The shortcomings were apparently overlooked by the Mine Safety and Health Administration, which approved the Crandall Canyon operators' plans to retreat mine - a method in which coal pillars supporting the mine are cut away, causing the roof to fall in.

Six miners were entombed in the Crandall Canyon mine when the enormous pressures of the mountaintop above caused a massive "bump," where coal explodes from the walls and pillars supporting the roof.

The study by the National Institute for Occupational Safety and Health said computer modeling "indicates that an elevated risk of bumps was present" in the area of Crandall Canyon where six miners were trapped by a massive coal outburst on August 6.

Sen. Edward Kennedy, D-Mass., chairman of the Senate committee that oversees MSHA, said the NIOSH report "raises very serious questions about whether the MSHA review process is strong enough and independent enough."

"Such questions about the review process are not just about Crandall Canyon - they have nationwide implications," he said.

MSHA did its own review of the models done by Agapito Associates Inc., the consulting company hired by the Crandall Canyon operators, before it approved the proposed mining. Kevin Stricklin, MSHA's administrator for coal-mine safety, said the agency's probe into the Crandall Canyon disaster would look at why those issues weren't caught.

Phone calls and e-mails to Agapito officials were not returned Tuesday.

The NIOSH report questions several of the assumptions Agapito made in running its computer models, designed to measure the stability of a mine structure contemplated in mining plans.

Even with those assumptions, the mining at Crandall Canyon appears to have been done in the margins of safety, with the consultant acknowledging that, based on the computer modeling, historical precedent indicated there was just better than a 50-50 chance the coal could be mined successfully.

In an Aug. 9, 2006, e-mail, sent days after Robert Murray's UtahAmerican Energy Inc. took over the mine, Agapito acknowledges that the "stability factor" in Crandall Canyon was well below what is recommended by NIOSH.

In deep mines, like Crandall, with comparable stability factors, "slightly more than half were successful, while the remainder encountered ground control problems," Agapito wrote.

Those "ground control problems" range from damaging bumps and pillar failure to catastrophic collapse trapping miners, according to the NIOSH database.

That information was not included in the consultant's formal memo to the company. Instead, the consultant relied on another, more sophisticated model - known as the LaModel - that Agapito said indicated the mining could be done safely.

Keith Heasley, who designed the LaModel for NIOSH and is now an engineering professor at West Virginia University, told The Tribune recently the validity of the model hinges on the accuracy of the data put in the program.

And the NIOSH report questions several of the figures and assumptions Agapito used:

* Agapito assumed the coal in Crandall Canyon was 80 percent stronger than typical coal;

* Pillars that had been partially extracted or collapsed were assumed to be helping to bear enormous pressures;

* The consultants predicted massive pressures could be managed by barrier walls half the width that NIOSH calculated would actually be needed.

When Murray's company bought the mine, there were two huge coal walls, some 450 feet wide, running on either side of the main tunnels.

On the other side were vast areas, completely mined, leaving behind rubble and leaving the coal barriers to bear tremendous pressure.

A standard formula used to calculate barrier widths shows the walls should have been 400 feet. The NIOSH computer models said the barriers should be at least 250 feet wide.

But leaving such thick coal walls would have made it impossible to dig the tunnels necessary to mine the valuable coal left in the barriers.

Agapito estimated that a 100-foot thick barrier wall would be adequate to handle the pressures and leave just enough room to dig the tunnels and recover the coal. Miners were also "slabbing" the barriers, or cutting 40 feet of coal out of the wall, weakening them further, records show.

The smaller barriers let pressures shift to the pillars, which were likely unable to bear the stresses, the NIOSH report said.

After a major bump in March damaged pillars and tunnels in the north barrier, operators moved to the south barrier, with Agapito arguing longer pillars could stabilize the roof and prevent future bumps.

Longer pillars would have little benefit without increasing the width as well, NIOSH reported. Stricklin said the March bump was not officially reported to MSHA. If it had been, additional scrutiny would have been given to the proposal to mine in the south barrier.