An MIT-connected project is changing how we view energy production.
It should have been bigger news. When the National Ignition Facility conducted a successful fusion experiment, we got a front-row seat to what could happen if this technology is ever perfected. But then it seemed like most people just forgot.
As part of a series of presentations this month, I heard from Bob Mumgaard, CEO and Co-founder at Commonwealth Fusion Systems, who talked about the ongoing race for new sources of clean power that could change the world, and how AI is helpful in doing groundbreaking research.
“Fusion is a reaction, and there’s a race right now to figure out the right machine to make that reaction happen here on Earth. The machine that would make the reaction would consume a fuel that everyone has access to – no need for uranium or plutonium and no nuclear waste – and would provide unlimited energy.”
Leading the Charge
Out of about 50 companies out there, CFS is the biggest. Mumgaard detailed advances that could lead to practical, production-level fusion processes. But challenges remain.
A reaction like this, he says, needs three things. It needs to be hot, dense and insulated.
Citing advances that outpace Moore’s law, Mumgaard suggested that a successful result in current research would lead to a plant that would outperform anything ever seen before. He explained that CFS purchased land on an inactive army base in Devens, Massachusetts, and created a commercially relevant demonstration facility called SPARC that uses high-temperature superconducting (HTS) magnets.
Specialists call SPARC a “tokamak,” a Russian word for a device using magnetic field for thermal nuclear fusion. It turns out that the Russians and the Europeans are also working on fusion using this tokamak approach.
Fusion at the National Ignition Facility
Over at the Lawrence Livermore Laboratory’s National Ignition Facility, they’ve been working on fusion research for years, and a couple of years ago, the first successful fusion experiment of its kind was completed.
In this case, scientists used lasers to hit small target pieces, trying to generate power, and for the first time, they created more power than they put into the reaction.
You can see a detailed tour of the plant and a lot of enthusiastic explanations of the science in play at this segment from 60 Minutes.
The NIF was built for $3.5 billion to ignite self-sustaining fusion,” intones the correspondent. “They tried nearly 200 times over 13 years. But like a car with a weak battery, the atomic engine would never turn over.”
You can also see in the segment that most experts were dubious about having commercial fusion anytime soon, for example, within a decade.
Fusion Power Now?
However, Mumgaard said his company announced the site on which CFS plans to build the world’s first grid-scale fusion power plant outside of Richmond, Virginia, bolstered by a collaboration with the utility, Dominion Energy Virginia. The proposed plant is expected to connect to the grid in the early 2030s..
“Investors are in it for the long haul, with the capital to see it through,” he said. Citing an exponential curve, he noted that this type of progress could help us to avoid higher temperatures and battle climate change, where we’ve already blown past the 1.5 degree standard, but may be able to keep global warming to two degrees, instead of, say, four degrees.
And AI is going to be important.
Specifically, a team at Princeton University has been able to figure out how to use AI to understand and forecast plasma instabilities as the magnets work on the plasma.
“By learning from past experiments, rather than incorporating information from physics-based models, the AI could develop a final control policy that supported a stable, high-powered plasma regime in real time, at a real reactor,” said Princeton Plasma Physics Laboratory research leader Egemen Kolemen at the time, about a year ago.
Now, with the advent of SPARC, we are that much closer to seeing how this methodology would work in practice.
In some ways, this is the most important application of artificial intelligence that you can think of. If it ends up saving our world from climate Armageddon, everything else is going to seem secondary.
