The future of nuclear power will hinge on beating the economics of emerging rivals, renewable power and gas, even as it faces safety concerns after the Fukushima meltdown a year ago.
Nuclear is facing more cautious growth in developing countries while fighting for its share in industrialised nations where the present fleet is nearing retirement.
That contrasts with earlier optimism, illustrated in a pamphlet published last year by the World Nuclear Association, the industry's global trade association, before Fukushima struck.
Nuclear power was "unparalleled in its capacity to generate electricity cleanly, safely and on a large scale ... because its fuel will be available for multiple centuries, its safety record is superior among major energy sources, its consumption causes virtually no pollution ... (and) its costs are competitive and still declining," it said.
Fukushima cast doubt on that safety.
Perhaps more significantly, plummeting costs of rival energy sources threaten its economics.
The key advantages of nuclear are that it supplies baseload (constant), low-carbon power, while a plutonium fuel source could last for centuries.
But the discovery of vast shale gas reserves in the United States plus rapid development of wind and solar power have discounted the value of a bountiful fuel.
Meanwhile gas is also cheaper (barely half the cost per unit of electricity for new U.S. plants, according to the Energy Information Administration), as is onshore wind, and solar catching up.
What's left is a slender advantage on narrow terms, beating the CO2 emissions of gas and intermittency of wind and solar - making progress on costs so vital.
Is it safe?
Was Fukushima a deadly warning (as green groups say), or unique (as many nuclear engineers assert)?
On March 11, 2011, a massive earthquake unleashed a 14-metre high tsunami which swept over protective walls at the Fukushima plant knocking out backup power and halting cooling of its reactors.
Engineers grappled to control overheating at three steel reactor cores, venting steam along with hydrogen and radioactive material into the surrounding concrete housing.
But the hydrogen ignited causing explosions which damaged buildings and leaked radioactive caesium into the surrounding countryside, dislocating nearly 80,000 people many of whom do not know when or if they can return.
Some experts have downplayed the wider implications for the industry: the earthquake and tsunami were unusually large; flooded, backup diesel generators were located in basements; and spent fuel ponds were also poorly positioned, vulnerably close to the reactors.
Other accidents have other explanations: at Three Mile Island in the United States in 1979, instrument error led to a loss of coolant, overheating, partial meltdown and a contained release of radioactive material. The lesson was a more rigorous protocol for dealing with cooling accidents.
At Chernobyl in Ukraine in 1986, a cavalier handling coupled with a "lethal design", where nuclear fission could continue even during an emergency, led to massive overheating, a fire and deadly explosion which ripped open the reactor core.
Experts will point to a much worse record in coal deaths, from mining and air pollution.
Nevertheless, technical explanations won't resurrect public faith that nuclear is completely safe.
Nuclear technology's future rests firmly on costs, where it already looks vulnerable: high costs reflect the size and complexity of reactors, and dependence on skilled labour, specific supply chains plus public or at least state planning support and permissioning.
The World Nuclear Association (WNA) estimates the upfront capital cost for nuclear at about $2,000 per kilowatt in east Asia, and $4,000-5,000 in the West.
That excludes financing, waste disposal, fuel and decommissioning costs, as well as those difficult to monetise risks around planning control.
Could stiffer carbon policies, penalising fossil fuels, help? Maybe, but low-carbon alternatives are emerging or catching up, including wind, solar, carbon capture and storage and biomass.
Meanwhile the next nuclear reactors in the developed world at least will be especially expensive, given a long construction decline and a resulting atrophying of talent and supply chains.
The next test will be a handful of new plants in Britain and larger rollouts in China, India and Russia, proving they can avoid budget overruns, delays (which have blighted the latest plants in France and Finland) and accidents.
This story was originally published by Reuters. Reproduced with permission.