A Renaissance for U.S. Nuclear Power? Not Anytime Soon

In many quarters, nuclear power generates enthusiasm. A single pound of reactor-grade uranium oxide produces as much electricity as 16,000-plus pounds of coal—enough to meet the needs of the average U.S. household for more than one year. And whereas burning coal emits carbon dioxide, sulfur dioxide, and nitrogen oxides, nuclear power generation is virtually emissions free.

The March 2011 earthquake off the coast of Japan and its destructive tsunami damaged not just the Fukushima nuclear plant; the partial meltdown there reminded the world about the risks associated with nuclear power, highlighting concerns about such accidents and the storage of spent nuclear fuel. In the United States, no new nuclear power plant has come online since the mid-1990s, and despite a brief period of enthusiasm in 2007 and 2008, construction on a handful of new plants is proceeding very slowly. The biggest stumbling block for U.S. nuclear power is the high cost of construction for new plants. Following Fukushima, industry observers have called for expanded regulatory oversight, which will likely cause these costs to increase further.

Nuclear plants are expensive to build, for several reasons. All stages of design, construction, assembly, and testing require highly skilled, highly specialized engineers, and differences in reactor design and site-specific factors have allowed little scope for spreading design and production costs across multiple projects. The cost of building a nuclear plant rose sharply after the Three Mile Island accident in Pennsylvania in 1979, in part because the regulatory process became more stringent and construction of new plants took longer and longer. Plants ordered during the 1970s took, on average, 14 years to begin commercial operation, compared with 4 years for coal plants and 2 years for natural gas facilities.

The extended construction time means that financing costs are a critical factor. Nuclear plants have historically faced a very high borrowing rate because the profitability and even the viability of such projects are threatened by several kinds of risk:

• Regulatory risk. Approvals are required at federal, state, and local levels and can be difficult to obtain—and sometimes impossible, as cancellations of permits have shown. Moreover, the trajectory of federal energy policy remains unclear: Will carbon taxation increase the cost of competing fossil fuels? Will clean energy standards disadvantage coal? Or will some other action boost the prospects for nuclear power?
• Market risk. Fossil fuel prices rise and fall with world economic conditions and political events. Natural gas prices typically determine the marginal cost of electricity, so a decrease in natural gas prices reduces profits for nuclear power plants that sell to wholesale electricity markets. Natural gas prices in the United States are currently so low that an economic argument for nuclear is difficult to make.
• Technology risk. Wind, solar, or technologies as yet undeveloped or unknown could become more cost-effective than nuclear, or carbon capture and storage could make coal more acceptable, or energy efficiency technologies could reduce demand for electricity.

Even excluding financing costs, current estimates of construction costs show that nuclear is substantially more expensive than coal and natural gas. A recent U.S. Department of Energy study reports overnight costs for nuclear of $5,300 per kilowatt, versus $2,800 for coal and $1,000 for natural gas. This study was completed just before the Fukushima crisis and thus does not incorporate any cost increases due to closer regulatory scrutiny.

Once constructed, nuclear power plants produce power at lower marginal cost than almost every other source of electricity generation. A series of studies at MIT has used cash flow models to estimate the lifetime costs of nuclear compared to coal and natural gas. These "levelized cost estimates" indicate that these reduced operating costs are not enough to make up for the initial difference in construction costs. Using current construction costs and fuel prices, the levelized cost of nuclear power in the United States is about 10 cents per kWh, compared to 7 cents for coal and 5 cents for natural gas.

What would it take for nuclear to become competitive? A modest carbon tax of $25 per ton of CO2 would make coal more expensive, about 9 cents per kilowatt-hour, and thus improve the prospects for nuclear, but natural gas remains the least expensive, at about 6 cents. Thus, under most conditions, a modest carbon tax is not enough to make nuclear economical. A substantial increase in natural gas prices, combined with a decrease in nuclear construction costs could change the story.

Might nuclear construction costs decrease over time? Learning-by-doing has been shown to make a difference in some industries, and its potential to push down construction costs was one of the arguments for the 2005 Energy Policy Act, which provides production tax credits and other subsidies for new nuclear plants. Several studies that disentangle the effects of learning-by-doing from other factors have found evidence of company-specific learning, but not much industry-wide learning. That undermines the argument for subsidies: if individual companies can learn by doing, they have real incentives for investment, and no government intervention is necessary. Supporting emerging technologies would be more efficient than diverting government funds to a proven, demonstrated technology like nuclear power, for which investors are not waiting on the sidelines to see whether the plants will work.

Another important issue is standardization. In France, for example, virtually all nuclear reactors are of the same design, and this uniformity has increased learning-by-doing in plant operations. Nuclear proponents argue that the next generation of U.S. nuclear plants will be more standardized, facilitating lower costs of construction and operation. Although the industry may indeed eventually coalesce around a very small number of designs, this is not obvious based on recent applications to the U.S. Nuclear Regulatory Commission. Among the 17 license applications received from 2007 to 2009, there is a mix of reactor designs made by five manufacturers, suggesting that the French approach is not going to be adopted here.

The chairman of one of the largest U.S. nuclear companies recently said that his company would not break ground on a new nuclear plant until the price of natural gas was more than double today’s level and carbon emissions cost $25 per ton. This seems to pretty well summarize the current state of affairs in U.S. nuclear power. Yes, there is a confluence of factors that could make nuclear power a viable, economic option, but unless the stars and planets align around nuclear, it seems unlikely that there will be much of a renaissance.

—Lucas Davis is an assistant professor at the Haas School of Business at the University of California, Berkeley. Reprinted with permission from Resources for the Future.