Up to 520 gallons of radioactive waste might have leaked from the inner shell of Hanford's Tank AY-102 sometime after January 2007, according to the first comprehensive report on the newly discovered leak.
The 444-page report released Tuesday by the Department of Energy details the myriad construction problems with the 42-year-old tank, which the report concluded likely led to corrosion of the tank at high temperatures.
"Tank AY-102 construction records detail a tank plagued by first-of-a-kind construction difficulties and trial-and-error repairs," the report said. The result was a tank that was not as robust as designed, it said.
However, because of construction and design improvements in the double-shell tanks built after Tank AY-102, other double-shell tanks likely don't have the same risk for leaks, the report concluded.
Tank AY-102 was the first of 28 double-shell tanks built at Hanford and the first discovered to have a leak from its inner shell. The waste is contained in the space between the inner and outer shells of the tank and has not escaped into the environment.
Hanford has 56 million gallons of radioactive waste left from past production of weapons plutonium and held in underground tanks. Waste is being pumped from 149 single-shell tanks, many of which have leaked waste into the soil in the past, into the newer double-shell tanks.
DOE's current tank farm contractor, Washington River Protection Solutions, began working on the new leak assessment report after suspicious material was found in August between the shells of Tank AY-102.
The report estimated that 190 to 520 gallons of waste leaked from the inner tank. However, a significant portion of the liquid evaporated, leaving about 20 to 50 gallons of drying waste, according to the report.
The tank has a capacity of more than 1 million gallons and holds about 850,000 gallons of waste, and the volume in the tank is monitored.
However, the tank is so large that 2,750 gallons would have to be lost for the level of the waste to drop 1 inch, said John Britton, Washington River Protection Solutions spokesman. In addition, atmospheric pressure changes the level of the waste, and about 75 gallons of water are added to the tank each day because the heat generated by the waste in the tank causes evaporation.
The leak is believed to have started sometime after January 2007, the report said. An inspection of the space between the shells in December 2006 and an ultrasonic inspection of the inner-shell walls in January 2007 found no evidence of leakage, the report said.
But in January 2007, liquid waste containing a significant concentration of potassium, which is found in only a few of Hanford's tanks, was pumped into the double-shell tank. One of the samples collected from leaked waste found on opposite sides of the tank had an unusually large concentration of potassium, the report said.
The construction problems started when the base of the outer shell was built with thin steel plates. Bulges were created when they were welded and when welds were redone, but the work was "eventually accepted so construction could proceed," the report said.
A layer of hard, insulating material was poured as a refractory over the steel bottom of the outer shell, but the 8-inch thick refractory cracked as the bulges moved, it said.
Welding the inner shell, which was on top of the refractory, also proved difficult, and 36 percent of welds were rejected upon inspection.
"Weld maps show welds being reworked as many as four times before passing radiography examination," the report said.
After welding was completed, the inner shell was heated up to relieve stresses and prevent welds from cracking. However, rainwater saturated the refractory and the tank bottom temperature could not be raised above 210 degrees Fahrenheit for two days while steam escaped from the water-soaked refractory, the report said.
After stress relief was completed, part of the refractory was found to be too damaged to be used. The outside 21 inches were excavated from beneath the inner shell and replaced with structural concrete. Pieces of Styrofoam were used to fill gaps in between the top of the refractory and the inner shell of the tank, the report said.
The tank began accepting waste in 1971 and likely began building up a thin layer of mildly corrosive sludge on its bottom when liquid waste was added between 1977 and 1984, the report said.
Then in 1998 and 1999, high-temperature sludge emptied from single-shell Tank C-106 was transferred into the double-shell tank.
"The sludge formed a blanket over the existing sludge and increased its temperature dramatically," the report said. "It is likely that the corrosion rate accelerated after the temperature increase."
DOE had planned to stage high-level radioactive waste at 15 of the double-shell tanks that would serve as feeder tanks to the vitrification plant under construction to treat the waste for disposal. Tank AY-102 had been planned to be the initial tank used to feed waste to the plant during commissioning, but that will not happen now.
Construction proceeded more smoothly on the second double-shell tank built, Tank AY-101, which had a 10 percent rejection rate of bottom welds, compared to the 34 percent of Tank AY-102. In addition, design changes were implemented for future generations of the tanks, with the newest one built in 1986.
Although the report speculated that Tank AY-102's problems may not be found in other tanks, DOE now is having Washington River Protection Solutions inspect six other of the earliest double-shell tanks to check for leaks.
The Hanford Advisory Board, concerned that double-shell tanks will need to hold waste for decades to come until all the tank waste can be treated, recommended earlier this month that DOE start work immediately to build new waste storage tanks.
DOE has said building one double-shell tank could cost about $100 million and building a group of six tanks could take five to seven years.
-- Annette Cary: 582-1533; firstname.lastname@example.org