Thousands of gallons of radioactive waste leaked from the inner shell of double-shell storage Tank AY-102 as Hanford workers pumped sludge from the tank during the weekend.
Workers were emptying the tank, Hanford’s oldest double-shell tank, because it had waste leaking into the 2-foot-wide space between its shells — called the annulus — in three places.
About 70 gallons of waste had been estimated in recent years to have leaked, drying in three separate patches. The leak detected this weekend added about 3,000 to 3,500 gallons.
Checks on Sunday and again Monday revealed that no waste breached the outer shell of the underground tank and none reached the surrounding environment, according to Department of Energy contractor Washington River Protection Solutions.
Work to retrieve sludge from the tank and move it to a sturdier double-shell tank stopped Sunday, and preparations were under way Monday to pump the leaked waste in the annulus back into the inner shell of the tank.
“Operations took a challenging, but not unanticipated turn,” said Rob Gregory, chief operating officer for the tank farm contractor, in a message to employees Monday.
An alarm for a level detector in the annulus sounded about 3:30 a.m. Sunday as sludge was pumped from the 1-million-gallon capacity tank.
Work to retrieve waste from the inner shell stopped when the detector showed 2 inches of waste had accumulated in the annulus.
The waste continued to seep into the annulus throughout the day, stabilizing at about 8 inches deep.
Early on Monday, the level of the waste in the annulus changed again. It dropped by about three-quarters of an inch, which would be about 180 gallons of waste, raising the possibility that it might be leaking from the outer shell.
Hanford workers checked a leak detection pit that sits like a saucer beneath the underground tank both Sunday and Monday, with no indication that waste had breached the outer shell of the tank. The level of the contents of the pit, which can collect precipitation, had not increased.
The pH level of the contents of the pit also were checked, with results early Monday evening showing they did not match the high pH waste in the tank.
The pit check does not rule out that waste could have leaked somewhere else from the outer shell but did not reach the pit.
However, another possibility under investigation is that some waste has flowed into the ventilation channels in the platform the inner shell sits on above the outer shell, reducing the level of waste measured in the space between the tanks.
The Washington Department of Ecology, the regulator on the project, has been working with Hanford officials since Sunday.
“There is no indication of waste leaking into the environment or risk to the public at this time,” the state confirmed in a statement Monday.
The increased rate of leakage likely occurred because waste retrieval work inside the tank dislodged material that was partially blocking leak sites, according to the Department of Ecology.
Increased leakage was an anticipated possibility as the state ordered DOE to empty the tank under a settlement agreement, and a contingency plan was developed to address it.
Before Washington River Protection Solutions started to remove waste from the tank in March, it placed a pump in the annulus.
The pump could not be hooked up while sludge was being pumped from the tank, but work was completed Monday to hook up a transfer line from the annulus to return waste to the main tank. Radioactive shielding was expected to be added Monday evening to the line to protect workers when waste is removed from the annulus.
Pumping from the annulus could begin within two days, DOE said Monday.
“We expect to resume sludge retrieval simultaneously with annulus pumping,” Gregory said in the employee message.
Waste was emptied quickly from the tank until Sunday.
DOE met a deadline with a day to spare in early March to start emptying an estimated 650,000 gallons of liquid waste that sat on top of about 150,000 gallons of sludge in the tank.
It started the more difficult work of emptying the sludge at the end of March, working mostly weekends. By Sunday, just 14 inches of sludge remained on the bottom of the tank, or about 46,000 gallons by Washington River Protection Solutions estimates.
Ninety-five percent of the liquid and sludge waste has been removed, DOE said.
DOE and Washington River Protection Solutions have a deadline under the settlement agreement to have the tank emptied enough that the cause of the leak may be investigated by March 4, 2017.