Potential Cost Of A Nuclear Accident? So High It’s A Secret!

Share on FacebookTweet about this on TwitterShare on LinkedInShare on Google+Share on RedditPrint this pageEmail this to someone

Catastrophic nuclear accidents, like Chernobyl in 1986 or Fukushima No. 1 in 2011, are very rare, we’re incessantly told, and their probability of occurring infinitesimal. But when they do occur, they get costly. So costly that the French government, when it came up with cost estimates, kept them secret.

But now the report was leaked to the French magazine, Le Journal de Dimanche. Turns out, the upper end of the cost spectrum of an accident at a single reactor at the plant chosen for the study, the plant at Dampierre in the Department of Loiret in north-central France, would amount to over three times the country’s GDP. Financially, France would cease to exist as we know it.

Hence, the need to keep it secret. The study was done in 2007 by the Institute for Radiological Protection and Nuclear Safety (IRSN), a government agency under joint authority of the Ministry of Defense and the Ministry of Environment, Industry, Research, and Health. With over 1,700 employees, it’s France’s “public service expert in nuclear and radiation risks.” This isn’t some overambitious, publicity-hungry think tank.

It evaluated a range of disaster scenarios that might occur at the Dampierre plant. In the best-case scenario, costs came to €760 billion—more than a third of France’s GDP. At the other end of the spectrum: €5.8 trillion! Over three times France’s GDP. A devastating amount. So large that France could not possibly deal with it.

Yet, France gets 75% of its electricity from nuclear power. The entire nuclear sector is controlled by the state, which also owns 85% of EDF, the mega-utility that operates France’s 58 active nuclear reactors spread over 20 plants. So, three weeks ago, the Institute released a more politically correct report for public consumption. It pegged the cost of an accident at €430 billion.

“There was no political smoothening, no pressure,” claimed IRSN Director General Jacques Repussard, but he admitted, “it’s difficult to publish these kinds of numbers.” He said the original report with a price tag of €5.8 trillion was designed to counter the reports that EDF had fabricated, which “very seriously underestimated the costs of the incidents.”

Both reports were authored by IRSN economist Patrick Momal, who struggled to explain away the differences. The new number, €430 billion, was based on a “median case” of radioactive releases, as was the case in Fukushima, he told the JDD, while the calculations of 2007 were based more on what happened at Chernobyl. But then he added that even the low end of the original report, the €760 billion, when updated with the impact on tourism and exports, would jump to €1 trillion.

“One trillion, that’s what Fukushima will ultimately cost,” Repussard said.

Part of the €5.8 trillion would be the “astronomical social costs due to the high number of victims,” the report stated. The region contaminated by cesium 137 would cover much of France and Switzerland, all of Belgium and the Netherlands, and a big part of Germany—an area with 90 million people (map). The costs incurred by farmers, employees, and companies, the environmental damage and healthcare expenses would amount to €4.4 trillion.

“Those are social costs, but the victims may not necessarily be compensated,” the report stated ominously—because there would be no entity in France that could disburse those kinds of amounts.

Closer to the plant, 5 million people would have to be evacuated from an area of 87,000 square kilometers (about 12% of France) and resettled. The soil would have to be decontaminated, and radioactive waste would have to be treated and disposed of. Total cost: €475 billion.

The weather is the big unknown. Yet it’s crucial in any cost calculations. Winds blowing toward populated areas would create the worst-case scenario of €5.8 trillion. Amidst the horrible disaster of Fukushima, Japan was nevertheless lucky in one huge aspect: winds pushed 80% of the radioactive cloud out to sea. If it had swept over Tokyo, the disaster would have been unimaginable. In Chernobyl, winds made the situation worse; they spread the cloud over the Soviet Union.

Yet the study might underestimate the cost for other nuclear power plants. The region around Dampierre has a lower population density than regions around other nuclear power plants. And it rarely has winds that would blow the radioactive cloud in a northerly direction toward Paris. Other nuclear power plants aren’t so fortuitously located.

These incidents have almost no probability of occurring, we’re told. So there are currently 437 active nuclear power reactors and 144 “permanent shutdown reactors” in 31 countries, according to the IAEA, for a total of 581 active and inactive reactors. Of these, four melted down so far—one at Chernobyl and three at Fukushima. Hence, the probability for a meltdown is not infinitesimal. Based on six decades of history, it’s 4 out of 581, or 0.7%. One out of every 145 reactors. Another 67 are under construction, and more are to come….

Decommissioning and dismantling the powerplant at Fukushima and disposing of the radioactive debris has now been estimated to take 40 years. At this point, two years after the accident, very little has been solved. But it has already cost an enormous amount of money. People who weren’t even born at the time of the accident will be handed the tab for it. And the ultimate cost might never be known.

Share on FacebookTweet about this on TwitterShare on LinkedInShare on Google+Share on RedditPrint this pageEmail this to someone