ASHINGTON, Sept. 7 — The emergency cooling systems that are
meant to protect nuclear reactors from melting down in case of a
ruptured water pipe could fail after a few minutes of use at most
reactors, according to a nuclear watchdog group that is citing a
government study to argue that the problem makes a catastrophe at
one power plant in New York 100 times more likely.
The group, the Union of Concerned Scientists, and a New York
environmental organization, Riverkeeper, plan to petition the
Nuclear Regulatory Commission this week to ask that the two Indian
Point reactors in Buchanan, N.Y., on the east bank of the Hudson
River, should be shut until corrections are made. The problem, they
argue, is that leaking water or steam would scour off pipe
insulation, paint and other materials, forming debris that would
clog the coolant pumps.
The Nuclear Regulatory Commission recognized the possibility
years ago, and in September 1996 classified it as a serious problem,
but does not anticipate that corrective action will be completed
until early 2007. A commission official said, however, that the
problem is complicated to solve and need not be fixed immediately
because the accident that would require use of the safety system was
unlikely in the first place.
David Lochbaum, a nuclear engineer with the Union of Concerned
Scientists, contended that the emergency core cooling system "is
virtually certain to fail at some plants."
"Right now you're relying on a pipe not breaking," he said.
According to Mr. Lochbaum and to data from the Nuclear Regulatory
Commission, the problem involves 69 plants of a design called
pressurized water reactors, in which the water that is used to carry
off the useful heat, and to keep the fuel from over-heating, is kept
at a pressure of about 2,200 pounds per square inch. If a pipe
breaks and the pressure is released, the water would boil into steam
because it is heated to more than 500 degrees. The steam could not
cool the fuel, and the fuel would melt.
So the plants are equipped with an automatic emergency core
cooling system. Drawing water from a tank outside the reactor dome,
the system can dump thousands of gallons a minute into the reactor,
making up for even a large leak.
In this design, water from a broken pipe would flow into the
reactor basement. The outdoor tank typically holds 125,000 to
300,000 gallons, and when it was nearly empty, the system would
start drawing water from the basement instead. The problem is that
if the water picks up debris along the way, that debris could clog
the screens over the pipes that lead back to the emergency
At the request of the commission, the Los Alamos National
Laboratory studied the 69 plants, and found that for some, the risk
of core damage was multiplied 100 times because of the debris
problem. It ranked the plants but did not name them; Mr. Lochbaum's
group used various detailed characteristics included in the report
to determine which plant was which, and discovered that the Indian
Point reactors were both in the worst five.
The plants' owner, Entergy, told the
N.R.C. in August, in response to a letter sent by the commission to
all plants, that it had analyzed the material available to become
debris, including "failed paints," and would train its operators in
ways to manage the problem, including pumping water in more
A spokesman for Indian Point, Jim Steets, said that he had not
seen the petition, but that "the N.R.C. has attached some level of
urgency, which we're complying with."
At the N.R.C., Sunil Weerakkody, the section chief for fire
protection and special studies, said that in decades of nuclear
plant operation, the emergency core cooling system had been used
only eight times, and that no accident had reached the stages at
which pumping from the basement was required.
"Our best knowledge says we won't even need that function," Mr.
The commission recognized the problem at Davis-Besse, the reactor
near Toledo, Ohio, where operators discovered that the vessel head
had been eaten away by acid, nearly all the way through. Had the
vessel ruptured, said Mr. Lochbaum and others, it would have blown
insulation off the head and into the basement and the screens. The
commission required the plant's owners to fix the problem before it
would consider giving permission to restart.
In an interview, Mr. Lochbaum pointed to that example, and to a
reactor in Michigan that was ordered to shut a few years ago because
it was found to have an unusually high potential for debris, as
precedents for the order he is seeking.
Studies for the N.R.C. found that Indian Point 2 could exhaust
the water in the external tank in less than 23 minutes, and unit 3
in less than 14, he said; Davis-Besse, in contrast, would have taken
35 minutes, and was more likely to survive the debris problem in
many kinds of pipe breaks, according to government data.
But Mr. Weerakkody said the Los Alamos study was only a quick cut
at the problem, to determine whether more work was needed, and had
used various "conservative" assumptions that might be too
"When you remove some of those conservatisms, the risk number
drops drastically," he said.