WASHINGTON -- After the Department of Energy spent almost $13 million to replace a body of Yucca Mountain research that had been tainted in an email scandal three years ago, a panel of scientists on Wednesday raised questions about the new work.

The Nuclear Waste Technical Review Board, an independent science body that reports to Congress, said it could not endorse a rebuilt computer model that would serve as a building block for DOE's bid to license a nuclear waste repository.

The water infiltration model estimates the rates by which there could be seepage into the Nevada mountain and down 1,000 feet where highly radioactive spent nuclear fuel would be stored and would need to remain safe and dry for thousands of years.

The review board concluded in a 30-page report the reworked model assembled by DOE and Sandia National Laboratories did not consider all available data, was not calibrated with other site information and did not consider likely significant evaporation.

As a result, revised estimates of water flow rates were about three times higher and less consistent with other lines of evidence compiled at the site, the board said.

Further, the Energy Department used a statistical adjustment to make infiltration estimates compatible with other models in its licensing computer program, the reviewers said.

"As used by DOE, the statistical modification of the infiltration estimates does not have a strong technical basis," the board said.

The board's opinion could put a new burden on the Yucca Mountain Project, where engineers and scientists face high bars to demonstrate and back up every detail of their work. The proposed repository is being designed to hold 70,000 metric tons of the most radioactive forms of nuclear waste. The Energy Department had no immediate comment. Yucca Mountain managers defended the new model at a review board meeting in March.

The Nuclear Regulatory Commission, which has asked for more information about the matter, could be the ultimate arbiter of whether the reworked infiltration estimates would be acceptable for licensing.

Nevada officials seized on the report as further ammunition in their fight against Yucca Mountain. They said it raises new doubts about the project.

"This shows the rework by Sandia was not legitimate," said Rep. Jon Porter, R-Nev., who held congressional hearings and sponsored audits on that segment of the project.

Porter said "it is obvious" DOE tried to "short cut" research to stay close to its schedules.

In the early 1980s scientists from the U.S. Geological Survey performed field investigations on water infiltration at Yucca Mountain and wrote a computer model that was folded into DOE's licensing package.

But much of that work was thrown out after the 2004 blockbuster release of emails written between 1998 and 2004 in which several USGS hydrologists suggested they were falsifying meticulous quality assurance records of their work.

The Energy Department launched several investigations. Project managers concluded the USGS science was defensible but its quality documentation was weak. They announced they would replace the USGS model as a way to rebuild confidence in the project.

Despite USGS problems with quality assurance, the Nuclear Waste Technical Review Board said it had a high opinion of the agency's science research. It recommended DOE go back and "requalify" the USGS work and put it to use.

The Government Accountability Office estimated DOE and other federal agencies spent $25.6 million to dissect USGS quality documents, retrace the infiltration model, review other computer models and software, and retrain workers.

The GAO estimated replacing the water infiltration model alone cost about $12.9 million.