CHARLESTON, W.Va. -- Federal coal mine regulators have concluded that a widely used emergency breathing device doesn't meet safety standards but have not yet taken any action to get the equipment out of the nation's mines.

This week, the National Institute for Occupational Safety and Health issued a long-awaited report on its investigation of problems with CSE Corp.'s SR-100 model self-contained self-rescuer, or SCSR.

NIOSH said an investigation of longer than two years has found a "critical" defect, in that too many of a random sample of SR-100 units had faulty oxygen starter devices.

The report said that five units out of 500 tested by NIOSH failed a test of an oxygen cylinder that kick-starts the devices. Under federal rules, no more than three of the 500 units could fail the test for NIOSH to be confident the SR-100 had an acceptable failure rate.

"Based on the findings of the investigation, the field-deployed CSE SR-100 units no longer conform to the minimum requirements for the certification" under federal rules, the NIOSH report concluded.

Though Pittsburgh-based CSE has come out with a new model with a different starter system, more than 70,000 SR-100 units are still in use in coal mines around the country.

NIOSH officials said Tuesday that they are still talking with the federal Mine Safety and Health Administration, trying to develop a joint plan to deal with the matter.

"We've had several meetings," said Maryann D'Alessandro, director of NIOSH's personal protective technology laboratory. "We'll continue to have meetings."

Through a spokeswoman, MSHA chief Joe Main on Tuesday declined an interview request, saying he wanted to review the new report and discuss it with NIOSH before commenting. MSHA was given the 24-page report on Friday, and NIOSH briefed top MSHA officials on their findings in early January.

The CSE emergency breathing devices have been under scrutiny for at least six years, when Sago Mine Disaster survivor Randal McCloy testified that the SR-100s of four of the 12 miners trapped by the Jan. 2, 2006, explosion wouldn't start.

The SR-100 model uses a chemical process to generate oxygen, based in part on a reaction with carbon dioxide being exhaled by its user. Over the years, coal miners had expressed similar complaints about SCSRs not starting or appearing to start slowly. Government and industry officials generally dismissed those complaints, saying the real problem was that miners didn't understand how to use the units properly.

Under federal law, all coal miners must be provided with a self-rescue device that will provide them with at least an hour's worth of breathable air to escape in the event of an underground fire or explosion.

MSHA and NIOSH jointly certify the devices as complying with federal regulations, and the agencies have an agreement through which they are jointly charged with dealing with any problems that come up later when devices are in use.

In 2010, MSHA and NIOSH launched a joint investigation of problems that were eventually traced to the oxygen cylinders used in the initial start-up of the SR-100 devices. Initially, CSE said it had instituted a "recall" of the troubled units, but later conceded it had not actually ordered coal companies to stop using the devices.

Scott Shearer, president of CSE, did not return a phone call Tuesday, but said in a prepared statement that his company has "received no final directive from NIOSH or MSHA regarding the SR-100."

"CSE voluntarily ceased production of the SR-100 when the company's internal quality control process identified a possible issue with the oxygen cylinder," Shearer said in the statement.

"We have remained committed to working with the agencies for some time to resolve any and all concerns," he said. "Once we receive final notification from the regulatory agencies, we will continue to develop appropriate next steps."

In lawsuits after the Sago disaster, families of the miners were investigating concerns that the SR-100 cylinders, made by a vendor for CSE, somehow leaked, leaving the units without enough oxygen to start properly. Those suits were settled, and the terms were kept confidential.

NIOSH said in its new report that an engineering firm hired by CSE had concluded that a loss of oxygen in the starter cylinders occurred at a threaded connection between the cylinder body and oxygen outlet assembly.

"When asked to determine the percentage of affected, assembled, SR-100s, CSE responded that it was less than 1 percent overall," the NIOSH report said.

"CSE could not identify a systemic cause or otherwise confine the failure to within certain lots," the agency report said. "Therefore, the failure could exist among all field-deployed units."

Reach Ken Ward Jr. at kw...@wvgazette.com or 304-348-1702.