In March 1998, the Ministry of Agriculture, Fisheries and Food cautioned residents of the English village of Seascale not to eat pigeons. It seems that a couple of dead pigeons removed from the garden of two sisters were found to have high levels of radioactivity. The pigeons had roosted on buildings at the nearby Sellafield nuclear plant, where they had become contaminated. British Nuclear Fuels, Limited, which runs the plant, checked for contamination around the sisters' home. A portion of their garden had to be cleared and disposed of as hazardous waste because of the radioactive pigeon droppings. [5]

Near Miss at Oyster Creek

A technician testing switches on May 2, 1979, at the Oyster Creek plant in New Jersey caused a false signal of high pressure inside the reactor vessel. This false signal automatically tripped the reactor, which caused the turbine to be shut down seconds later. The turbine trip caused the plant’s internal power supplies to transfer from the auxiliary transformers to the startup transformers.

However, one of the two startup transformers was unavailable due to maintenance. Two of the three feedwater pumps were powered from the unavailable startup transformer and stopped running when the auxiliary transformers de-energized. The third feedwater pump automatically tripped during the feedwater system transient. Operators couldn’t restart this feedwater pump because its auxiliary oil pump was broken.

With the loss of all three feedwater pumps, the only water supply to the reactor vessel was provided by the two control rod drive pumps, which didn’t provide enough water to the reactor vessel to make up for the all the steam being pumped out. Consequently, the water level dropped until it was a mere twelve inches above the top of the irradiated fuel assemblies in the reactor core. The normal water level is more than ten feet above the core.

About 36 minutes after the scram (i.e., the rapid shutdown of a reactor by insertion of its control rods), the operators restarted one of the feedwater pumps, and reactor vessel water level was quickly restored to the normal operating band. A few minutes’ delay would’ve probably caused the reactor core to be uncovered. As at Three Mile Island Unit 2 just 35 days earlier, uncovering the reactor core could’ve triggered its meltdown. [6]

Please Don't Flush the Toilet While the Reactor is Running

A sign on the bathroom door at the University of Florida’s research reactor in the 1980s warned:

"Please Don't Flush the Toilet While the Reactor is Running."

The cooling water system for the reactor was connected to the city water main. This connection also supplied water to the toilet. The reactor automatically shut down at least five times because a flushing toilet had affected its cooling water flow.

The good news, on a relative basis, was that this research reactor was equipped with a direct connection to a well-supplied cooling water system for "risky" experiments. [7]

How could a conscientious person in the bathroom really tell whether the reactor was running? To flush or not to flush, that's the question.

Fatal Accidents

On December 9, 1986, the reactor at the Surry Unit 2 facility in Virginia scrammed from full power. A pipe going to one of the main feedwater pumps ruptured, releasing about 30,000 gallons of hot water into the turbine building. Some of the water flashed to steam. Eight workers were scalded by the water and steam. Four workers later died from their injuries.

The pipe's metal walls were originally about half an inch thick, but they had been eroded over the years to less than 10 percent of their original thickness by the water flowing through the pipe. This thinning weakened the piping until it broke.

The high temperature in the turbine building following the pipe rupture caused 62 fire sprinklers to discharge. After the sprinklers activated, water from the pipe rupture and sprinklers flowed into a control cabinet for the Halon fire-suppression systems in the emergency switchgear rooms. As a result, both Halon systems discharged. Water from the sprinklers also entered the control cabinet for the fire-suppression systems in the cable tray rooms. Both systems discharged carbon dioxide into the cable tray rooms located directly above the control room.

Water also got into a security card reader. The water caused this reader to send a continuous signal to the security computer, which overloaded the system and locked all areas controlled by card readers. This lockout enabled personnel to leave these areas, but prevented anyone from entering them.

Halon and carbon dioxide leaked into the control room. Carbon dioxide, which is heavier than air, flowed down the turbine building hallway. The control room door to the turbine building had been blocked open to allow personnel to enter the room. Due to the security computer lockout, the card reader at this door was not working properly. Because the pressure in the emergency switchgear room was higher than that in the control room, Halon leaked through floor penetrations into the control room. Operators in the control room reported shortness of breath, dizziness, and nausea from inhaling the Halon and carbon dioxide. [8]

This was not the first time that steam mortally injured nuclear plant workers. In fact, it wasn’t even the first time at the Surry plant. On July 27, 1972, three workers manually opened valves to allow steam to bypass the turbine and go directly to the condenser. After adjusting several valves, steam shot into the area through a small gap in a vent line. Two workers were badly scalded and rushed to the hospital. Both later died of their injuries. [9]

'Bama Boo-Boo

Workers at Farley Nuclear Plant Unit 2 in Alabama first achieved a nuclear chain reaction on May 8, 1981. They ran the plant through its initial operating cycle uneventfully. The plant was shut down on October 24, 1982, for its first refueling outage.

Four days later, personnel discovered that the containment spray header isolation valves were locked in the closed position. This system sprays borated water into the containment following an accident to reduce the pressure and temperature. The system also sprays sodium hydroxide into the containment following an accident to remove radioactive iodine from the air.

Or at least that's what the system could do if its isolation valves were open. A check of the records revealed that the valves had always been closed. Calculations determined that the radiation doses to the thyroids of plant workers and members of the public would have exceeded federal limits had an accident occurred with containment spray disabled.

The reason for the valves being in the locked closed position instead of locked open position was attributed to a modification during the plant's construction. The modification lengthened the stem of the valves, which made the valves appear open when they were actually closed. That excuse lost favor when further research discovered that the same modification had been made to the valves on Unit 1. The Unit 1 valves were found to be in the locked open (i.e., proper) position. [10]