written 31 December 2023
published 7 January 2024
At COP28, the latest United Nations Climate gathering, nuclear power received more attention. Saudi Arabia committed to developing nuclear electrical generation. China is constructing 21 large nuclear reactors. Some people believe a massive nuclear build out will avert the climate crisis. The 436 reactors now operating produce about 10 percent of the global electricity. It would take 10,000 additional reactors to completely decarbonize the global economy.
It is true that an operating reactor produces no greenhouse gases (GHG), but when the whole life cycle of a reactor is analyzed, including construction and fuel enrichment, a standard 1,000MW reactor releases GHG comparable to a natural gas power plant. Even that evaluation is incomplete, as it excludes complete decommissioning of a large nuclear plant (never been done), and long-term storage of high level nuclear waste (not yet done even after 70 years).
Nuclear corporations were blackmailed into business. After the atomic destruction in Japan, the US government wanted a happy face for the atom, so Atoms For Peace promoted "power too cheap to meter". The electrical industry was told to develop nuclear power, or the government would do it, putting them out of business. This was a bluff, but nobody knew it then.
Economically, nuclear power is a bust. Reactors are large, expensive, and centralized, making construction more an art than manufacturing. Costs consistently comes in over budget and behind schedule, making nuclear power more expensive than solar or wind, even including storage. Even operating an existing nuclear reactor is more costly than building renewable projects. While solar, wind, and battery costs are dropping every year, nuclear costs keep increasing. Small modular reactors (SMR), heralded as the salvation of the nuclear industry, suffer the same cost problems, plus a lack of customers. The only SMR project in the US was just canceled due to cost overruns.
Uranium is a finite commodity, and used inefficiently. A reactor core contains tons of highly processed enriched uranium. After a few years, when only 5 percent of the uranium has been consumed, the core must be replaced. When fission byproducts build up, performance degrades to the point of economic inefficacy. Millions of tons of highly radioactive "spent" fuel are stored at reactor sites. The best uranium deposits have already been developed, leaving only poorer quality ore. Most low level enriched uranium comes from Russia.
But the real economic costs come when a reactor breaks. Designed to last for 40 years, decisions were made in the beginning with incomplete information, with multiple units built on those designs in order to make nuclear construction seem profitable. So far, the worst US designed reactor failures were the 40 year old units at Fukushima, in 2011. Complete cleanup cost estimates are over $1T. Actual repairs have yet to begin, because radioactivity is too high for even robots to function for very long, let alone humans.
The only reactors still operating in California are the 40 year old pair at Diablo Canyon, near San Luis Obispo. Heavy radioactivity embrittles metal, making it more prone to shock failure. Several earthquake faults have been identified near the site, including one right through the plant. PG&E has done embrittlement tests, but refuses to release the results to the public, claiming "proprietary rights". The Diablo Canyon reactors were recently granted a 5-year extension, with no changes required to the existing, aging equipment.
A reactor failure due to a seismic event could affect a large area of central California, from LA to San Francisco and inland to Nevada, depending on which way the wind blows. But PG&E would not be liable for any damages beyond $13B, due to the Price Anderson Act, a sweet heart deal the US made when the nuclear industry began. Every liability insurance policy written has an exclusion for nuclear damages. This all helps the nuclear industry seem profitable.
Nuclear power highlights a fundamental capitalist problem: the conflict between safety and profits. Each reactor is so powerful, that any accident can become catastrophic faster than humans can react. It is so expensive, that the incentive is enormous to cut costs to be more profitable. Add in limited corporate financial liability, and you get a recipe for disaster.
Fukushima shows the "small probability, high impact" nature of a failed nuclear reactor. The economics of even a properly operating reactor fail basic capitalist reasoning. To leave a habitable planet for our descendants, we have to do better.
