ATLANTA, GA — Carbon monoxide (CO), a colorless, odorless, tasteless, nonirritating gas, causes thousands of deaths per year. A seminar at ASHRAE’s Winter Meeting here took a look at consumer apathy, CO monitor performance, and contractor negligence in the 100-plus deaths attributable to fossil fuel-burning heating systems every year.

According to Thomas E. Greiner, Ph.D., Iowa State University, “This room could be full of carbon monoxide, and we wouldn’t know it.”

Once the effects start, people in a room with CO may notice them, but they may not be able to make clear decisions. Effects can include “CO intoxication,” which impairs judgment to the point that people affected by it don’t get themselves out of danger in time.



Neurological Damage

Although there are more known emergency room cases of CO poisoning than CO-related deaths per year, poisoning can also be more difficult to diagnose.

One case that was fairly easily detected involved wedding guests who experienced CO intoxication at a church, said Greiner. After about 30 minutes, the bride’s mother and the pastor were lying on pews, complaining that they felt hot. They didn’t know what was going on.

They were saved when more guests came for the reception, wondering why these people were behaving strangely; one guest realized something was wrong and called the local fire department. CO levels were measured at 600 ppm.

The gas entered the building because the church’s absorption air conditioning unit was rusted out at the air handler. A local newspaper made light of the event, but one year later some guests were still feeling adverse effects, such as memory loss, personality change, and more severe neurological effects.

The brain and heart are typically damaged first, Greiner said. Early symptoms are headache and an achy, flu-like feeling. Alz-heimer-like symptoms, Parkin-son’s Disease, muscle twitching, and immune-related responses can result from prolonged exposure to higher levels, he said, and can persist well beyond the event.

Symptoms such as apathy, disorientation, amnesia, and urinary and fecal incontinence can also relapse, making them still more difficult to be properly diagnosed. “There is a gross underestimation [of CO poisoning] by doctors because it mimics so many other things,” he said. CO levels can be detected in the bloodstream.



Dangerous Practices

Consumers may be putting themselves at risk by not paying attention to their CO monitor warnings and neglecting heating system maintenance. They also can put themselves at risk by hiring unqualified heating contractors.

“People will think that they have the repair completed, they’ve called the heating contractor, but the alarm goes off again and they are later told that the repair was inadequate,” Greiner said. He noted that consumers also have an alarming tendency to shut off CO monitors when they continue to sound. “People think it’s too sensitive.”

Greiner cited a case involving renters who had lived in a house five years. When their furnace’s pilot wouldn’t stay lit, they called out a contractor who “fixed” it.

After the repair, the home’s smoke detectors went off. The landlord replaced the smoke detector battery, but there was still a problem.

The home had a patched-on garage that was not heated. Continuing investigation found that the chimney was plugged with dead birds, and it was cleaned out. Burned furnace wiring was also replaced.

The family now had a CO monitor and it went off. They called the university.

When Greiner came out, he found that the furnace’s air intake shutter was completely closed off by the first contractor in order to keep the pilot lit. “Some of our heating contractors are extremely dangerous,” he commented. The wiring had burned because of flame rollout.

Nonetheless, the pilot was burning with a blue flame. Greiner later pointed to the need for contractors to perform a full combustion analysis on an appliance whose operation is in doubt. “Under many conditions, the flame will continue to burn blue. In typical scenarios of multiple CO alarms there is over-firing, especially in conversions from lp to natural gas and vice versa.”

While Greiner was investigating this home, he said he saw something that he wished he’d been able to catch on video. He watched a bird perch on the edge of the chimney. Soon it wobbled a bit, and then it disappeared — down the chimney.



Consumer Apathy

Paul K. Clifford, Ph.D., said that his company, Mosaic Industries, studied field experiences to determine what causes CO alarm activation and how homeowners respond.

In 1994, a Chicago, IL, ordinance was enacted requiring the installation of residential CO detectors. From Oct. 1 through Dec. 20 of that year there were 8,500 emergency responses in the city. (Recall the notorious Thanksgiving Day that year, when the city’s fire department was kept hopping due to alarms that sounded because of a combination of oven use and the right atmospheric conditions.)

