- Energy too cheap to meter.

That was the promise of nuclear power in the 1950s, at least according to Lewis Strauss, chairman of the Atomic Energy Commission.

That promise has not yet come to pass.

But with some incredible new technologies, perhaps it still could.

The question is, should it?

(upbeat music) Energy isn't scarce, it's everywhere.

Seriously, literally, all mass is energy.

The trick is getting at it.

Burn coal and you liberate a tiny bit of the energy locked in its chemical bonds.

That's easy and cheap to do, but the energy you get is pathetic per kilogram of coal, and worse per ton of carbon dioxide released into the atmosphere.

At the other end of the spectrum is the energy released when particles of matter and antimatter are brought together.

They annihilate each other, releasing 100% of the energy contained.

Sounds great, except that antimatter is incredibly difficult to create and store.

In between breaking chemical bonds and matter/antimatter annihilation we have nuclear energy.

The strong nuclear force holding nuclei together contains an enormous amount of energy.

The Sun is powered that way.

Releasing a mere 0.4% of the mass of hydrogen nuclei as it fuses them into helium.

But that's enough to power the Sun for 10 billion years.

Practical fusion power stations are a holy grail of energy production, but are still a long way off.

Until then our only viable source of nuclear energy is fission, which means breaking very heavy nuclei into more stable smaller parts.

If we want to convert mass into energy, fission gives us the most bang for our buck.

Unfortunately, that bang can be literal.

Use of nuclear energy may risk the proliferation of nuclear weaponry, and there's also the problem of nuclear waste and the specter of horrible accidents.

This last one was painted in terrifying detail in the recent dramatization of the Chernobyl disaster.

Nuclear reactors sounds scary because the disasters are pretty epic.

However, the reality is that far far more people die from straight up air pollution due to coal-fired power plants than ever died in nuclear reactor accidents.

In fact, the radioactivity around coal fire plants is also higher due to the traced but completely uncontained radioactive products of coal burning.

But the most compelling attraction is that nuclear power doesn't directly produce carbon emissions.

In fact nuclear power may be our most sure path to reducing carbon emissions and halting climate change.

But can we do nuclear power safely enough?

There are modern ideas, including the much hyped Thorium Reactor, that suggest maybe we can.

Before we can understand those, we'll need to review how nuclear reactors work.

Every fission reactor exploits the same phenomenon.

Certain very large nuclei like uranium and plutonium can split into smaller nuclei when hit by a single neutron.

When these nuclei split they release energy and fast-moving neutrons.

Those new neutrons can smash into nearby nuclei, breaking them up and releasing more neutrons.

If you have enough of these heavy nuclei, if you exceed what we call critical mass, then neutrons produced in every fission trigger at least one more fission.

That's a chain reaction, a domino effect.

That can be a runaway chain reaction in which each split nucleus causes multiple other nuclei to split, resulting in an explosive release of energy.

That would be an atomic bomb.

But if you can regulate the process, make sure that each nucleus splitting causes on average only one other nucleus to break, then the reaction can be controlled.

It can be made to produce a steady amount of heat that is used to power a turbine, often just by boiling water.

The most common commercial power plants use uranium fuel, in particular the isotope Uranium-235.

Uranium-235 has 92 protons and 143 neutrons.

It makes up less than a percent of naturally occurring uranium, which is mostly Uranium-238 with three extra neutrons.

U-235 is useful because it's highly fissile, which means it has a high probability of intercepting a stray neutron and splitting.

It's fissile in the presence of the fast-moving neutrons created by its own fission.

But it's many times more fissile if those neutrons are first slowed to become so-called thermal neutrons.

On the other hand the more stable Uranium-238 is only fissile to fast-moving neutrons and not at all to slow neutrons.

In fact, is much more likely to absorb slow-moving neutrons.

The cheapest way to do commercial fission is to take advantage of Uranium-235's high fissibility to these thermal neutrons.

To sustain thermal fission in uranium you need to enrich it by a few percent, increase the proportion of U-235 relative to U-238 so that more neutrons get created and fewer get absorbed.

You also have to slow down those neutrons into the sweet spot for splitting U-235.

To do this, thermal reactors use some sort of moderator.

The most common moderator is plain old water.

Because hydrogen nuclei in H2O are around the same mass as neutrons, they absorb a lot of momentum in neutron collisions and conveniently that same water can also work as a coolant.

It takes heat away from the uranium fuel preventing meltdown, to where it's needed, which is to drive a turbine either directly or via a secondary loop of water.

I just described, very, very crudely, the principles behind the light water thermal reactor.

These are the most common because they're cheapest, but let's talk about the problems.

First, there's safety.

Every major disaster has been with a thermal reactor due to a cooling failure.

In Three Mile Island, the water escaped a jammed hatch.

In Chernobyl, water boiled increasing the neutron count.

And in Fukushima, a tsunami knocked out the water pumps.

The common issue is that water cooling requires active effort to maintain and so is prone to disruption.

Modern light water thermal reactors addressed the failures of the past and repeats of these disasters are very unlikely.

But unforeseen failures are still possible especially due to human error.

Even the smartest nuclear engineer can have a Homer Simpson moment.

One way around the coolant issues is to use molten metals or molten salts.

These can be liquid over a very large range of temperatures reducing the chance of accidental boiling, and they allow the system to be operated at much higher temperatures, which increases efficiency, and at much lower pressure than water.

The high pressures required for water-cooled reactors add a lot of complexity and size and potential to explode.

Perhaps the worst downside of the common modern reactor is the waste.

They use only around 1% of the uranium extracted from the ground, the U-235.

Some of the U-238 gets converted to fissile plutonium by absorbing neutrons.

