Nuclear energy has been with us for eight decades, with decadal ups and downs for the last half-century.
- In the 1950s, nuclear energy was going to give us energy too cheap to meter
- In the 60s, the perceived risks of radiation fed nascent anti-nuclear sentiment.
- Nuclear power began to gain momentum again during the 1970s until the meltdowns at Three Mile Island and the Chernobyl accidents dealt it a double blow, knocking it out of contention in the 1980s and 1990s.
- In the first decade of this century, carbon-free baseline energy looked promising; however, the meltdowns in Fukushima knocked it out of contention once again.
Now, once again, we find ourselves thinking that carbon-free energy might not be a bad idea after all, this time to power our AI data centers and server farms. This time, there’s a new wrinkle: Instead of complex bespoke behemoths cooled with high-temperature and high-pressure water, it seems as though simpler and smaller reactors cooled with low-pressure liquids (metal or salt), stamped with advanced safety features, and produced on an assembly line might be a better way to go. Enter the American SMR (small modular reactor).
I’m an easy convert to the idea of SMRs – I spent four years living, eating, working, and sleeping less than 200 feet from an SMR; it was powering the submarine where I spent those years. Before that, I’d spent another year learning the care and feeding of a different SMR and two years teaching my sailor students to do the same. When I moved from the training reactor to the one on my submarine, I knew the plant layout, how to operate all of the equipment and systems on my watch stations, and how to record my logs – I was helpful from the very start because both reactor plants were virtually identical. That similarity, like those submarine reactors in the '60s, '70s, and '80s, is one of the biggest things SMRs have going for them – SMRs are essentially identical; if a sailor could operate one, then he could operate all the others in that class of reactor (in my case it was designated an S5W reactor (the 5th submarine reactor plant designed by Westinghouse).
The reactor plants used by the nuclear Navy meet many of the criteria and have many of the advantages of the current generation of SMR designs – including commonality of design and economies of scale, as well as the relative ease of personnel transitioning from one plant to another of the same design. But many of the newer plants have advantages beyond those. Unlike the high-pressure water-cooled reactor I worked on, with its ever-present concerns about leaks of high-pressure and high-temperature water, many of the new designs are cooled with liquid metal or liquid salt that doesn’t need to be under pressure to keep from boiling and that lack of pressure makes the plants easier to design, easier to build, and much less likely to spring a leak.
There’s been a lot of news in the area of SMRs lately – new plants are being proposed or built in the home of fossil-fuel, Texas, along with Wyoming, and other states (e.g., Nebraska) are interested as well, all following in the footsteps of a Chinese reactor that’s well underway. Most of the proposed American small modular reactors (SMRs) are being financed by tech companies to power their data centers and server farms, which are notorious energy hogs. We might not be able to escape technology. Still, at least it can stop belching carbon dioxide and other pollutants as it serves us an endless supply of cat videos, homework, and poorly written articles into our news feeds.
Here, too, SMRs have advantages over the large reactor paradigm: size and resilience. A data center, for example, isn’t likely to need 1500 MW of electricity, and in a remote location, chances are that the surrounding communities won’t be able to use all that energy either. An SMR can be sized to produce the necessary energy, and if the data center grows to the point of requiring more, it can simply add another reactor. This modularity brings up another advantage! If there’s an accident at a large reactor plant containing over 1000 MW worth of fission products, all 1000 MW are taken offline when the reactor shuts down for any reason. Compare that to a cluster of 10 SMRs of 100 MW each – the loss of a single smaller reactor only takes 10% of the cluster’s power offline, and in the event of a serious accident, there’s only a tenth the amount of radioactivity available to be released; a release is less likely with the newer passive safety systems.
Despite my experience in the Navy, I’ve been more or less agnostic about nuclear energy, largely because of all the factors that conspire to make each plant unique, complex, slow to license and build, and expensive. But I’m cautiously excited about SMRs because they lack all of those drawbacks. I’m keeping my fingers crossed this time and hoping we finally do nuclear right and that it can finally live up to its potential.
