The Department of Energy needs more information to make sure its leak-prone underground tanks will last as long as they are needed to store radioactive waste, according to a report to Congress by the Government Accountability Office.

"DOE and its contractors lack comprehensive information about the condition, contents and long-term viability of Hanford's aging underground waste tanks," the report said.

DOE is emptying the contents of 142 tanks built from the '40s and '50s into 28 newer and sturdier double-shell tanks to await treatment and then disposal. The double-shell tanks are generally structurally sound, the GAO agreed with DOE.

However, "the condition of the older, single-shell tanks -- nearly half of which are confirmed or presumed to have already leaked -- is much less certain," the report said.

Together the 170 tanks hold about 56 million gallons of waste, enough to cover a football field to the height of a 15-story building. They have leaked an estimated 1 million gallons of waste, although none is currently known to be leaking.

The GAO is concerned that by the time the last single-shell tank is emptied of waste it may be 91 years old, even though the single-shell tanks were built to hold waste for only a decade or two.

"The likelihood of a major failure of a tank increases with time," the report said.

It referred to a 2007 study by the Washington State Department of Ecology that said the probability of a single-shell tank leaking may double about every 10 years. The study estimated that about 41 of the single-shell tanks designated as sound could leak waste into the ground by the time they are emptied.

DOE has removed pumpable liquid from the double-shell tanks, but about 3 million gallons remain, some of it in small pores among the solid waste. In addition, work in tanks, including to remove solid waste, may require liquid to be added to tanks.

DOE is in the early stages of a study to determine the structural integrity of the single-shell tanks. It plans to spend about $800,000 in the current fiscal year to plan the study and $2.5 million next year to begin it.

Now DOE has limited means to assess the integrity of tanks and determine if leaks have occurred, according to the report. DOE monitors the level of waste in the tanks, lowers cameras into the enclosed tanks to take pictures and has built monitoring wells around the perimeter of the tanks. But it has not thoroughly examined any of the seven tanks that have been emptied so far, the report said.

Delays in building the $12.2 billion vitrification plant have added uncertainty to the tank waste program, the report said.

Although ways of treating some of the waste sooner are being considered, the main plant is not expected to begin operations to treat waste for disposal until 2019. That will free up space in double-shell tanks for more of the waste in single-shell tanks.

DOE had agreed under the legally binding Tri-Party Agreement to have the single-shell tanks emptied by 2018. Now it is proposing to regulators that the tanks be emptied by 2040, and a DOE analysis completed early this year looks at postponing emptying the single-shell tanks to 2047 given delays in finishing the vitrification plant.

However, DOE still may be overly optimistic about the time it needs to empty tanks, the report said.

It's currently emptying tanks at the rate of about one a year. That pace is determined in part by the Hanford budget, the need to test and develop new technologies to retrieve waste and limits on the remaining space in double-shell tanks until some treatment starts.

Hanford officials have indicated they can retrieve 1.7 tanks per year until 2019, the report said.

To determine how long the tanks may last, DOE needs to know more about what's in the tanks, the report said. The tank waste includes 46 different radioactive elements, the remains of 240,000 tons of chemicals added to the tanks during Hanford's plutonium-production years and miscellaneous material, such as discarded equipment, concrete, casks of experimental fuel elements and plastic bottles of plutonium and uranium.

"Beyond the general characterization of tank wastes, DOE lacks knowledge of the specific proportions of constituents in each tank," the report said. Understanding the types and quantities of waste may be critical to determining how long the tanks may be safely used, it said.

More knowledge also could help DOE make the best decisions weighing costs versus risk, the report said. Retrieving the last 15 percent of the waste from a tank can cost as much or more than removing the first 85 percent of the waste, the report said.

"In the absence of specific risk analyses, the accompanying reduction in risk, if any, is unclear," the report said.

DOE disagreed that it lacked the information to make sound decisions on managing the tank farms. As work progresses it will be performing additional monitoring and characterization of waste, it responded.

"This is consistent with sound project management principles, where longer term conceptual planning gives may to cost-effective detailed work planning for nearer term activities," wrote Ines Triay, acting DOE assistant secretary for environmental management, in a letter to the GAO.

The GAO is recommending that DOE give priority to assessing the long-term viability of single-shell tanks; report risks to the public, workers and environment on a regular schedule every three to five years; and work with its regulators to develop a realistic schedule for emptying tanks and to include risk information in its management strategy.