Skip directly to content

Carbon Monoxide Alarm

Carbon Monoxide Alarms and its History

How does a carbon monoxide alarm Work?

The history of the carbon monoxide (CO) alarm starts with a long term development of the CO sensor and the alarm market.  Quantum Group Inc. (Quantum) was founded in 1982 to focus on forensic science for litigation.  One of Quantum’s customers wanted a CO safety shut off device because he was paying out 10% of sales in CO litigation settlements.  This company was Atlanta Stove Works (ASW).  In 1984, ASW and the Gas Research Institute (GRI) partially funded the CO sensor development with about $1 million each, which was injected over a period of about 4 years. Quantum and several utilities companies also contributed money to the development cost. This led to a carbon monoxide alarm that could also shut off appliances. These products were field tested in 1986.  Later improvements led to a ventilation control that was launched in 1987.  Next, Quantum Eye®, a visual color indicating carbon monoxide badge, was launched in 1989 along with field testing of the carbon monoxide alarms. The CO alarms passed the American Gas Association Laboratory (AGA) tests as well as shock and vibration tests performed by other nationally recognized laboratories including Wylie Labs.  In 1990, these CO alarms were sold to Fleetwood for the recreational vehicle (RV) market. These first CO alarms used a photoelectric smoke alarm chip technology (known as an Application Specific Integrated Circuit (ASIC) made by Motorola.

Today carbon monoxide alarms use electronic circuits usually involving a small microprocessor to calculate when to initiate alarm signals based on exposures to hazardous CO concentration and time.  An accumulation of CO over time must be calculated based on a formula similar to the Coburn equation or a simple look up table.  The graph and the equation showing the relationship between CO concentrations and exposure times versus percent carboxyhemoglobin (%COHb) are given in the Underwriters Laboratories (UL) 2034 CARBON MONOXIDE ALARM standard, which costs over $300 to purchase. 

Under certain conditions for short term exposures to elevated levels of CO, i.e., the %COHb levels are below the saturation levels: one may assume there is a linear relationship of CO concentration uptake. This assumption does not hold for low levels over long time periods and it is best to use the full Coburn-Forster-Kane (CFK) equation Please Click to see the complete CFK equation(www.coheadquarters.com/CFKEqu1.htm ).  

One simplification of CFK equation that has been used is the following:

%COHb= (3.317 x 10-5) X (CO ppm) X 1.036 X (RV) X (t)

Where:

CO ppm = carbon monoxide concentration in parts per million (ppm)

RV (Respiratory Volume) = volume of air breathed in liters per minute (lpm) (Typical adult is about 5.2 lpm)

t = exposure time in minutes

This formula was taken from Stewart, R.D., Peterson, J.E., Fisher, T.N., Hosko, M.J., Baretta, E.D., Dodd, H.C., Herrmann, A.A., "Experimental Human Exposure to High Concentrations of Carbon Monoxide", Arch. Environ. Health, 26, 1-7, 1973.

CO detection technology may be based on chemical reactions causing an optical change, an electrochemical change or a semiconductor change in resistivity.  Changes in any of these signals in response to CO may be used to trigger an alarm.  Most carbon monoxide alarms require an external power supply to detect the presence of CO.  These types of carbon monoxide alarms will not function without a battery or other power source.  However, today it is possible to use a lithium battery that will last for the life of the product, which typically ranges from5 to 7 years.   We know from experience with smoke alarms that the majority of fire deaths in homes are the result of missing or dead batteries. For this reason CO alarm product with a long life power supply is clearly needed.  Regulations requiring carbon monoxide alarms to have a tamper resistant battery such as one soldered to the  printed circuit (PC) board are expected to follow the smoke alarm requirement in California SB 1394 (the law states that after Jan. 1, 2014 all battery powered smoke alarms must be able to be powered for ten years by a lithium battery.  Ten (10) years is the life of a smoke alarm while 6 years is the typical life of a carbon monoxide (CO) alarm.   Even though a carbon monoxide alarm with a long life battery is not required yet, you may soon be able to purchase one by 2014.  For example, a lithium battery powered CO alarm from Quantum Group Inc. and other manufacturers are currently in design and development phase and should be available by then.

