During the public meeting, proponents of including the monitor requirement cited statistics and test results to contend that the devices are more reliable than ever before. The question of reliability was a key reason the monitor requirement was removed following the first public review of the draft standard.

Test Data

Mark Goldstein, president of the Quantum Group and president of the manufacturer group Carbon Monoxide Safety & Health Association, reported on reliability testing by Quantum. He said tests were conducted both to the Canadian Standards Association (CSA) 6.19-01 and to Quantum’s own standards, which Goldstein said were even a bit higher than the CSA. He said testing involved pulling units off the assembly line and subjecting them to a range of residential environments including humidity variations. He said as of January 2003 there were “no failures of any kind” of those units tested. “That’s where we are at the moment, ” he said.

Goldstein noted the CSA testing procedures maintain that even if there are no failures, “you can’t reduce the sample size.” If a failure turns up “you have to increase the sample size. Now in sample testing, you can’t say there are no failures, but you have a higher confidence level than if you’ve had some failures.” He went on to say, “If it alarms too early, you have a failure. If it alarms too late, you have a failure.”

With reference to the 62.2 standard, Goldstein said, “CO detectors are needed and should be required. I’ll ask ASHRAE how it should be done.”

Bobbie Gee, president of Senco Sensors, did not focus on statistics. “We’re in the life and death business and we get caught up in being 2 ppms off [as grounds to question reliability]. I’m not going to get upset that the alarm didn’t exactly go off when it was designed to. Doggone it, we should have these in the homes. We need to call the families of victims [who died due to carbon monoxide] to keep our heads straight.

“Our devices have been tested and tested and tested. We have a great sensor and we would like to get them in everybody’s home.”

Isaac Papier, general manager of Underwriters Laboratories, reported on his organization’s testing of a wide range of monitors purchased by UL staff from various commercial outlets. He said no monitor is going to be perfect. “Some alarmed early, some alarmed late, some failed,” he said, “but there was enough of a safety margin to get people out” of their homes if necessary.

“Yes, we should be able to make a perfect device,” he continued. “But can it be affordable? I do not see that happening. We aren’t in a perfect state yet.”

David Mills of UL reported that the organization has been accumulating monitor data for 3-1/2 years. He said testing was done to various levels of ppm contaminants. He said the higher the ppm, the quicker the alarm went off. “Even with most of the late units, the situation was not a matter of life or death.”

Paul Clifford of Mosaic Industries was less upbeat. He said his studies “showed overall poor reliability.” He said that his tests on alarms purchased at retail stores found that 4 percent of the alarms tested sounded false alarms in clean air in one day. He also noted that seven of 10 brands purchased at retail stores “did not alarm at hazardous CO levels.”

The public forum also found spontaneous involvement from the audience. When one presenter said, “I can’t image one person in this room that doesn’t have a carbon monoxide device,” a voice chimed in, “I don’t.”

When one speaker did not provide statistics, an audience member said, “We came to hear data. Do you want to avoid data?”

Even the timing of the data came in for some criticism. “What the hell took you people so long?” said one audience member. “Why have we waited until the last minute to get data? You are putting us at a disadvantage.”

Publication date: 02/10/2003