The increased incidence of false positives was one of the things that prompted UL 2034 regulations to improve the performance of CO monitors. “Are these really false alarms?” Clifford asked.

Mosaic’s study found a considerable degree of consumer skepticism:

  • There is a 15% rate of installation in cities with no CO alarm ordinance.

  • There is only a 20% to 25% rate of installation in cities with an ordinance.

    • Perhaps more disturbing were the consumers’ responses to the alarms. On activation:

  • 62% called no one.

  • 23% assumed it was a false alarm.

  • 30% knew a probable source but did nothing.

  • 58% performed no investigation.

  • 10% called the utility.

    • Gas Technology Institute (GTI) chamber tests from 1998 showed many late alarms. This “flies in the face of earlier experience” and runs counter to the consumer perception that the units are too sensitive.



    Performance Questions

    According to testing, the lower the relative humidity, the worse the units’ performance, Clifford said. Other results included:

  • Out-of-box failures, with some variation of failure rates between brands;
  • Poor control of sensitivity, which did not meet UL compliance;
  • False alarms;
  • Many performance failures at toxic levels;
  • Failure to alarm at low relative humidity;
  • Increasing failure to alarm/ late alarms with unit aging; and
  • Inaccurate digital displays. The worst was accurate within plus-minus 200 ppm.
    • “Many detectors do not respond to concentrations lower than 100 ppm,” he said. Many models also alarm unpredictably over the course of CO exposure, he said.

    Alarm performance is unrelated to UL rating, he continued. The best-performing brand on the market is not UL rated. However, there were many unsupervised failures across the board. And at 5% rh, all brands were unresponsive to CO, he said.

    Mark Goldstein, president of the Quantum Group in San Diego, CA, and also president of the manufacturer group Carbon Monoxide Safety and Health Association (COSHA), questioned these results in a phone interview. He pointed out that units have a 48-hr preconditioning requirement to stabilize alarm performance, and that changes from this procedure can skew results.

    He also pointed out that UL only tests units to 15% rh since in non-lab conditions, “When combustion is coming into the house, you’ve got condensation,” which generally results in higher relative humidity levels, Goldstein said. He said that units are typically calibrated to higher relative humidity levels because of this.

    Natural gas is comprised of chemical compound CH4, he explained. When burned, it releases one molecule of CO and four of H2O. “At 400 ppm, you’ve got 400 molecules of water for every molecule of carbon monoxide,” he pointed out. Testing to 5% rh was, in his opinion, “irrelevent” to possible field conditions.

    Certification and testing bodies such as UL, CSA, and CENELEC (EU) are all working to create standards that would hopefully improve the accuracy and predictability of the units, Clifford said. Innovations to the Canadian Standards Association’s 6.19-00 would entail:

    • Less than a 1% failure rate (lack of compliance to specification) at time of manufacture, and a 90% confidence level; and

    • Reliability of production and test methods. A maximum of 37.7% could fail after three years, with no more than 23.1% unsupervised failures.

    Clifford pointed out that “CO detectors are a small part of the solution.” People may install them in lieu of regular service by a qualified contractor, then may not respond when the alarm goes off.



    CPSC Endorsement

    According to Max H. Sherman, Ph.D., who is with Lawrence Berkeley National Laboratories and also is on the committee for ASHRAE Standard 62.2, “Ventila-tion for Acceptable Indoor Air Quality,” the standard will not mandate the use of CO monitors in residential applications.

    The standard will address source control, since “You can’t set ventilation rates without considering sources you want to control.” But it will not include unusual sources, like a collapsed chimney or extremely high outdoor levels.

    More common conditions that can lead to high CO include back drafting, slippage, and unvented combustion.

    However, the U.S. Consumer Product Safety Commission (CPSC) stands by its statement that “Every home should have a CO alarm,” said spokesman Ken Giles.

    “If ASHRAE has information on individual brands that are defective, we would like to have it. That’s the kind of empirical, scientifically valid research we could use,” he said. “We could negotiate a recall.”

    COSHA’s Goldstein said, “ASHRAE should do what they think is best,” although he questioned the validity of some of the research presented at the Winter Meeting.

    He added, “I think CO detectors are saving lives.”

    Publication date: 02/12/2001