We’ve seen the mushroom clouds. We’ve seen men in haz-mat suits wanding down terrified children with Geiger counters at Fukushima. We’ve seen Chernobyl’s lugubrious “zone of alienation,” its vacant towns reclaimed by wolves. We’ve seen umpteen doomsday thrillers about radiation-induced mutation on a scale somewhere between AIDS and a zombie apocalypse. Scary and compelling, these images shape our understanding of nuclear fission, the one force we’ve all grown up believing would destroy civilization.
Such forebodings haunt the controversy over the Indian Point nuclear power plant, perched on the Hudson River 35 miles north of Midtown in the village of Buchanan, ground zero of a 50-mile spewing radius that contains nearly 20 million people. Entergy Corporation, the plant’s owner, wants to renew the licenses for its two functioning reactors, which expire Sept. 28, 2013 and Dec. 12, 2015, respectively. Critics from Governor Cuomo to environmental groups like Riverkeeper vehemently oppose the relicensing, prodded by the conviction that an accident at the plant would turn New York into an uninhabitable wasteland.
To its opponents, Indian Point is the worst place in the world to put the worst threat in the world. But are those fears realistic? What would an accident at Indian Point be like? Do science and history of nuclear accidents elsewhere bear out the visions of radioactive desolation that drive the opposition to Indian Point? Or could New York live with a meltdown at Indian Point rather than die from it?
We need to think hard about these questions, because lurid speculation can box us all into an energy policy that’s aimed at the wrong disaster.
In the legal arena, the battle over Indian Point is more about fish casualties than human ones. The State of New York, joined by Riverkeeper, is fighting the plant on several fronts. To get the Nuclear Regulatory Commission to renew its licenses, Entergy needs the state Department of Environmental Conservation (NYSDEC) to issue it a Water Quality Certification, which NYSDEC pointedly refused to do in 2010, arguing that the plant’s everyday operations and effluents present an ongoing risk to people and wildlife. (Entergy is appealing the decision, and the reactors will keep operating past their license expiration while the process continues.)
The splashiest of NYSDEC’s complaints are that cancer-causing radionuclides from spent-fuel pools, pipes and tanks have leaked into the groundwater under Indian Point and are oozing into the Hudson. The numbers do sound dire: up to 600,000 pico-curies of tritium per liter in some monitoring wells. That’s 30 times the EPA drinking water limit, on top of 13 times the limit for strontium-90, an isotope that accumulates in bone; up to 100 times the limit for cesium-137. But because EPA water standards are extraordinarily strict—and because radiation is a much weaker carcinogen than most people realize—the contamination is actually pretty innocuous.
If you were to drink a half gallon of water containing those contaminants at those concentrations every day for 25 years, you would have a one percent chance of getting a fatal cancer from them, on top of the 20 percent lifetime risk you already face. That would make Indian Point groundwater roughly as carcinogenic as beer. Diluted into the immense volume of the Hudson, those radionuclide concentrations fall to harmless levels far below drinking water limits.
The danger to fish is more substantial. The plant swallows and expels 2.5 billion gallons of water per day to cool its generators, with usually fatal consequences for any creature sucked inside. The result is what Riverkeeper calls “the slaughter of more than a billion fish and other river organisms every year.”
That carnage feels slightly less heart-rending when you itemize the “other river organisms.” Of the 1.3 billion massacred, according to the state, about 1.18 million are adult fish trapped on the intake screens while the other 1.29882 billion are largely “fish eggs and larvae”—also known as “plankton.” That’s a lot, but since a breeding female lays anywhere from a few thousand to several million eggs, it’s not clear how big a dent Indian Point makes in overall fish populations. (Entergy argues that most of those creatures, being far down the food chain, would die before maturity anyway.)
But the public anxiety that drives and hardens policy positions isn’t about fish kills. At the an open meeting on Indian Point last May, a crowd of anti-nuclear activists, politicians and citizens had potential catastrophe on their minds: earthquakes, flash floods and design failures that might someday snowball into a disaster.
