Only two U.S. nuclear sites are in compliance with federal fire regulations. How confident can we be that the Nuclear Regulatory Commission has things firmly in hand?
On an ironically clear and placid day in August 2007, a three story tall cooling tower at the Vermont Yankee nuclear power plant collapsed, goring a massive hole in the center of the structure and spewing asbestos, rotting wood, plastic panels, and thousands of gallons of water onto the bank of the Connecticut river. It later emerged that several employees had expressed concerns about the tower, and that, in the days before the collapse, others heard odd noises from within the structure.
Plant administrators refused to allow reporters onto the property for three days. They insisted the tower was of minimal import, and continued running the reactor.
Nine days after the collapse, Yankee was forced to make a full emergency shut down – known as a SCRAM at boiling water reactors, like those at Japan’s crippled Fukushima plant and Vermont’s Yankee - after a critical reactor valve failed. The proper lubrication had been neglected.
“Our nuclear plants are like snowflakes, they’re all different and they can all melt.”
—former NRC commissioner
The plant’s response mirrored its reaction to a transformer fire, in 2004. In June of that year, as a stream of dense, chalky smoke billowed from the compound, the governor’s office became irate with the plant for not providing sufficient information, according to an official who worked for the state at the time. A month later, the plant was cited by the NRC for “Failure to make timely notification of status upon declaration of unusual event.”
The narrative seems more befitting a Soviet-era republic than modern New England, but strings of jarring failures are what many in southern Vermont, and those across the river in neighboring New Hampshire, have come to expect from Yankee.
In April 2004, the plant announced that two fuel rods were unaccounted for, only to find them three months later in the spent fuel storage pool. In August 2006, materials leaving the plant were four times more radioactive than federal limits permit. In 2008, radiation exposure forced the evacuation of 25 employees after a fan was placed too close to the top of the reactor vessel. In May 2008, a crane dropped a 360,000-pound spent fuel cask four inches onto a concrete floor, and, in August 2009, administrators admitted that they had failed to monitor casks for radiation leaks since they had been installed the previous June. In January 2009, the plant sprung two radioactive leaks, one in a “safety sensitive area.” The next month, administrators excavated 135,000 cubic feet of contaminated soil from the plant. And in September 2008, after repairs to the collapsed cooling tower were completed, concerns about further possible rot and corrosion at the plant emerged. It also became clear that Entergy, the New Orleans based corporation that runs Yankee (as well as New York’s Indian Point plant, and eight other nuclear sites), had failed to make renovations to the two cooling tower cells that are critical for an emergency shut down.
The ongoing failures at Japan’s Fukushima plant have put new emphasis on concerns about nuclear safety here in the United States. And on the heels of recent massive regulatory failures that left our financial markets in chaos and the Gulf of Mexico blackened, one wonders whether American nuclear regulation, handled by the Nuclear Regulatory Commission (NRC), in Rockville, MD, is comparably insufficient.
An NRC Investigator General report from 2002, I found, showed that only 53 percent of the employees at NRC felt it was “safe to speak up in the NRC,’ and less than half -- 48 percent -- felt “that management actually trusts the judgment of employees at their level in the organization.”
A 2006 report observed: “Significant reservations still exist about the Differing Professional Opinions (DPO) program. Some employees feel comfortable raising an issue and going through the DPO process. However, a number of employees do not feel comfortable doing so, out of fear of retaliation.”
One former NRC commissioner, speaking on background last June, told me: “our nuclear plants are like snowflakes, they’re all different and they can all melt.”
According to John Grobe, the Associate Director for Engineering and Safety Systems at NRC, “Approximately one-half of the core damage risk at operating reactors results from accident sequences that initiate with fire events.”
Nonetheless, only two U.S. nuclear sites are in compliance with federal fire regulations; all others continue to operate with exemptions, a stopgap system that was implemented more than thirty years ago and was never intended as a permanent solution.
At a public meeting in July 2008, NRC commissioner Gregory Jaczko, whom President Obama has since appointed to chair the commission, noted: “[these are] simple, straightforward regulations and I don’t think there is one plant right now that is in compliance with those regulations.” He continued, “That's simply unacceptable for a regulator.”
The problem, he pointed out, is a longstanding one: “We have never really been able to have a clear set of criteria that we enforce as a regulatory body in fire protection. To this day, I do not think we do.”
His assessment remains accurate.
He outlined a key piece of NRC Title 10 fire regulations Appendix R, set fourth by the NRC in 1980, dealing with the fire protections for reactor control cables.
“You have to have a three hour fire barrier,” around cables connecting control rooms to reactors, “you have to have an hour [barrier] with fire suppression, or you have to have 20 feet of separation. And if all of those fail, you have to have an alternate shut down capability.”
Plants today, rather than relying on automated fire detection and suppression systems, are using a patchwork of manual actions, that is actions by plant workers outside of control rooms to enact dozens of shutdown procedures, in the event of a serious fire: “The reality is that we have licensees that are using unapproved operator manual actions,” Jackzo said in 2008. The NRC declined to make the chairman available to discuss the matter.
Reactors rely on two sets of control cabling, a primary and a backup. Close to a thousand miles of such cables run through plants. The problem is that reactor designs do not physically separate the two sets of lines, and a disaster that damages one will likely affect the redundant system, as well.
Nuclear supporters hold that plants cannot afford to meet the regulations that Jaczko outlined. Voices within the industry have wisely refrained from arguing this explicitly, and prefer to attack the regulations as overkill that fail to identify and address genuine problem areas. As one industry advocate put it, the regulations make “no distinction between a room with a tank of diesel and a room that’s bare concrete and steel.”