A firm founded by Bill Gates bets on a novel nuclear reactor

SINCE HANDING over the reins as Microsoft’s chief executive in 2000, Bill Gates has been best-known for his philanthropy. The Bill and Melinda Gates Foundation, one of the world’s largest charities, has given billions of dollars to vaccination drives, family-planning clinics, research into drug treatments for malaria and more.

But Mr Gates has not abandoned the business world entirely. On June 2nd TerraPower, a company he founded in 2008, announced that it would build a demonstration of an exotic, high-tech nuclear power station in Wyoming. The firm’s Natrium reactor is one of a gaggle of new designs that have emerged in recent years, as engineers try to come up with cheaper, simpler nuclear power plants that can provide low-carbon electricity with fewer of the cost and safety worries that have plagued the industry in the past.

The Natrium reactor makes two big changes to the standard nuclear-power-plant design. It replaces the liquid water that normally courses through the core with hot, liquid sodium (natrium, in Latin). And instead of using the heat generated by the reactor to make electricity directly, it first employs it to heat a tank of molten salt that acts as a giant battery. The upshot, the firm hopes, will be a cheaper reactor that is better suited to power grids that will increasingly be dominated by intermittent sources of energy such as wind turbines and solar panels.

Start with the reactor itself. Most nuclear power plants are light-water reactors (LWRs), a technology developed in America in the 1950s. They use ordinary water both to cool the reactor core and to increase the intensity of the chain-reaction by moderating the speed of the neutrons that are emitted when uranium atoms split. Thus slowed, these neutrons are more likely to go on to split more atoms in turn.

Natrium employs hot, liquid sodium as a coolant, and dispenses with the moderator entirely. This is another idea that dates back to the 1950s, but one which has never been widely deployed. Yet sodium offers several advantages as a coolant, says Chris Levesque, TerraPower’s boss. The liquid sodium’s high temperature—around 500°C—makes the reactor more efficient. At the same time, liquid sodium is much less corrosive to pipes than hot water. And though the water in LWRs is pumped through at high pressure, Natrium is designed to operate at close to atmospheric pressure. That means pipes, containment buildings and the like can be less beefy without affecting safety. TerraPower reckons its reactor needs only 20% of the concrete required by an LWR of equivalent power, which helps keep down costs.

The firm’s second big idea is its molten-salt energy-storage system. Inspiration for this came from the solar-power industry, says Mr Levesque. Solar-thermal systems (in contradistinction to the more familiar photovoltaic ones that generate electricity directly) have, for several years, used similar tanks to store excess solar energy harvested during the day. In Natrium’s case, the sodium coolant transfers heat from the reactor into the molten-salt tanks. A separate set of pipes then removes heat from the tanks and uses it to produce electricity.

TerraPower hopes this arrangement will let the new reactor ramp its power output up and down, depending on the price of electricity. This is something that LWRs struggle to do. The firm’s demonstration plant will usually produce 345 megawatts (MW) of electrical power. But by releasing the energy stored in the molten-salt tanks, it will be able to boost that to 500MW for over five-and-a-half hours. This should be a useful trick as power grids fill up with wind and solar farms that are likely to cause power prices to fluctuate more than they do at present. Combined with lower construction costs, TerraPower hopes such agility will make its plant more economically attractive than older designs.

It all looks good on paper. But then, nuclear power always does. The industry has been plagued by delays and cost overruns for decades. Existing sodium-cooled reactors, most of which are experimental, have a spotty record. A plant in Japan suffered a serious fire in 1995 and was shut down for over a decade. The Superphénix reactor in France, built in 1974, proved extremely unreliable, and was offline for years at a time. It closed for good in 1998. (Other reactors, such as the Fast Flux Test Facility in Washington state, have better records.)

The Union of Concerned Scientists, an American not-for-profit organisation, argues in a report published in March that sodium’s advantages as a coolant are counterbalanced by drawbacks. One is that a reactor which ran too hot might see its power output rise as a consequence. Unlike water, the loss of which shuts a reactor down for lack of moderation, sodium slightly damps the chain-reaction. If bubbles of sodium vapour formed in the coolant, that damping effect would diminish, risking a dangerous feedback loop of rising temperatures and growing power output.

The physics of such judgments are tricky. Few countries have as much nuclear experience as France, which generates around 70% of its electricity that way. Yet in 2015 French regulators said they could not determine whether sodium-cooled reactors are significantly safer than modern LWRs. TerraPower, moreover, insists that its Natrium plant is designed in a way that makes runaway reactions impossible.

America’s government, for its part, thinks the technology has merit. It is chipping in $80m to help TerraPower build the demonstration plant, which the firm says should be ready by 2028. In the meantime, says Mr Levesque, TerraPower has been fielding inquiries from electricity firms interested in its technology. Whether Mr Gates’s bet on a nuclear-power revival will pay off remains to be seen. ■

For more coverage of climate change, register for The Climate Issue, our fortnightly newsletter, or visit our climate-change hub

A version of this article was published online on June 9th 2021