Leaky language obscures danger

December 1, 2009

By Marlene Lang

I received mail asking about the "leak" at Three Mile Island on Nov. 21. There was no "leak." Chicago-headquartered Exelon Corp. is replacing the steam generator at its Three Mile Island Unit 2 nuclear reactor, as is being done at other old plants around the country. These alloy generators were made to last the life of the plant, 30 to 40 years, and no plan was made back then for how to replace them. Today, these dinosaur plants are applying for license extensions and must replace their ailing parts.

The new TMI generator arrived very recently at its home just south of Harrisburg, Pa., site of the infamous 1979 meltdown inside the Unit 1 reactor.

On Nov. 21, workers were cutting a pipe in the steam cooling system, where radioactive water circulates when the generator is functioning. Somehow, there was a change in air pressure - this is the unexplained part of the incident - and that caused a "wind" inside the pipe, which in turn stirred up invisible radiation dust that was lurking on the pipes, blowing it on workers.

The exposure was low for the 150 workers inside the containment building around the reactor, and no radiation was detected outside the plant.

But it's the fact of the mild exposure being "accidental" that is scary, considering that 3,000 workers are coming and going on Three Mile Island for the generator replacement project.

And bear in mind there is on-site high-level radioactive waste storage on that island for spent uranium fuel.

This whole matter of keeping old plants operational beyond their intended functional life is iffy. Three Mile Island received a 20-year extension of its operating license from the NRC in October.

A plant in Crystal River, Fla., also is replacing its old steam generator, and workers there on Nov. 23 found a gap as they were cutting through the dome that encloses the reactor to contain radioactivity - a 4-foot-thick lead-and-concrete structure. Workers discovered a 60-foot crack, 2 inches wide in some places. Owner Progress Energy doesn't know if the opening was caused by the cutting action, or if it existed before the steam generator replacement project began.

This containment dome is supposed to withstand natural disasters. We are assured the Nuclear Regulatory Commission will conduct a special inspection to figure out how that gap got there.

The aging Beaver Valley plant, 25 miles outside Pittsburgh, received its license extension just weeks ago. And even as I wrote this column on Nov. 24, the plant declared an "unusual event." According to an initial report, there was an internal system leak. "No radioactive release occurred," but a valve opened under pressure during the shutdown of a heat removal system, and it remained "lifted" for nine minutes, leaking radiation into a relief tank. The relief tank did not leak. What a relief.

THIS IS PART ONE OF A TWO-PART COLUMN. PART TWO WILL RUN WEDNESDAY.

Nuclear regulators tippy-toe around problems 

By Marlene Lang

The following is part two of a two-part column. Part one ran in Tuesday's paper.

If I question the presence of "tee th" in the Nuclear Regulatory Commission's oversight of the nuclear power industry, it is with good reason. I've read hundreds of inspection reports as part of a monthslong freelance writing project related to the 30th anniversary of the March 1979 accident at Three Mile Island.

Most of the code violations identified by NRC inspectors at nuclear power plants, I've found, are not immediately dangerous to workers or to the public. The many minor incidents and "findings" - for which the plants are rarely cited - earn labels "of very low safety significance."

As long as no huge disaster actually results, the findings get "Green" ratings, the lowest on a color scale. White is next on the scale of danger, representing "low to moderate risk." Yellow means the safety risk is "substantial" and red, of course, means "high."

I've never seen a Yellow or Red. Exelon's Peach Bottom plant actually earned a White hand-slap in 2008 after a long investigation of its problem with contracted security guards sleeping on the job and after the NRC determined the culture in the plant was such that people were afraid to report dozing guard sightings to their managers.

Another unusual White citation was given to Entergy Nuclear's Vermont Yankee plant in 2006. The plant shipped a contractor's machine that is used to crush radioactive uranium fuel rods before they are stored in pools outside plants. The big box arrived by flatbed truck at a Pennsylvania nuclear plant, giving off four times the amount of radiation allowed by the Department of Transportation. The NRC later reported finding a highly radioactive sliver of metal and two small "hot particles" had fallen inside the crusher. The "White finding" meant the plant was under beefed-up NRC oversight of its shipment practices for a whole year.

NRC reports are full of glossy language that tippy-toes around the fact of screw-ups, which, in NRC-speak, is "the human performance attribute." In literature classes, we called this a euphemism.

It is much more palatable to speak of an "inadequacy" being "associated with the human performance attribute" than to blurt out that our safety from nuclear radiation seesaws precariously upon the fulcrum of human competence or human error.

Here is an example, from a routine inspection report, pulled with little fuss from my own files. I'd describe it as typical. The NRC handed down a no-citation "barrier integrity" violation in 2006 to the Peach Bottom nuclear power station in Delta, Penn., just downriver from Three Mile Island.

Stop and think about "barrier integrity." It means preventing leaks.

The report explains, or rather, obscures, the fact that stuff was found in a reactor valve; stuff that wasn't supposed to be there.

"Personnel inadequately accomplished foreign material exclusion recovery procedures," preventing a "primary containment isolation valve closure" on April 5, 2006, according to the NRC.

Deeper in the report - you have to stay on task and dig for actual details - I read that an alert inspector had noticed a valve had appeared partially opened while control room lights indicated it was closed. The valve was dissembled, and "four pieces of metal resembling shim stock and a 3-inch length of 1 / 8 -inch diameter weld rod" were recovered from the valve body after they prevented the valve from closing.

While I am not an expert on these systems, I've read enough to understand that these valves are part of an enclosed cooling system and that cooling is very important when nuclear power is being produced, as is enclosure. Bad things happen if the reactor gets too hot or if the closed system finds an opening.

Again, listen to the NRC's dissociative language: "The finding is more than minor because the failure of a containment isolation valve to close is associated with the Barrier Integrity Cornerstone attribute of systems and component performance and affected the objective to provide reasonable assurance that physical design barriers (containment) protect the public from radionuclide releases caused by accidents or events. The finding was determined to be of very low safety significance, since the finding did not represent an actual open pathway in the physical integrity of the reactor containment."

We the public who are being safeguarded from radionuclide releases need to remember at all times that this is an industry with a very low error tolerance and that "minor" safety infractions are most often the result of the "human performance attribute."

Often it is little more than luck that prevents an "actual open pathway in the physical integrity of the reactor containment" - also known as: a leak.

Marlene Lang's column can be found at: www.southtownstar.com/news/lang/index.html