A failed pump has once again stopped work with Hanford's new Mobile Arm Retrieval System to retrieve waste from an underground tank of radioactive waste, ending hopes that it would be emptied this year.
The Department of Energy is required to empty that tank and six others before October 2014 to meet a court-enforced consent decree. DOE already notified the state of Washington that it may not be able to empty two of the seven tanks on time.
However, the latest setback on emptying Tank C-107 with the Mobile Arm Retrieval System, or MARS, may not mean that now three of the seven tanks are at risk of not being emptied by the consent decree deadline.
DOE and Washington River Protection Solutions have developed a plan to reschedule work to get as many of the tanks emptied as they can before October 2014.
Two of the seven tanks could be emptied to regulatory standards before October 2013, according to information provided by DOE to the Washington State Department of Ecology, the regulator on the project.
MARS began pumping waste from Tank C-107 in fall 2011, raising hopes that the slow and difficult work to empty radioactive waste from enclosed, underground tanks would become more efficient.
It's the largest and most robust system tried yet to empty tanks and required the unprecedented step of cutting a 55-inch diameter hole in the top of the tank and installing a new riser large enough to lower the MARS system into the tank.
But in about 22 months, it has operated for just 450 hours for reasons that have had little to do with MARS.
"MARS itself is doing a very good job to break up the waste," said Nancy Uziemblo, a scientist at the state Department of Ecology.
Typically, Hanford workers have to use at least two, and sometimes more, technologies to empty each tank. But MARS was designed to allow one system to attack different types of waste within a tank.
In the hours it has operated it has finished removing sludge with the consistency of peanut butter from the tank and also gotten out some of the hard layer of waste that was beneath the sludge.
The system is designed to allow equipment to get very close to the waste and spray liquid at high pressure to erode away the harder layer, Uziemblo said.
MARS has removed 230,000 gallons of waste from Tank C-107 out of an estimated 253,000 gallons of waste in the tank when it started work. Before the pump inside the tank failed, Hanford officials believed the remaining waste in the tank might be retrieved in a matter of weeks.
Instead, work to replace the pump may stretch until the end of December.
The waste from Hanford's 149 single-shell tanks -- including the 70-year-old Tank C-107 -- is being emptied into newer double-shell tanks.
Earlier delays in getting Tank C-107 emptied were mostly caused by difficulties in double-shell tanks, including a broken pump in the tank that was supplying liquid waste to spray inside the single-shell tank. Liquid waste is used rather than water to prevent the creation of more contaminated waste.
The pumps in the single-shell tanks also fail periodically, although usually last longer than the 450 hours that the Tank C-107 pump was used, said Jeff Lyon, the Department of Ecology manager for tank waste storage. However, it's a "very hostile environment," he said, and the pump has been in the tank for about 22 months.
DOE is shifting work from Tank C-107 to Tank C-105, which is one of the two tanks that DOE had notified the state it might not have emptied by the consent decree deadline. The two tanks share an electrical system, so work may only be done on one at a time. That system may be modified to allow work to more easily switch between the two.
Tank C-105 will be emptied using another MARS, this one equipped with vacuum technology to introduce as little liquid into the tank as possible. Tank C-105 is not on the list of about 68 tanks suspected to have leaked. However, because of contaminated soil around the tank, Washington River Protection Solutions is taking a conservative approach and limiting the liquid used in waste retrieval.
Work to pump the tank could begin this winter.
Work already is under way to empty Tank C-101 and could be completed this month, Uziemblo said. Work to remove its 88,000 gallons of waste started in mid December.
The other tank that could be emptied by October is Tank C-110, where work to empty the remainder of the tank waste will be done sooner than had been planned.
Hanford workers will empty the remaining 17,200 gallons at the bottom of the tank with the Foldtrack, an 800-pound machine that extends to 12 feet long to fit through the narrow riser that provides access into the tank. Before it's lowered completely onto the floor of the tank, it folds in half, placing its two crawling tracks in parallel with a plow blade for pushing the waste in front.
The tank has waste with the consistency of sand, that sits in drifts that may hide larger chunks of waste underneath. The Foldtrack is planned to push the waste toward a pump in the center of the tank to allow it to be removed.
"I'm pleased with what I see," Uziemblo said.
To get tanks emptied Washington River Protection Solutions has brought in additional crews and work is being done in two shifts, she said.
Because of the hotter than normal weather this summer, workers have been coming in early in the morning to allow more progress to be made.