Few energy questions divide thoughtful people as sharply as nuclear power. Some environmentalists call it essential to stopping climate change; others call it a dangerous, costly distraction. This level sets out each case as its best defenders would make it.

A short history

Fission releases energy when a heavy nucleus splits in two. In 1954 a small reactor at Obninsk in the Soviet Union became the first nuclear plant to supply electricity to a power grid. Britain's Calder Hall followed in 1956 and the American plant at Shippingport, Pennsylvania, in 1957. Officials of the era promised power so cheap it would hardly be worth measuring. Construction boomed in the 1960s and 1970s, then slowed sharply after the accident at Three Mile Island in Pennsylvania in 1979 and the far worse disaster at Chernobyl in 1986. Today nuclear plants supply roughly a tenth of the world's electricity, from about four hundred reactors in some thirty countries.

The case for

Supporters start with climate. Counted over their whole life, nuclear plants emit about as little carbon per unit of electricity as wind farms. They run around the clock whatever the weather and use little land. Sweden and France cut the carbon in their electricity faster than almost any other countries, largely by building reactors alongside hydropower.

Supporters also argue that fear has outrun the evidence. Measured by deaths per unit of electricity produced, nuclear power is among the safest sources, far safer than coal, even when Chernobyl and Fukushima are counted. Air pollution from burning coal kills more people in a typical year than nuclear accidents have in the whole history of the industry. In this view, closing reactors early usually means burning more fossil fuel and causing more deaths, not fewer. They add: keeping an existing, safe reactor running is usually far cheaper than building a new one, because the building cost has already been paid.

The case against

Critics start with money and time. Recent reactors in Europe and the United States have run years late and far over budget. Olkiluoto 3 in Finland began construction in 2005, was meant to open in 2009, and began regular operation only in 2023. The two new units at Vogtle in Georgia finished in 2023 and 2024, years late and at more than double their expected cost. Critics argue that every dollar and every year spent waiting for a reactor could cut more carbon if spent on wind, solar and efficiency.

They also argue that low death counts understate the harm of accidents. After Fukushima in 2011, more than a hundred thousand people were evacuated, and the stress of evacuation itself was linked to over two thousand deaths among elderly and sick people. A rare accident can cost hundreds of billions of dollars and shake public trust across the world. Finally, the skills and materials of civilian nuclear power can help a country seeking weapons.

The waste question

Spent fuel stays dangerously radioactive for thousands of years, but its volume is small. It is cooled in pools and then stored in steel and concrete casks. The hard part is permanent disposal. America's planned repository at Yucca Mountain, Nevada, stalled amid state opposition. Finland's Onkalo repository, deep in ancient bedrock, is expected to be the first permanent disposal site for spent fuel, showing that the problem is at least as political as technical.

New designs

Many companies now promise small modular reactors. Supporters expect them to be cheaper because they would be built in factories as many identical units, the same learning curve logic that made solar cheap. Skeptics point out that small reactors lose some economies of size, and that the leading American design suffered a setback when NuScale's planned project in Utah was cancelled in 2023 after its expected costs rose. Meanwhile South Korea and China have built large reactors at far lower cost than recent Western projects, suggesting that steady building with a skilled workforce matters as much as design.

Underneath the numbers lie different values: how to weigh a small risk of rare catastrophe against the steady harm of fossil fuels, and how far to trust the institutions that run dangerous technology.

Worked example: France and Germany

After the 1973 oil shock, France chose to build dozens of reactors in about fifteen years. Nuclear power came to supply around two thirds of French electricity, making it among the lowest carbon in Europe. But in 2022, cracks found in pipes and a drought that limited cooling water forced many reactors offline, and France imported more electricity than it exported for the first time in decades.

Germany went the other way. After Fukushima, its government decided in 2011 to close all its reactors, and the last three shut in April 2023. Renewables grew to supply more than half of German electricity, but coal plants ran longer than they otherwise would have. One study estimated that the phase-out raised carbon emissions and caused around a thousand extra air pollution deaths a year.

Apply the ideas. France shows the climate benefit of steady nuclear power and the risk of depending on one aging fleet. Germany shows a public choosing to accept more fossil pollution for a time to remove a risk it judged unacceptable. Neither choice was free.