The Tech - Online EditionMIT's oldest and largest
newspaper & the first
newspaper published
on the web
Boston Weather: 36.0°F | Mostly Cloudy

This column incorrectly references the “Electrical Power Research Institute.” It is the “Electric Power Research Institute.”

Article Tools

In the aftermath of the Fukushima Daiichi nuclear disaster in Japan, many wondered what the event’s impact would be on the nuclear renaissance in the United States. Those who follow the nuclear industry didn’t need eight months of hindsight to give an answer: what nuclear renaissance?

The outlook for U.S. nuclear power has worsened considerably in the past five years. Where once there were plans for new reactors at more than 30 different sites, today there are only five, and even those planned reactors might disappear. Only one is actually under construction, and to credit the industry with breaking ground on a new reactor is overstating its prospects.

However, none of this gloom is the result of Japan’s tsunami. On the eve of the Tohoku earthquake, U.S. nuclear power looked just as moribund as it is today. The cause of this decline is not renewed concerns about safety, or even that old red herring, waste disposal — instead, it is simple economics. Other technologies, particularly natural gas, offer much cheaper power than nuclear both today and in the foreseeable future.

In 2009, the MIT Future of Nuclear Power study released an update to its 2003 estimate of the costs of nuclear power. Estimating a capital cost of $4,000/kW and a fuel cost of $0.67/MMBtu, the study’s authors projected a cost of new nuclear power of 6.6 cents/kWh. Using the same modeling approach, the cost of electricity from a natural gas plant with capital costs of $850/kW and fuel costs of $5.16/MMBtu would be 4.4 cents/kWh.

What’s worse, the estimate of 6.6 cents/kWh assumes that nuclear power is able to secure financing at the same interest rate as natural gas plants. In reality, credit markets assign a significant risk premium to nuclear power, bringing its total levelized cost of electricity to 8.4 cents/kWh, nearly twice the cost of natural gas power. Unless the capital costs of new nuclear power plants turn out to be significantly less than what experts expect, or natural gas prices rise considerably in the near future, there is little reason to believe that any new nuclear plants will be built without significant subsidies.

This is not to say that nuclear power could not make a comeback within the next 10 to 20 years. But before nuclear can once again be considered a credible competitor to fossil fuels, four changes must happen.

The first problem facing nuclear power is its high capital cost. It is possible, of course, that one of the new reactor designs put forward by General Electric, AREVA, Westinghouse, or elsewhere will prove to cost significantly less than the $4,000/kW estimate being circled about, but such chances are low, and it is just as likely that these reactors will come in at higher-than-expected costs. If AREVA’s recent forays in Finland and France are any indication, $4,000 might prove a rosy number. Assuming we don’t live in a world in which nuclear power costs less than what we expect, nuclear is going to need to lower its capital costs, ideally to somewhere in the area of $3,000/kW.

Reducing capital costs starts by ending the obsessive focus of universities on improving the fuel utilization of reactors. A significant fraction of nuclear engineers at MIT and other institutions across the world are spending their time researching new reactor types, developing enrichment and reprocessing technologies, or involving themselves in other areas in which the goal is to lower the cost of fueling a reactor. This is folly; fuel costs represent roughly 10 percent of nuclear power’s total cost, and only five percent is the uranium itself — even if nuclear fuel was free, it wouldn’t be enough to make nuclear cost-competitive with natural gas. Instead of funding yet another Generation IV reactor design or one more study of closed fuel cycles, the Department of Energy and Electrical Power Research Institute should put serious money into the development of technologies that offer potential improvements to capital costs — of the research at MIT, two good examples are high thermal efficiency fuels (fuel pins with geometric designs that make it easier to remove heat from the reactor), and nanofluid coolants, both of which could make it possible to get more power from the same reactor.

The second problem facing nuclear power is its high borrowing costs. To some extent, this problem is a natural consequence of nuclear power plants taking a longer time to build than natural gas plants and having a much higher construction risk (the capital cost of natural gas plants is well-established relative to that of nuclear power). And likewise, to some extent, this problem might resolve itself over time, both as the completion of nuclear plants helps nail down the true capital cost of nuclear power, and as vendors add smaller, modular reactor designs to their list of offerings. But much of the reason behind the high interest rates on loans to nuclear construction is that the industry is scoring an own-goal. In the current relationship between utilities and reactor vendors, utilities are asked to absorb all of the costs of a vendor’s overruns — if a reactor ends up costing a couple billion dollars more than the vendor quotes, it’s the utility that is expected to make up the difference.

This is terrifying for a utility’s creditors. The largest utilities in the United States have market capitalizations in the area of $30 billion, while most hover closer to $5 billion. If a nuclear project should fail, the utility might go completely bankrupt, leaving nothing to those foolish enough to lend them money. Accordingly, nuclear projects face higher borrowing costs than other electric projects. It doesn’t have to be this way — if reactor vendors and construction companies helped share the project risks posed by nuclear plants, borrowing costs would be lower. It is also possible for the U.S. government to shoulder some of the risk — but after Solyndra, few legislators have an appetite for letting energy companies push their risks onto the taxpayer.

Next, the United States is going to have to adopt some form of carbon tax on electricity generation, or offer a comparable subsidy to the nuclear industry. An appropriately sized carbon tax of $20/ton CO2 would raise the cost of natural-gas-generated electricity by 0.7 cents/kWh, while having a negligible impact on nuclear power.

And finally, the nuclear industry is just going to have to catch some luck and see natural gas prices rise. That’s a tall order, given the new resources being opened up by hydraulic fracturing and the slowed consumption of natural gas brought about by the recession. But it’s not entirely outside of the realm of possibility — the futures market for natural gas has been wrong before.

Nuclear power is down, but not out. With a proper R&D focus, good business practices, appropriate policy, and a little luck, the gulf that separates nuclear power from its competitors may yet be bridged.

Still, it’d be best not to hold your breath.