Fusion ignition is an historic achievement, but we’ll need machine intelligence to achieve clean nuclear energy
Fusion energy has long been “50 years away.” Even as time marched on, that 50-year gap never seemed to get shorter. Suddenly, however, things are starting to happen. Early in 2022, the Joint European Torus tokamak released 59 megajoules of energy in a 5-second pulse, more than doubling the previous high. JET’s larger sibling, ITER, is slated to go online in 2025. Then, at the end of 2022, the Lawrence Livermore National Laboratory in California announced it had achieved net gain, fusion ignition generating more energy than was put in. It was a world first, and it prompted President Joe Biden to set a goal of the U.S. having a commercial fusion reactor within a decade.
“The president has a decadal vision, to get to a commercial fusion reactor within 10 years,” energy secretary Jennifer Granholm said. “This shows that it can be done.”
That 50-year chasm may have closed rapidly, but there is still a gap to breach before fusion energy powers a grid. The Department of Energy is hoping AI and machine learning can help make it happen.
Major Breakthrough
The Dec. 5 experiment at LLNL that achieved fusion ignition yielded 3.15 megajoules of energy, surpassing the 2.05 megajoules that 192 lasers blasted onto a capsule of fusion fuel. For the first time in 50 years of laser experiments at the lab, the output topped the input.
“The pursuit of fusion ignition in the laboratory is one of the most significant scientific challenges ever tackled by humanity, and achieving it is a triumph of science, engineering, and most of all, people,” director Kim Budil said.
Researchers had figured out a way to get 8% more energy out of the lasers, enabling them to burst thicker capsules that could hold more fuel, made up of deuterium and tritium hydrogen isotopes. When those isotopes fuse, they produce helium nuclei and neutrons that can be turned into heat energy.
The extra power from the lasers and capsules put the experiment over the edge, making fusion ignition a reality.
“Crossing this threshold is the vision that has driven 60 years of dedicated pursuit—a continual process of learning, building, expanding knowledge and capability, and then finding ways to overcome the new challenges that emerged,” Budil said.
But Still a Long Way to Go
This all sounds extremely promising for fusion energy, which could be the best path to net zero. But there are still challenges to overcome. That the LLNL’s National Ignition Facility was able to achieve fusion ignition and get positive power generation is indeed a major accomplishment. But the true amount of energy that went into the process was more like 322 megajoules, the amount required to power the lasers. That’s more than 100 times what the reaction yielded.
“It’s a big milestone, but NIF is not a fusion-energy device,” David Hammer, a nuclear energy engineer at Cornell University, told Nature.
The ignition facility wasn’t built so much for energy efficiency, but to have the biggest and most powerful lasers for nuclear weapons research. Fusion ignition was, up until now, a distant dream. Budil herself acknowledges that this is only a first step.
“I don’t want to give you a sense that we’re going to plug the NIF into the grid: That is definitely not how this works,” she said during a press conference in Washington, D.C. “But this is the fundamental building block of an inertial confinement fusion power scheme.”
The Department of Energy also says there are many science and technology developments still needed before fusion ignition energy can power our everyday lives. But now there is a blueprint to follow. The foundation has been laid.
Research Funding
One thing the LLNL and Department of Energy do have is a huge amount of data. So much data, they couldn’t possibly pore over it all themselves. So, the energy department is doling out $33 million in grants over the next three years to researchers that can use AI and machine learning to sort through it all and find any insights human scientists might have missed.
The department is accepting submissions through March 15 and will likely fund six to 10 projects. The hope is these researchers and their machines can make fusion systems more efficient, maybe figuring out a better way to power the lasers, or better yet develop a fusion ignition technique that doesn’t rely on lasers.
“Recently, DoE initiated an agency-wide activity to accelerate fusion energy research, development, and demonstration activities to realize a commercially relevant fusion pilot plant (FPP) on a decadal timescale, an effort known as the Bold Decadal Vision,” the call for proposals says.
If the right machine can provide a solution, there surely will be more funding sent its way. If we are to build a commercial fusion reactor in the next decade or even two, AI and machine learning will definitely help power it.
Leave a Reply