Some conjectures are almost as elaborate as a Bruce Willis movie. In his 2007 report for Riverkeeper, Risk-Related Impacts from Continued Operation of the Indian Point Nuclear Power Plant, mathematician Gordon Thompson imagines terrorist squads paragliding in with anti-tank missiles, a 9/11-style plane-crash into a containment building, and a ten-kiloton atomic bomb. (Reactors cannot explode like an atom bomb, so terrorists would have to bring their own.) Public officials take those notions very seriously; the state’s 2007 petition to the Nuclear Regulatory Commission cites terrorist spectaculars and fuel pool fires in its bill of particulars against Indian Point’s relicensing.
But how likely are these events?
Beneath their chaotic particulars, nuclear accidents have two common features. One is a loss of cooling water in the reactor. The fission chain reaction in Indian Point’s reactors would instantly shut down in that event, but the residual decay heat of radionuclides would make the fuel rods heat up and boil their volatile fission products (radioactive iodine and cesium) into the air along with high-pressure steam and explosive hydrogen gas. That sets the stage for the second element of a spew: some physical shock that cracks the steel and concrete containment barriers so that those airborne radionuclides can find an opening to the outside world.
Rare as these once-a-generation mishaps are, no plant is absolutely proof against them. But Indian Point is markedly less vulnerable than, say, the Fukushima Daiichi plant. One decisive reason is that it’s not in a tsunami zone, so it’s beyond the reach of the only natural force that has ever walloped a plant hard enough to cause a meltdown. It also has a surfeit of the emergency diesel generators, fuel tanks and pumps that keep water flowing in a crisis, some stashed high above any possible flood waters, some buried in earthquake-proof settings. The chances of them being knocked out at one blow, as at Fukushima, are vanishingly small.
Then there are the spent-fuel pool fire scenarios, which have replaced reactor meltdowns as the top nuclear nail-biter in the public mind. There’s a logic to that. Densely packed with years-old but still somewhat hot fuel rods, Indian Point’s pools, like those at most nuclear plants, hold much more radioactive material than the reactors do. If they were to spring a leak and lose their cooling water thanks to an earthquake or bomb, the zirconium fuel cladding could heat up and burn. No one knows how big a spew that would cook off, but some argue it could dwarf 2011’s Fukushima and 1986’s Chernobyl releases.
That storyline isn’t completely implausible, but a chain of low likelihoods would have to cooperate. Indian Point’s pools are set in bedrock, partially below ground, with massive volumes of water and walls rated to withstand airplane crashes. (A 2008 safety evaluation by a panel of outside experts, avowedly independent though paid by Entergy, found that the thick reinforced concrete walls of the reactor buildings could withstand the impact of a Boeing 767 airliner at 350 mph without major damage.)
It would take a big leak, from a very big jolt, to outpace the staff’s efforts to spray in water. And contrary to blogosphere lore, the fuel rods would likely not burn even if cooling water drained away. According to a new Nuclear Regulatory Commission study, air circulation would usually suffice to cool the assemblies if they have all been in the pool longer than a few months, which is the case 92 percent of the time. And the other 8 percent of the time a release is still almost impossible unless sprinkler spigots and fire hoses somehow malfunction. Much of the drama at Fukushima surrounded the struggle to keep the upper-story fuel pools full as the water boiled off. That struggle succeeded for the simple reason that spent fuel is spent—cool enough that a fire hose can handle it.
As unlikely as these scenarios are, still, it could happen here: some calamity, familiar or novel, might plunge Indian Point into a meltdown. What happens then?
The most influential answer comes from Edwin Lyman, a physicist and senior scientist with the Union of Concerned Scientists. His 2004 study for Riverkeeper, Chernobyl on the Hudson? The Health and Economic Impacts of a Terrorist Attack at the Indian Point Nuclear Plant, is the bible of Indian Point disaster scenarios. It’s easy to see why, given the study’s mix of apparent scientific rigor and frightening conclusions. Using computer models and data developed by the NRC, Mr. Lyman estimates that an Indian Point spew could kill more than 600,000 people, displace millions of refugees and cost trillions.