But most of it is either unused or converted to heavier non-fissile elements.

These are the so-called transuranic actinides, elements heavier than uranium on the actinide sequence of the periodic table.

They are very radioactive and have half-lives of tens of thousands of years.

That means they're dangerous on geological timescales, and there is literally no place on earth we can guarantee that containment vessels will be safe against earthquakes, volcanic activity, or eventual crushing by Ice Age glaciers.

A possible solution to this nightmare waste disposal issue is to try to burn all of the heavy nuclei as fuel.

One way to do this is to use fast neutrons.

A fast reactor doesn't try to slow down the neutrons, that means U-238 can split along with U-235 and along with any actinide that happens to be produced by neutron absorption.

The waste products of a fast reactor are the fission products, much smaller nuclei than the actinides.

Some of these are incredibly nasty, like Cesium-137, but they have half-lives of centuries not tens of millennia.

So safe storage is at least plausible.

Fast-neutron reactors get to be smaller than their slow thermal cousins, because they don't need a neutron moderator.

That makes them ideal for things like submarines.

The issue with these guys is that you need much more enriched fuel.

The U-235 content needs to be over 20%, several times higher than in a thermal reactor.

And that's just because the overall fission rate is much lower per fast neutron compared to slow neutrons.

That enrichment is expensive and so after abundant natural uranium deposits were discovered and fuel became cheap, commercial interests opted for the thermal reactor, even though it wastes 99% of the fuel and leads to eons of looming environmental catastrophe.

Fast reactors also have the advantage that they can create or breed their own fuel.

Although fast neutrons don't keep nuclei as easily, when they do hit they liberate more free neutrons than when a slow neutron causes fission, typically two to three neutrons per split.

That means you have one neutron to contribute to the fission chain reaction, and at least one more to be absorbed by a non-fissile element to turn it into something fissile.

An element that can do this is called fertile.

For example Uranium-238 is fertile because it can absorb a neutron and be transformed into Plutonium-239.

A typical breeder reactor includes a reactor core burning highly enriched uranium or plutonium, surrounded by a blanket of fertile material that cycles into the core as it becomes fissile.

Thermal and fast reactors have different advantages and disadvantages regarding nuclear proliferation.

The waste of a thermal reactor isn't fissile, but it could be bred into fissile material.

The ultimate waste products of a fast breeder reactor are not dangerous in this way, but the intermediate products include weapons-grade plutonium, which you definitely don't want in the wrong hands.

Some of the advantages of both of these reactor types can be achieved by switching to a completely different fuel, thorium.

That's the thorium reactor, Thorium is another actinide two spaces lighter on the periodic table compared to uranium.

It's not naturally fissile, but it is fertile.

Upon absorption of a neutron it decays into Protactinium-233 and then into Uranium-233.

And U-233 is nicely fissile.

It's even better than U-235 and plutonium-239, because it absorbs fewer neutrons which means better neutron economy and more importantly, on average uranium-233 produces slightly more than two neutrons per split, even when it's split by a slow-moving neutron.

That means it's possible to breed new uranium-233 from thorium in a thermal reactor.

You don't need a fast reactor.

There are different ways to build a thorium reactor, but perhaps most promising is the liquid fluoride thorium reactor or FLTR, LFTR.

In this design both thorium and uranium-233 are bonded with fluorine and dissolved in a molten fluoride salt, beryllium or lithium fluoride.

The fission in the uranium produces heat and neutrons to sustain fission and to breed more uranium from the thorium.

The uranium and thorium can either be mixed together or separated with a thorium blanket surrounding the uranium core.

In either case the molten salt containing the uranium also transports heat out of the core to secondary circuits that ultimately power a turbine.

The actual fusion only happens in the reactor core, because that's where the moderator slows down the neutrons to make fission much more likely.

In this case the moderator is a lattice of graphite channels through which the fluid flows.

Graphite is particularly great because it slows neutrons without absorbing them.

When the fluid is away from the graphite neutrons speed up which means fission slows down.

Because it's in liquid form the fuel can be quickly drained from the reactor in emergencies.

A plug with a low melting temperature will melt if the core gets too hot or if power supplying a cooling fan goes out.

The fuel then drains to a tank where fission is impossible.

In addition, built the right way, the liquid fuel becomes less fissile as temperature increases.

That's because at high temperatures thorium is increasingly good at absorbing neutrons.

So not enough neutrons are left to continue the fission.

This whole setup is a great example of passive or walkaway safety, meaning that in the event of an emergency even if every mechanical or human mechanism failed, the reactor would simply power down.

Another compelling advantage of the LFTR and molten salt reactors in general is that they can be small because they don't need giant structures to handle the high pressure water.

In fact, it was for use in submarines and aircraft that molten coolant reactors were first conceived.

But now this compactness and modularity means they could be inserted into the current electrical grid to replace coal or natural gas plants.

Or you know, on a lunar or Martian settlement or a starship.

That same modularity poses perhaps the biggest risk.

If small thorium reactors became widespread they'd become less easy to regulate and monitor.

We'd want to be very careful that the reactor design leaves the weaponizable U-233 completely inaccessible without enormous effort.

Nuclear power is a possible solution to our dire energy and climate challenges.

The question is do we need it or can we meet those challenges with renewables like wind and solar, assuming significant advances in battery tech.

I don't know the answer and I'd love to hear your opinions.

What I do know is that we face an enormous hurdle in our progression as a technological species.

One which may take all of the ingenuity we can muster.

We should think very carefully about whether the power of the atom is necessary to survive and thrive into the next technological stage and send us to greater distances and further futures in space-time.

(dramatic music)