power plants.

Smith-Kevern cited a 2011 paper by Robert Rosner of the Energy Policy Institute at the University of Chicago and Stephen Goldberg of the Argonne National Laboratory that found modular plants could cost $3 billion to $5 billion.

But Ed Lyman, a senior scientist with the Union of Concerned Scientists, said capital cost per kilowatt – not the cost of building a reactor itself – is what matters.

“Small plants, of course, cost less than large plants, but they also generate less electricity,” Lyman said. “And with the economies of scale factor, small plants will cost more per kilowatt than large plants unless there is some major cost savings somewhere to offset this factor.”

Lyman cited a 2007 report by Westinghouse when he told the Senate Energy and Natural Resources Committee, “At best, the capital cost of four 335-megawatt reactors was slightly greater than that of one 1,340-megawatt reactor.”

The Nuclear Energy Institute, on the other hand, has taken issue with Lyman’s use of the report and has said the Westinghouse numbers constitute a preliminary evaluation that found the capital costs of small and large plants are “practically equivalent.”

Lake Barrett, a former DOE official and nuclear consultant, said groups for and against small modular reactors are using dueling reports to argue their cases, but cost estimates aren’t firm and could change as the plants are developed.

Factors that must be considered, he said, are fixed costs like security, licensing and siting; whether the plants can be mass-produced; and uncertainty about whether Congress will price heat-trapping carbon emissions.

“How successful SMRs are going to be, we really don’t know yet,” he said. “If the world is going to phase out older coal plants because of carbon concerns, you have to replace them.”