Mr. Lyman’s meltdown scenario, involving a plane crash possibly supported by a terrorist ground assault and an inside man, generates a terrifying “peak” body count of a half million. There’s only one problem: his results bear no resemblance to any real nuclear accident.
Let’s start with Lyman’s estimate of “early fatalities”—that is, prompt deaths from acute radiation syndrome (ARS), marked by sloughed-off skin and bloody. Lyman’s average estimate predicts 696 ARS fatalities, while his peak estimate envisions 43,700 of these horrible deaths. But nothing remotely like that has ever happened. At Chernobyl, a total of 28 men died of ARS, most of them tasked with trying to smother the blazing nuclear bonfire; no cases occurred beyond the plant boundary. At Fukushima there were likewise no cases of ARS among civilians—and not a single one among the thousands of emergency workers battling the meltdowns at ground zero.
Lyman’s overprediction of early fatalities flows from aneye-popping overprediction of radiation levels. At a distance of 12 miles from Indian Point, he calculates that the radiation dose in the first week would range from 198 rem in the average case to a peak of 1,490 rem, a uniformly fatal dose. For Midtowners 35 miles from the plant, the exposures would be 30 rem in the average case to a peak 307 rem, a frequently fatal dose. These numbers are drastically higher than any seen at real spews. The most heavily exposed people near Fukushima , in the district of Namie close to the 12-mile evacuation line, got a dose of 2.7 rem over four months. Even at Chernobyl, the average dose among emergency workers was only 12 rem and among evacuees from the exclusion zone just 3 rem.
To put those numbers in context, the UN Scientific Committee on the Effects of Airborne Radiation (UNSCEAR) puts the average lifetime Chernobyl dose in contaminated areas of Belarus, Ukraine and Russia at 900 millirem; the average American’s yearly background dose is 240 millirem. In Colorado, background radiation is about 500 millirems above normal because of high elevations and radon gas.
The biggest civilian Fukushima dose is roughly the additional dose you get from living six years in Denver—or being a flight attendant, which has a higher occupational exposure than working in a nuclear power plant. (Lots of radiation up there.)
Mr. Lyman’s projected “latent cancer fatalities” similarly soar above precedents. His estimates range from 40,000 in the average case to 603,000 in the peak case, even assuming massive evacuations to limit radiation exposures. Compare those numbers with the 27,000 eventual cancer fatalities from the Chernobyl spew estimated by Lyman’s UCS colleague Lisbeth Gronlund, and with Fukushima cancer estimates that range from a few hundred to a couple thousand.
But the cancer cases that scientists actually observe after real nuclear accident are much smaller than those calculated estimates.
UNSCEAR estimates that 6,000 thyroid cancer cases have occurred post-Chernobyl because children drank milk from cows that ate iodine-contaminated grass, which the Soviet authorities didn’t warn people about. (Thyroid cancer is readily treatable, so only 15 deaths resulted.) And Chernobyl cleanup workers who got the highest doses may face an elevated risk of leukemia, an effect that wavers on the brink of statistical significance. Otherwise, UNSCEAR finds “the vast majority of the population need not live in fear of serious health consequences.”
Finally, Mr. Lyman’s study forecasts the permanent relocation of anywhere from 684,000 to 11 million people, with costs of up to $2 trillion.
It’s not clear that even temporary evacuations are necessary in most situations. A study by Stanford’s Mark Jacobson estimates that if the 100,000 people evacuated from the Fukushima 12-mile zone had simply stayed there, forever, the elevated radiation would have caused at most 245 extra cancer fatalities; he notes that the stress of the evacuation itself killed some 600 people. “Sheltering in place”—that’s emergency-speak for “relaxing at home”—can be a remarkably effective response to an accident, especially in New York’s big stone tenements: the EPA manual says that the shielding in my third-floor apartment would reduce my dose from external fallout by a factor of 30. Doing nothing could be the best strategy in a spew.
