(Bloomberg Opinion) -- Twenty-five years ago, as a young physicist, I worked on research linked to fusion energy. Nuclear fusion powers the sun and stars through reactions that turn hydrogen nuclei into helium nuclei, and if we could master the process on Earth, we'd have a safe and virtually limitless source of clean energy. At conferences every year, scientists from around the world gathered to discuss the achingly slow progress then being made. Government and university labs have now been trying unsuccessfully for more than 50 years.
The fusion conferences still go on — this month in Opole, Poland, and at the University of Wisconsin in September — but there’s a different tone, and more excitement, as the research finally seems to be bearing fruit and bringing us closer to fusion energy as a viable energy source. A flurry of new startups aim to achieve it within only 15 years, spurred by the belief that nimble private companies may succeed where lumbering government projects have not. Think Elon Musk’s Space Exploration Technologies Corp.
It’s probably naive to expect similar success for fusion energy. It took SpaceX 13 years to become the first private company to launch and then re-land a rocket capable of achieving orbit, and the company had the advantage of a deep base of scientific understanding of rocket technology. In the race for fusion, in which the extreme conditions required to create it make the scientific challenge immeasurably harder, the slow and methodical government labs may still turn out to have the edge.
Conceptually, fusion seems easy: It only requires getting hydrogen fuel to a sufficiently high pressure and temperature, then keeping it there long enough for the crucial reactions to take place. The main obstacle is instability, arising from the many surprising ways things can go wrong. In one approach, known as inertial confinement fusion, or ICF, intense pulses of laser light are used to compress a tiny pellet of fuel in an attempt to reach temperatures and pressures comparable to the interiors of stars. Despite decades of refinement, the violent compression always mixes hot and cold parts of the imploding fuel together, limiting the results.
But a series of experiments last year at the Lawrence Livermore National Laboratory’s National Ignition Facility did finally pass a landmark, with the implosion releasing more energy from fusion reactions than was put in to make it happen. Computation and machine-learning algorithms may help researchers build on this achievement through improved experimental designs.
Another approach uses magnetic bottles to hold matter at extreme temperatures but much lower densities than the compression experiments. This is the goal of the International Thermonuclear Experimental Reactor, the world's largest fusion project, being built in France; it hopes to prove the possibility of sustained fusion by 2035. Algorithms are also improving prospects here, where the key instabilities come through “disruptions” involving sudden catastrophic discharges of super-hot matter and electric current that damage key components of the reactor. One research team recently found that algorithms can be trained on data to spot these damaging disruptions before they occur, giving scientists a warning of 30-thousandths of a second, enough to take steps to avoid a breakdown altogether.
Such mainstream approaches build upon decades of accrued learning. Could any of the new private firms beat them to the finish line? One called Commonwealth Fusion Systems and linked to the Massachusetts Institute of Technology plans to build a working reactor by 2025. First Light Fusion, a company spun out of Oxford University in the U.K., plans to use a radically different approach completely unlike the mainstream fusion efforts. More than 20 other startups have emerged in the past few years.
In a report last year, the U.S. science and technology advisory group JASON offered a fairly pessimistic view on the prospects for near-term, low-cost fusion success, in part by looking to the development history of other key technologies, including solar and wind energy. These came to fruition by gradual, incremental improvements through better designs and new materials, and much more slowly than most industry experts expected. JASON put practical fusion energy still 30 years in the future.
I asked Omar Hurricane, chief scientist of the ICF program at Lawrence Livermore National Laboratory, for his thoughts. He describes himself as a “pessimistic optimist.” After 50 years of research, he told me in an email, scientists now understand the extreme conditions required for fusion. They still can't create them in the lab but are gradually getting closer. He expects the startups to make some interesting advances and rapid initial progress, but they will run into more difficult problems soon thereafter, as the other fusion programs have.
“Science and technology follows a learning curve, so with anything new we get rapid initial progress,” he says, “but the rate of learning steadily slows as the easy problems get solved.”
Ultimately, success is likely, he expects, although the timing may be difficult to predict.
“I am optimistic that the laboratory fusion problem is a solvable one,” he says. “But Mother Nature makes the fusion problem diabolically hard because the required conditions are so extreme.”
He suggests that investors in private-sector fusion efforts have patience, and they shouldn't expect too much, too soon.
“Getting to fusion is going to be a long hard march that will require a lot of discipline to stick with and long-term vision to realize,” he says.
To contact the editor responsible for this story: Brooke Sample at bsample1@bloomberg.net
This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
Mark Buchanan, a physicist and science writer, is the author of the book "Forecast: What Physics, Meteorology and the Natural Sciences Can Teach Us About Economics."
©2019 Bloomberg L.P.