Staying In the comfort zone
ORNL has provided the Nuclear Regulatory Commission with valuable scientific data and computer codes to help the agency decide whether individual nuclear power plants can continue to operate safely.
Multiple gas centrifuge cascades were constructed at the Portsmouth Gaseous Diffusion Plant in Ohio and successfully operated until the program was discontinued in 1985.
America's nuclear power plants, which provide 20% of the nation's electricity, have operated relatively safely since the startup of the first plant in 1957. To ensure the continuity and improvement of these facilities' safety records and to avoid accidents such as the 1979 loss-of-coolant (LOCA) accident at the Three Mile Island plant, the watchdog of the nuclear power industry--the Nuclear Regulatory Commission (NRC)--relies partly on the work of ORNL researchers.
NRC's research program at ORNL, led by Julie Simpson, addresses newer as well as older threats to the safe operation of U.S nuclear power plants. The new threats include electromagnetic interference from wireless technologies and attacks by terrorists and computer hackers. The traditional threats include lightning, human error, and challenges to the integrity of reactor pressure vessels (RPVs), such as corrosion, pressurized thermal shock, and irradiation.
License To Keep Operating
Of the nation's 103 nuclear power plants, 18 plants that currently have 40-year operating licenses have had their licenses extended by NRC, allowing them to operate another 20 years. Data and modeling by ORNL researchers have helped guide the NRC in making these decisions.
Researchers in ORNL's Heavy-Section Steel Irradiation Program, led by Tom Rosseel and Randy Nanstad, both of the Metals and Ceramics (M&C) Division, have analyzed whether the steel in different RPVs has become too embrittled from neutron irradiation to continue safe operation. "Based on our test data and computer modeling," Nanstad says, "it looks like many of the nation's RPVs could last 60 to 80 years instead of the expected 40 years."
"We write reports that help NRC understand the damage mechanisms in irradiated steel containing various elements, such as copper and nickel," Rosseel says. "In this way, NRC can develop standards, codes, and regulations to make sure that nuclear power is safe for the public."
Electric utilities seeking license renewal must convince the NRC that their RPVs are not susceptible to brittle fracture even though the vessel walls have reduced ductility and fracture toughness after exposure to neutron radiation at operating temperatures of about 288 o C (550 o F) for 40 years. One concern is that if the RPV walls are overcooled by a sudden drop in water coolant temperature during a pressurized thermal shock (PTS) event, small flaws in the wall and/or its welds could grow into cracks that propagate through the wall, introducing the possibility of RPV failure.
A re-evaluation of the NRC's stringent PTS rule could result in earlier nuclear plant license renewals. An updated ORNL computer code is being used for this re-evaluation. Called the FAVOR (Fracture Analysis of Vessels: Oak Ridge) code, it was developed by Terry L. Dickson of the Computational Sciences and Engineering Division (CSED) for ORNL's Heavy-Section Steel Technology (HSST) Program. NRC engineers applied FAVOR to three pressurized-water reactors for which license renewals have been sought.
"Our preliminary results show that conservatisms in the current PTS rule can be reduced while continuing to provide reasonable assurance that public health and safety are protected," Dickson says.
Hole in a Head
HSST researchers have been involved in analyzing a recent nuclear power plant incident. On February 16, 2002, during an outage for refueling, workers making an NRC-required inspection at the Davis-Besse Nuclear Power Station in Ohio found long cracks in three control rod drive mechanism (CRDM) nozzles penetrating the RPV head, the dome-shaped upper portion of the ferritic steel vessel housing the nuclear core of the pressurized water reactor (PWR). Upon further inspection of the outside of the RPV head, the workers observed a buildup of white powder, identified as boric acid deposits. They also found that the cracks in the CRDM nozzles allowed the boric-acid-containing cooling water to leak outside the vessel.
After the powder was removed on March 7, workers identified a cavity 5-in. wide and 6-in. deep between two nozzles (through which control rods are withdrawn to start up the reactor and returned to shut it down). What caused this irregularly shaped cavity to develop?
Investigation found that, because boric acid is corrosive to the ferritic steel forming the RPV wall, the leaked water ate its way through the 6-in.-thick wall until it reached the backside of the austenitic-steel layer that protects the inside of the RPV wall from corrosion by the borated cooling water. This stainless-steel cladding (with a nominal thickness of 0.24 in.) was pushed outward into the cavity by the high pressure--2165 pounds per square inch (psi)--of the 600 o F water inside the vessel. The thin cladding deformed but did not break, holding the cooling water within the vessel. It was remarkable that the cladding held up so well because it was later discovered that the thin stainless-steel layer was flawed-- it contained defects with an estimated average depth of 0.05 in.
Today the Davis-Besse reactor remains shut down. NRC is investigating the structural integrity of reactor vessel heads at 69 PWRs.
After these discoveries, NRC sought analytical and experimental assessments from Paul Williams and B. Richard Bass, both of CSED, and Wallace J. McAfee of the M&C Division. "NRC wanted to know how close the reactor vessel head was to failure," Bass said. "What were the probabilities of failure based on what can be ascertained from the conditions?"
The HSST Program is supporting an accident sequence precursor assessment to identify the factors contributing to the event. HSST researchers are also trying to determine whether the flawed cladding would have failed by ductile tearing or plastic collapse and what would have been the size of the rupture.
Williams is developing analytical computational codes to calculate probabilities concerning how and when the Davis-Besse RPV would have failed with continued operation. In addition, McAfee is conducting a series of 14 high-pressure clad-burst experiments on flawed clad disks that will be completed in 2004. Bass says that these ORNL large-scale tests provide essential data against which model predictions can be compared, thereby improving analytical tools for predicting the likelihood of failure.
ORNL has a growing nuclear safety research program, and it's safe to say that the Laboratory is providing useful answers to NRC's challenging questions.