Microsoft recently signed an agreement with Constellation Energy to restart the 50-year-old Unit 1 nuclear power plant at Three Mile Island. Microsoft would be the only customer, buying all the electricity to power their data centers. Three Mile Island Unit 2 was the one that melted down in 1979, Unit 1 was shut down in 2019 because the electricity produced was too expensive and the plant not profitable. Constellation estimates it will cost $1.6 billion to restart the plant. They are asking you, the US Taxpayer, for a $1.6 billion loan guarantee since there is a lot of risk in doing something that has never been done before. If it succeeds, Constellation gets the profit. If it fails, you pay to clean up the mess. Pretty sweet deal. In addition, you will pay them $200 million per year in federal tax credits on the sale of that electricity to make it less expensive for Microsoft and more profitable for Constellation. Over the 20-year deal, that’s another $4 billion. Yes, renewables get subsidies, too, all power generation does. Renewable electricity, however, is put out on the grid and benefits everyone, but in this case your money will go directly to Constellation and Microsoft.

It is no coincidence that nuclear power is being sought by some of the largest and most profitable companies in the world. They need lots of electricity and can afford the high cost of nuclear power, as long as you help them pay for it.

The fallout (pun intended) from nuclear accidents like Three Mile Island, Chernobyl and Fukushima put a damper on new construction worldwide. Nuclear proponents argue that this fear is misplaced, and perhaps it is. Nuclear reactors are statistically very safe, at least until something goes wrong. Nuclear currently provide 18% of US electricity, slightly less than renewables which provide 21%. The average age of US nuclear plants is 42 years, two years beyond their original 40 year design life.

There have only been 5 new reactors built in the US in the last 35 years. The newest reactor, Vogtle in Georgia, took 11 years to build. It was 7 years late and $17 billion over budget. Lazard, a global energy consulting firm, put the price of electricity from Vogtle at $190/MWh, twice as expensive as electricity generated from wind, solar, or fossil gas. It is cost, not fear, that keeps nuclear power in decline.

The nuclear industry has consistently overpromised and under-delivered with cost overruns and long delays. This is not unique to the US. Hinkley Point C is a two reactor plant under construction in the UK, their first new reactor in 30 years. First conceived in the 1980s, construction finally began in 2017 by the French firm EDF with a planned completion of 2025 at a cost of $25 billion. In 2024, EDF announced that estimates for the final cost could rise to $59 billion and be delayed to 2029-2031. Like Vogtle, the electricity eventually generated by Hinkley Point will be the most expensive in Britain.

The answer, say the nuclear power proponents, is a new modern design of small modular reactors, or SMRs, that can be built on assembly lines. By shrinking the size and simplifying the design, it is claimed, they can be cranked out like donuts with thousands spread all over the country. (Whether it is wise to sprinkle thousands of unproven nuclear reactors around is not mentioned.) Yet, here again, the nuclear industry has over-promised and failed to deliver.

Small nuclear reactors are hardly new. They were tried before in the 1960s. Several were built and most promptly shut down because they were unreliable and the electricity they generated was two to three times more expensive than other non-nuclear sources.

Yet, here we go again. NuScale Power has the only SMR design licensed to operate in the US. In 2017, anticipating approval, NuScale announced a power plant in Idaho at a cost of under $3 billion (which included a $1.36 billion subsidy from the Department of Energy). The electricity generated would be bought by a western consortium of municipal power producers. In 2021, the inevitable cost issues with nuclear power surfaced and customers started to dropping out. By 2023, costs had ballooned from $3 billion to over $9 billion and the project was canceled. Since 2015, the federal government (that means you) has given $586 million to NuScale, which has not yet produced one kilowatt of electricity.

Undeterred, another small reactor design is now being touted as ‘the next great technology’, the Natrium reactor from TerraPower, which plans a new reactor in Wyoming. At a projected cost of $4 billion, the reactor is backed by a $1 billion investment from Bill Gates, another $1 billion from some of his billionaire friends, and a $2 billion subsidy from you, courtesy of the US government. Unfortunately, it is not small, nor modular and won’t be ready for at least 7 years. Bill Gates has admitted that this project will be expensive and the electricity uncompetitive. His solution? “We have discussions with utilities about building tens of these, but, you know, we really only have a huge impact and success if we get past 100.” That would more than double the total number of reactors currently operating in the US.

Mr. Gates is referring to a concept called the Learning Curve Effect, known as Wright’s Law. It states that whenever production of a unit doubles, the required labor and cost for the next unit drop. Common sense suggests that the more you make of something, the easier and cheaper it gets. While the nuclear industry struggles with one-off designs, with dreams of making 100 that will take decades to build and require billions in federal help, the solar, wind and battery storage industry is deploying millions of units annually, delivering electricity to the grid, with costs dropping with every one made. Despite all the SMR happy talk, nuclear power is doomed to be an expensive niche supplier of an ever decreasing amount of our electricity, not just because of safety, but because of competition.

During the conference call announcing the termination of the UAMPS project in November 2023, NuScale’s Chief Executive Officer explained the decision by saying: “Once you’re on a dead horse, you dismount quickly. That’s where we are here.” The metaphor of a dead horse seems appropriate for small nuclear reactors.

Tom Meara received his engineering degree from Gannon University. He is now retired, having spent his career with GE and Hewlett Packard. He is a Jamestown resident.