When the giant winter storm Nemo hit New England two weeks ago, the Pilgrim Nuclear Power Generating Station in Plymouth, Mass., lost outside power for several days. Diesel backups took over operating the reactors' cooling system. Pilgrim has the same kind of reactors that failed less than two years ago at Fukushima, Japan, after an earthquake and tsunami crippled offsite power and emergency back-ups. The Pilgrim incident comes as the U.S. nuclear industry is fighting proposed new safety measures meant for a crisis that might begin exactly this way.
Of the 104 reactors in the U.S., 31 are very like those in Fukushima, that lost power, melted down and exploded. I'm entering one of them, the Dresden Nuclear Station, about an hour southwest of Chicago.
View GE Mk I & Mk II Reactors in the U.S. in a larger map
After Fukushima, the Nuclear Regulatory Commission studied what happened. Should it require new safety measures here, even though a crisis is very unlikely?
It's not zero," says Charles Casto, director of NRC Region III, the Midwest. "Hah. The probability's not zero; it's something."
Region III has about two dozen reactors. We spoke at Casto's office. Nuclear regulation, he said, is about possibility more than probability.
"You take your best -- based on history...you know, what has history shown you that the probability would be?" Casto said. "But that doesn't mean zero."
The Fukushima reactors, and their 31 U.S. cousins, including Dresden and the Pilgrim Station, are old Mk I and II boiling water reactors, built by General Electric. The safety enclosures for the reactors are too small. If their cores start to melt down, the containments could fail in several ways, including radioactive hydrogen gas building up and exploding ...as at Fukushima. There's an increasingly politicized dispute between the industry and the NRC over how to make preventing meltdowns safer.
Dresden, the first commercial nuclear plant in the country, is operated by the largest nuclear energy company in the U.S. -- Exelon Generation. It looks dated but extremely well maintained. It hums.
Gregory Roach, the senior NRC inspector here, showed me back-ups on backups on backups, flood protection, fire protection. An hour into the tour, we came to the part I most wanted to see -- the venting system.
"We showed you where the hardened vents were for the dry well, so now this is downstream," Roach said.
He was pointing to a pipe overhead that exits the container wall. The vent system...if power fails, and the back-ups fail, and radioactive hydrogen builds up, you can probably save the reactors by venting the hydrogen.
"And this pipe goes up to the main stack, 300 feet, and then releases into the environment," Roach said.
The Japanese vents mostly didn't work; that's what caused the explosions.
"So you learn a lesson of Fukushima Daiichi," Casto said. "And then you put in standards and say your vent must be able to operate under those conditions."
A year ago, the NRC issued a new order: vents must be reliable.
"Does that mean during an accident you have to be able to get access to it?" Casto asked. "Does that mean you have to do it remotely?
The industry and the NRC are working on it; the industry has until the end of 2016.
Now, here's what the fight is about: do the vents at these 31 plants also need filters? Because, in the best of conditions, some radioactive gas and particles may escape in venting -- and in the very, very unlikely worst of conditions, a lot could escape.
"You have to establish the 'what if'," Casto said. "What if it does happen? What if the improable happens?"
Reactor cores hold dozens of tons of radioactive material. At Pennsylvania's Three Mile Island in 1979, about half the core melted, but the containment held, and the venting was relatively minor. We don't yet know how much escaped from Fukushima; it's too dangerous to go look. But there are scenarios in which it's possible to lose a good part of those dozens of tons through the vents. With filters, virtually all of it is captured.
To a layperson, nuclear regulation can be almost as daunting as nuclear physics. The Dresden plant is a wonder of machine technology -- in it I got reacquainted with the idea of awe.
The NRC process for new rules...that's messier.
The chances of a reactor ever needing a filter are so small that you can't justify the cost. But the NRC staff concluded that the consequences of no-filter could be so bad, they should be required anyway. By NRC procedures, if the staff wants to override normal cost-benefit standards, the five commissioners have to vote to approve, and the fight is on.
A week after the NRC staff testified for filters in January, 21 House Re-publicans on the Energy and Commerce Committee sent the NRC a letter admonishing the staff. There is plenty of money at stake.
"External filter vents would be an additional approximately $15-20 million per unit," said David Czufin, an engineer who runs the Dresden plant for Exelon.
Exelon has 10 more of these reactors, so filters could cost the company more than $200 million -- on top of many other precautions, he said.
"Additional connections, additional equipment, stored equipment for readiness..."
Exelon does plan to spend $400 million in the next three years for post-Fukushima modifications, some of that for vents -- but not filters.
"The one thing I would tell you is that the Fukushima event, we have learned a lot from it," Czufin said. "We continue to look at how the plant in Japan was operated differently from my plant."
He's right; they are different. The independent Japanese commission on Fukushima says so: U.S. plants are much better regulated, run and prepared. And nuclear is already at a competitive disadvantage, according to Bloomberg New Energy Finance, natural gas plants can produce electricity for about 40 percent less. If the industry loses this one, it's going to hurt.
A last question for the director of NRC Region III, Charles Casto. Is he satisfied now that under circumstances like Fukushima Daiichi, he has the technology to operate vents at the stations that are under his authority?
"Yes, we believe that we have reliable vents and that the operators can do a controlled vent in a reliable way during an accident," Casto said.
But unusual things do happen, as at the Pilgrim Station. Without the filters, there is a very small chance that those vents might become a kind of radioactive fire hose. That's what the fight is about.
The NRC Commissioners are voting on this now -- a process that can take weeks, or longer.