Since there are no currently active contests, we have switched Climate CoLab to read-only mode.
Learn more at
Skip navigation
Share via:



We can't prevent catastrophic climate change unless we convert electric power to non-carbon sources. We need reliable, dispatchable, abundant power, at low cost.

One near-term option is molten salt reactors: nuclear reactors with liquid fuel.

While liquid thorium reactors have gotten a lot of attention, simpler molten salt reactors using uranium fuel have similar advantages and are easier to develop, and several startup companies are attempting it.

The U.S. has the capital and expertise to make a big impact in this area, reducing its own emissions and exporting proliferation-safe reactors to other countries. To make it work we need regulatory reform.

Which proposals are included in your plan and how do they fit together?

Molten salt reactors were featured in the 2011 plans:

Those advocated thorium breeders, which are more complex to develop. This proposal shifts the emphasis to uranium-fueled reactors.


Oak Ridge National Laboratory ran a molten salt reactor for four years in the 1960s. The experiment was a success, but the project lost funding, primarily because light water reactors were already being successfully used in submarines. Safety concerns raised by MSR proponents were discounted. In retrospect, that wasn't the best decision.

A conventional "light water" nuclear reactor uses solid fuel. The coolant is water, under pressure of up to 160 atmospheres to keep it from turning to steam. Refueling happens only every 18 months, so at the beginning an excess of fuel is required. Gaseous radioactive fission products build up over time. Continuous cooling is required. In disaster scenarios, hydrogen from the water can build up and explode, as we saw at Fukushima.

Despite all this, reactors have built up an impressive safety record, and even accounting for Chernobyl and Fukushima, nuclear is one of our safest energy sources. But it's taken a lot of complicated engineering and safety mechanisms to keep it safe.

A molten salt reactor is inherently safe, due to the physics of the fuel and coolant. It has a strong negative feedback: when the fuel heats up, the nuclear reaction slows down. Noble gas fission products are continuously removed. The fission products of most concern in conventional reactors (cesium, strontium, and iodine) are strongly bound into fluoride salts. There's nothing to drive a chemical explosion, and everything's at atmospheric pressure. Small amount of fuel can be added as often as desired, so there's no need for excess reactivity in the reactor. If there's a leak, everything radioactive will drip out and cool into rock, rather than venting to the atmosphere.

If the reactor loses electric power, or the fuel somehow overheats, a frozen plug melts and all the fuel drains to a tank designed to passively cool it. With many of the fission products removed, there's less decay heat than in conventional reactors.

Molten fluoride salts have been used for decades in aluminum refineries, with well-established safety protocols.

What about waste?

Almost all long-term nuclear waste is essentially unused fuel, left over from our inefficient reactors. Some MSR designs can consume this fuel, leaving behind only the fission products. Encase them in glass and bury them, and they'll be back to the radioactivity of the original ore in about three centuries.

Some designs are also strongly resistant to weapons proliferation.


With inherent safety and simple construction, MSRs can be much cheaper than conventional nuclear. With designs amenable to mass production, they can be rapidly deployed. With better fuel efficiency, they won't be limited by uranium production.

They have the potential to create a true nuclear renaissance, restoring the early vision of atomic power for a peaceful, abundant, and environmentally friendly civilization.

We'll highlight two U.S. companies attempting to bring them to market.


ThorCon's design is very similar to the 1960's reactor at Oak Ridge, purposely avoiding new technology. Since Oak Ridge already did four years of testing, ThorCon hopes for a short pathway to a production reactor.

ThorCon's team has experience with shipbuilding, and its primary innovation is shipyard construction. The reactor would be built the same way we build ships, in prefabricated blocks that can be plugged together. Each block would be tranported to the plant site by barge. Reactor cores would be sealed units; they run for four years, cool for four years, and then are shipped back to the factory for processing.

ThorCon's design is actually simpler to build than ships of similar size, which we successfully mass-produce. ThorCon estimates that a one large shipyard could produce 100 GWe of capacity per year. ThorCon estimates costs to be as low as 3 to 5 cents per kWh, with a plant cost of $1200/kW.

ThorCon's design also has disadvantages. It requires 20% enriched uranium, which legally qualifies as "low-enriched" but is still much higher than conventional reactors. Fuel utilization and waste production are better than light-water reactors but not as good as some alternate designs.


Transatomic is a U.S. startup out of MIT, with a reactor design that makes significant upgrades to the Oak Ridge reactor.

Oak Ridge used a graphite moderator (which slows down the neutrons so they're more likely to cause fission). Transatomic uses zirconium hydride; since it's more effective, only 50% of the core has to be moderator, compared to the 90% with graphite. Transatomic also uses a fuel salt which can hold 27 times as much uranium.

This design lets Transatomic hold much more uranium in a given volume. Consequently, it can use very low-enriched uranium, as low as 1.8% U-235.  Fuel efficiency is 75 times better than light water reactors. Using known uranium reserves, Transatomic's reactor could supply all the world's energy needs for 4000 years.

The little waste which is produced is primarily fission products with relatively short half-lives, requiring containment only for a few centuries. The reactor can also consume existing long-term nuclear waste; there's enough in existing waste stockpiles to power the world for 72 years.

Since the reactor uses very low-enriched uranium and produces a fuel mix which is unsuitable for weapons, it's strongly resistant to proliferation.

The reactor's high power density makes it small enough to be built in factories and shipped to the plant site. Transatomic estimates a capital cost of $1.7 billion for a reactor producing 520 MWe and 1250 MWt.


The U.S. has the capital and expertise for a vibrant nuclear industry. What it doesn't have is a regulatory structure compatible with new reactor designs. This in addressed in the 2014 proposal:

Public support

In the U.S., extensive nuclear expansion might seem like a non-starter in terms of public support, and certainly there will be opposition. But this isn't unique to nuclear; wind farms haven't exactly had a smooth pathway either. We're going to face political challenges no matter what we do.

Reactors that produce insignificant and short-lived nuclear waste, with easily-understandable safety advantages, could be a big help. Anecdotally, the author has described these reactors to anti-nuclear members of the Green Party, who responded that they sounded like a pretty good idea.

Explanation of the emissions scenario calculated in the Impact tab

What are the plan’s key benefits?

What are the plan’s costs?

What are the key challenges to enacting this plan?


Related plans