Can nuclear fusion solve all our energy worries?

Short answer, perhaps some of them, but very much not in the short term. It’s going to take a long time to scale up and if we want to address climate change we still need to do all the other stuff too.

Nuclear fusion is one of the most beloved tropes in science fiction where it can be found powering everything from cameras to cars (Back to the Future) to Iron Man’s suit in the MCU and entire planets everywhere else. It’s small, it’s powerful and, it’s a useful plot device for being able to say ‘here we are in the future’ as well as provide all the energy for whatever doohickey currently needs it.

And, of course it’s now no longer science fiction. We’ve been able to do small scale fusion for a while now, replicating the same thermonuclear reaction that powers the sun. However, it’s always required more input than we get out at the other end. And while technically interesting, that’s not really of any practical value.

A team at the Lawrence Livermore National Laboratory in California though has just announced that it has produced the first fusion reaction that produces more energy than it requires to get started. You can read all the technical details in this good Guardian explainer, but essentially they used 2.05MJ (megajoules) of energy to produce 3.15MJ of energy out of a millimetre-sized capsule.

This is the culmination of seven decades of research and no wonder people are excited. Basically the Livermore team achieved what is called ignition; the point at which the energy released from smashing two atoms together to form one (deuterium and tritium to form helium if you must know) spreads to other nearby atoms and becomes self sustaining.

But it also has to be said that it’s very small scale so far. The 1.1MJ produced isn’t even enough to boil two kettles of water. And there are other big gotchas lurking in the wings too, such as the fact that the lasers used to heat the fuel and start the reaction took about 500MJ to power; or that the team producing the reaction can only do it about 10 times a week when the frequency you need to power a domestic appliance like a kettle is more in the 10 times a second window; or that the energy produced by the reaction is generally in the form of things like x-rays and gamma rays and would need to be converted into something that we can genuinely use and can come out of a plug socket.

Promising the moon

None of which apparently stopped US energy secretary Jennifer Granholm stating boldly that Joe Biden had set a target of “a commercial fusion reactor within 10 years.”

Politicians got to politic. Apart from Kennedy’s famous 1962 promise to land a man on the moon before the end of the decade, these promises are usually so much hot air. The actual scientists involved reckon three decades minimum, perhaps closer to five before we can scale up nuclear fusion to power plant size. They could, of course, be being over cautious; progress as we know is zipping along at a faster rate than ever, and we may be able to use AI and Machine Learning to start to crack even some some of the knotty real world engineering problems involved in scaling all this up to something useful that bit quicker.

But it’s also a dangerous comment when seen in the context of global warming and meeting net zero emission targets. Essentially politicians can — and will, just watch them — use it as a ‘get out of jail free’ card, saying that there’s a technological solution to all this carbon problem just coming down the pipe. The problem is that even if we can get large scale fusion reactors commissioned, built, and on stream by 2050, say, it will be too late to reverse catastrophic changes in global climate patterns caused by global warming.

Frankly, it already probably is.

That’s not to say we shouldn’t continue the research, however. We absolutely should. There are two competing methods, the inertial confinement approach that the Livermore lab just had success with and the rather more researched method of magnetic containment using a doughnut shaped torus (as in the header image above). The latter has thought through more of the engineering practicalities in scaling the process up, but has yet to achieve ignition. If it does, then all bets are possibly off and you can start shaving whole chunks of time off the timeline.

We also might ramp up research in the wake of the new achievement. The Fusion Industry Association — yes, there is such a thing — estimates that there’s roughly $4.8bn of private money being spent every year on developing fusion technology. By comparison, Meta has spaffed $15bn on its metaverse-focussed Reality Labs venture since the start of 2021. Perhaps we need to think about our priorities a bit more…