Each winter, news stories of carbon monoxide-related deaths remind the public that CO is the most common form of poisoning in the U.S.

Each winter, news stories of carbon monoxide-related deaths remind the public that CO is the most common form of poisoning in the United States. A fire marshal usually warns that carbon monoxide monitors are an absolute requirement in every home, as essential as fire and smoke detectors. Homeowners without monitors are considered as indifferent about their families' safety as those who keep loaded pistols in plain sight.

Since hvac contractors have a daily responsibility to eliminate one of the potential sources of CO, they are sometimes encouraged to take the next step and promote the use of monitors that can warn residents before it is too late.

It is even suggested that monitors can be used as a marketing device that will allow contractors easier access to customers and provide a means for acquiring regular service and maintenance work.

But are consumer-grade CO monitors, detectors, and alarms sufficiently reliable for contractors to give them their seal of approval by selling and installing them? Although the products definitely save lives, and homeowners are safer with some level of protection than none at all, critics complain that the record of performance of many units is unsatisfactory or cannot be verified.

(Editor's note: The carbon monoxide probes, analyzers and monitors pictured with this article are commercial-grade products designed and built for the hvac trade. Opinions about residential, consumer-grade monitors do not apply to these products.)

The silent killer

The dangers of carbon monoxide poisoning are widespread and well documented. Carbon monoxide is a colorless, odorless, poisonous gas produced by the incomplete burning of solid, liquid, and gaseous fuels.

According to the federal Consumer Products Safety Commission, more than 200 people in the United States die every year from CO produced by fuel-burning appliances, including furnaces, ranges, water heaters, room heaters. Still others die from CO produced while burning charcoal inside a home, garage, vehicle or tent, or by leaving cars running in attached garages. Beyond the deaths, several thousand people each year go to hospital emergency rooms for treatment for CO poisoning.

The initial symptoms of CO poisoning are similar to the flu, but without the fever. They include headache, fatigue, shortness of breath, nausea, and dizziness.

Fifty parts per million is the maximum concentration a healthy adult should sustain over an eight-hour period. A concentration of 400 parts per million can be life-threatening within three hours.

According to a study in the American Journal of Emergency Medicine, cities with ordinances that require detectors have lower rates of CO-related deaths. In Chicago, which requires detectors, 0.4% of those exposed to carbon monoxide died. In Los Angeles, the rate was 15%.

According to the CPSC, "CO detectors-alarms always have been and still are designed to alarm before potentially life-threatening levels of CO are reached. The (Underwriters Laboratory) standard 2034 (1998 revision) has stricter requirements that the detector-alarm must meet before it can sound. As a result, the possibility of nuisance alarms is decreased."

CPSC said it worked closely with Underwriters Laboratories to help develop the UL 2034 safety standard for CO detectors-alarms.

"CPSC recommends that every home have a CO detector-alarm that meets the requirements of the most recent UL standard 2034 or the IAS 6-96 standard in the hallway near every separate sleeping area," the agency said.

Testo carbon monoxide analyzer.

Legislation mandates monitors

Requirements for CO alarms have been established in several states and municipalities. Effective Oct. 1, 1994, Chicago became one of the first U.S. cities to adopt an ordinance requiring the installation of CO detectors-alarms in all new single-family homes, and in existing single-family residences with new oil or gas furnaces.

On Feb. 8, 1999, then-New Jersey Gov. Christine Todd Whitman signed into law a bill that requires the installation of carbon monoxide detectors in multiple dwellings, hotels and boarding homes. CO monitors remain optional in single-family homes.

In 2001, the Rhode Island General Assembly enacted a law which requires most new and existing single-family homes and other residential properties to be equipped with both a working smoke detector and carbon monoxide detector prior to transfer of the property or issuance of a certificate of occupancy as of Jan. 1, 2002.

Carbon monoxide detectors must emit an audible signal with a minimum rating of 85 decibels at a distance of 10 feet and have a "power on" indicator.

A recently passed New York law requires every one- and two-family home, condominium and co-operative unit sold after Dec. 1, 2002 to have an operating carbon monoxide detector.

"This bill means greater residential safety for all New Yorkers," said Douglas Troutman, government affairs manager for the National Electrical Manufacturers Association.

The NEMA government affairs department sent a letter to New York Gov. George Pataki on behalf of the association's carbon monoxide detector and sensor manufacturers urging him to sign the bill, calling it a "most important life safety issue."

The Bacharach Monoxor II carbon monoxide analyzer measures and displays CO levels from 0 to 2,000 ppm.

Contractor involvement

If CO monitors save lives, and are sometimes even mandated, should contractors have any special concerns about installing or recommending the purchase of specific units?

For their part, most manufacturers of residential monitors do not necessarily encourage the trade to get involved in their use. The units are readily available from multiple retail outlets and do not require special skills to install.

The manufacturers are also careful not to overstep the claims they make for the products. Warranties clearly state that the products meet the requirements of UL 2034 and are subject to stated levels of error.

But if a trusted local contractor were to insert himself or herself between the product and the homeowner, he or she may be implying a level of performance that the monitors cannot always deliver.

In June 2002, the Gas Research Institute issued its final report, "Evaluating the Performance of Residential CO Alarms," prepared by Paul K. Clifford of Mosaic Industries. The 172-page report covered alarms available during the period from February 1999 through December 2001.

The study was commissioned after the 1998 revisions to UL 2034 that were meant to correct many of the problems that had existed previously in consumer-grade CO monitors.

According to Clifford, when CO alarms were first widely installed in 1994, they suffered from a number of drawbacks including numerous false alarms. After two years of CPSC-sponsored workshops and hearings, in October 1996, CPSC staff made recommendations for extensive modifications to UL 2034.