In defending his study, Mr. Lyman pointed out to me some differences between his scenario and previous accidents. He assumed that terrorists could trigger an almost immediate meltdown and a huge release. His model stipulates that Indian Point would release 67 percent of its radioactive iodine and cesium within a few hours; by contrast, the Fukushima “release fraction” was just a few percent and the spew took a day to get rolling. Chernobyl released about half its iodine and cesium, but Lyman says the graphite-fueled reactor fire—something that can’t occur at Indian Point—produced a chimney effect that lofted the plume higher and therefore diluted it more than might happen in an Indian Point release. It’s debatable to what extent these factors account for the gap between model and reality, but they highlight just how dependent nightmare forecasts are on happenstances that usually break the other way, even in a major accident.
Scary as they seemed, both Chernobyl and Fukushima fell short of the doomsday hype. Chernobyl’s death toll was tragic, but no higher than the toll from many a dam break. Evacuations rendered the area a wasteland, but not an uninhabitable one: thousands of peasants filtered back into the exclusion zone, and the nuclear plant itself kept running its other two reactors with a full workforce for 14 years after the accident.
At Fukushima, UNSCEAR predicts that there will be no discernible health effects from the radiation. In the evacuation zone, radiation levels have declined by two thirds since April of 2011 thanks to radioactive decay and rainfall; evacuees are returning and farmers are growing rice in some places. Last February, scientists from France’s nuclear safety agency spent three days in Fukushima City and made side-trips into the evacuation zone. When they got home, their dosimeters said they had absorbed less radiation than they would have had they spent those days in Paris.
Rather than an Indian Point meltdown, these scenes of flagrant survivability suggest a low-stakes affair with limited contamination that would clear rather quickly, and health effects that would be too small to measure. New Yorkers who worry about such things might move. I would stay, though I might spend even more time huddled in my apartment. If the fallout were, say, four times what the EPA thinks I can handle, that would raise my fatal cancer risk, if the risk factors are right, from 20 percent to 21 percent. The lingering radiation would take its place amid the mild health worries—bus exhaust, West Nile virus, cheeseburgers—that shadow city life. New York, eternal zone of alienation, would feel a bit more alienating, but rents might go down.
Even though the odds are a thousand to one that the plant will ever kill anything besides fish or spew anything more exciting than low-carbon electricity, our fixation on risks obscures the very real environmental and health benefits Indian Point confers by abating air pollution and greenhouse emissions.
Pollution from coal-fired power plants kills 13,000 Americans every year, a toll far worse than nuclear accident scenarios. A recent study by NASA climate scientist James Hansen estimated that nuclear power plants have saved 1.8 million lives by displacing fossil-fueled electricity, and could save a further 7 million by mid-century.
Indian Point’s opponents are right that the plant could be replaced by other power sources, but the alternatives all have unacknowledged costs and contradictions. Riverkeeper touts wind power too, but building enough wind turbines to equal Indian Point’s output would cost $14 billion (not counting backup gas plants for windless spells) and stir ferocious NIMBY opposition, even among environmentalists. (Robert F. Kennedy, Jr., Riverkeeper’s “Chief Prosecuting Attorney,” is a vocal opponent of the Cape Wind project—which just coincidentally would spoil the view from his family’s Hyannis Port estate.)
And all of these schemes miss the biggest picture of all: every electron Indian Point generates is one that’s not generated by fossil fuels.
A few days before the Nuclear Regulatory Commission’s Indian Point meeting, scientists announced that the atmosphere’s carbon dioxide level had reached the 400 parts-per-million milestone. Listening to the assembled environmentalists at the meeting, passionate in their concern for their planet as they denounced the state’s biggest supplier of clean energy, I couldn’t help thinking: this is how we get to 401.