By 1998, approximately 20 million units had been shipped and an estimated 8% to 15% of homes had at least one alarm installed. According to Clifford, their track record was unimpressive.

A multiyear field survey found that more than 70% of CO alarm activations occurred at harmless CO levels. The alarms depended on conditions of ambient temperature, humidity, and outdoor CO levels.

"False alarms diminish consumers' confidence in CO alarms," the report stated. "After a false alarm, some consumers may intentionally disable their alarms or may ignore future alarms. Sixty-two percent of consumers in a study responded to an alarm activation by ignoring it.

"Several cities (Albany, Chicago, Toronto) passed closely watched ordinances requiring CO alarms in residential housing. Their poor experience with mandatory installation has discouraged other municipalities from following suit, and even in ordinance cities, installation rates have stalled at low levels," according to the report.

Standard revision

In October 1997, UL released a revised standard that became effective for alarms manufactured after October 1998. This revision requires that individual alarms be subject to multiple tests and that sensitivity testing and selectivity testing be performed using the same alarm.

But according to the report, "We conclude that the overall performance of alarms offered to the public has not improved with the new revision. Many UL-certified alarms continue to fail the basic sensitivity requirement of the UL standard.

"The incorporation of a UL 2034 standard requirement for time-of-manufacture and in-service testing will go a long way to assuring that CO alarms fulfill their basic function. Unlike 2034, the recently developed Canadian Standards Association 6.19-01 Standard for Residential CO Alarms has incorporated key CPSC and GRI recommendations for reliability, including specifications for alarm longevity and time-of-manufacture testing."

Of the 71 units and 14 brands tested in the study:

  • 4% alarmed in clean air;

  • 35% failed to alarm at hazardous CO levels;

  • 37% failed to work properly when tested at an absolute humidity of 1,320 ppm (two times greater than the lowest absolute humidity required by the UL 2034 standard);

  • Three of seven retail brands with digital displays were accurate to ? 30%, with one of the three accurate to within ? 10% of the true concentration. The other four brands were not adequately accurate; and

  • Seven brands showed decreased sensitivity to steadily increasing CO concentrations in comparison to fixed concentrations.

    Three brands identified as "best-performing" consistently worked within the UL 2034 specification. They never exhibited a false alarm, and none failed to alarm at dangerous CO doses. However, two of these brands, including the best performing brand with the most accurate digital display, have become unavailable on the market and are no longer available to consumers.

    Alarms of the six most poorly performing retail brands often failed to actuate at dangerous CO levels.

    ASHRAE not ready to require

    The issue of CO detector requirements was discussed at a Jan. 24 special session during the American Society of Heating, Refrigerating and Air-Conditioning Engineers Winter Meeting in Chicago. The meeting was held by the committee writing the proposed standard 62.2P, "Ventilation and Acceptable Indoor Air Quality in Low-Rise Residential Buildings."

    At various times in its preparation, the proposed standard has included a requirement that carbon monoxide alarm devices be installed in all residences. The current draft of the proposed standard, recently released for its fourth public review, does not contain requirements for the devices.

    The requirement was removed following the first public review. Commentors expressed concern about the unreliability and expense of alarms, and said that alarms are not needed if there are requirements on carbon monoxide sources.

    Members have rejected requests to return the requirement to the proposed standard based on reports questioning the reliability of devices currently in the marketplace.

    "To partially accommodate commentors on both sides of the issue, the committee included no device requirements in the current draft, but agreed to reopen the issue when the CO alarm industry could demonstrate a more reliable set of products," said Dr. David Grimsrud, chair of the committee's indoor air quality subcommittee.

    The committee recently was contacted by the industry with information that new, reliable devices are now available in the marketplace.

    Unless publication of the currently draft was significantly delayed, the issue of CO devices would be addressed through future addenda to the proposed standard, according to committee Chairman Max Sherman.

    Results of the ASHRAE session were not available at press time.

    The CO-Experts "Low-level carbon monoxide `health' monitor" provides a digital readout for CO levels at 5 ppm and above, and alarms visually and audibly beginning at 10 ppm.

    Better testing recommended

    One of the sharpest critics of consumer-grade residential CO monitors is George E. Kerr, president of CO-Experts. Kerr has been involved in "early detection" for almost 50 years and has served as an advisor to the U. S. Consumer Product Safety Commission, the American Gas Association, and the Canadian Gas Association, is a voting member of the CSA 6.19 Technical Committee on CO Alarms, and serves on the UL-2034 STP on Carbon Monoxide Alarms and Gas Detectors.

    Kerr designed and manufactures a CO monitor that is not approved by UL. It could not be approved since it detects and warns of CO levels far below those allowed by UL Standard 2034. Thus, for those required to install UL-approved monitors, Kerr recommends installing his unit alongside, but not in place of, the approved retail units.

    Kerr warned SNIPS readers to "cover their rear ends carefully" if they decide to recommend or install consumer-grade CO monitors. He said that units without displays, and UL 2034-listed units that will not display CO readings below 30 ppm, are not suited for the trade.

    Kerr wants to see wholesale changes in the way consumer units are tested. "Manufacturers need to put their money where their mouths are," he said. "They need to test units at the time of manufacture. Retail testing is too late. The lifetime of the unit could have expired by the time you buy it. They need to retest for effectiveness every 90 days for the declared lifetime of the unit. That could be up to 20 times in one lifetime."

    Kerr admitted that he is considered a purist by some of those involved in setting standards, but he makes no apologies. "Unless we push for improvements, we won't get them."