The carbon monoxide alarm is regulated by an ANSI standard, which is also known as the UL 2034 carbon monoxide alarm standard.  It deals with standalone CO alarms of all types including battery and AC with battery backup (hardwired and plug-ins, i.e., each CO alarm has a power source but it is not a part of a system with a central panel such as a security or fire panel.  CO alarms for systems are regulated by UL 2075. Interconnect AC powered carbon monoxide alarms are not considered systems as there is no central panel.

Carbon monoxide is extremely poisonous and the health effects are often deadly.  Even though many survive they are most often left with diminished mental capacity and other long term symptoms. 

The current monoxide carbon monoxide alarm Standard, ANSI/UL 2034,  is based on % carboxyhemoglobin (%COHb) in the blood as a biomarker for time to alarm.  This standard states clearly no alarm below 5% COHb but must alarm before 10% COHb is reached.  The carbon monoxide exposure  caoable of creating carboxyhemoglobin levels in a human between 5 to 10 percent  determines when the alarm signals. Please see CO concentration over time for the UL Standard actual tests are summarized the following Table 1 below and the graph from the LBNL While Paper.  Additionally, in 2009, UL2034 standard was also revised to add a CO ramp test to simulate a real-world situation where CO concentration is expected to rise at a semi-constant rate of 0 to almost 500 ppm in 30 minutes.

Table 1 UL Carbon Monoxide Gas Tests

CO Concentration (PPM)

Exposure Duration

Alarm Limits

30

30 days

Must not alarm

70

4 hours (240 minutes)

Must alarm within 60 to 240 minutes

150

50 minutes

Must alarm within10 to 50 minutes

400

15 minutes

Must alarm within 4 to 15 minutes

 

The CO ramp test involves a continuous flow of CO at a rate of 16 ppm per minute for 30 minutes at which point the CO concentration should be between 465 to 495 ppm and must be maintained within this range for an additional 3 minutes after the CO supply is cut off At this concentration the alarm should respond after 19 and before 30 minutes.  These gas tests are performed at various environmental condition including temperature and humidity from minus 40 °C (-40 °F) to plus 70 °C (158 °F) and several other in between.  The relative humidity is tested under extreme and normal temperature from 7.5% to 95% RH.  For example, the high humidity test for residential alarms is 52 °C at 95% humidity preconditioned for 168 hours (7 days) and then run all four carbon monoxide exposure tests in Table 1 above.  The RV (or unconditioned space such as garage, car, attic) test condition for high humidity is 66 C at 93% RH for 10 days preconditioning followed by the same CO tests in Table 1.  This is a very extreme test and no people could survive the environment.  These gas tests are performed at various environmental condition including temperature and humidity from minus 40 °C (-40 °F) to plus 70 °C (158 °F) and several other in between.  The relative humidity is tested under extreme and normal temperature from 7.5% to 95% RH.

History of the Carbon Monoxide Alarm

Over 150 years ago Sir William Robert Grove was born July 1811 and passed away August 1896. He was an outstanding physical scientist but also pursued a legal career.  Sir Grove anticipated the theory of the conservation of energy, and was the pioneer in fuel cell technology.  His fuel cell running backwards is an electrochemical sensor. This was an important step to modern sensing. He also invented the first electric light bulb, which was later improved by Thomas Edison. His great contribution was the law of conservation of energy.

The majority of chemical sensors are called electrochemical in nature, which represent almost 58% of the total market today.  Optical sensors represent about 24%, mass based sensors about 12% and thermal sensors about 6% of today’s sensor market, according to the Electrochemical Encyclopedia, Electrochemical Sensor by Jiri Janata, School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, GA 30332-0400.

Dr. Mark K. Goldstein gives Stan Blachman the most credit for the AGA CO standard.  Stan formally head of Standards at the American Gas Association Laboratories near Cleveland, OH and after he retired from AGA, Mr. Blachman became Executive Director of the Carbon Monoxide Safety and Health Association (COSHA) in about the early 1990s.   Dr. Goldstein, one of the inventors of the biotechnology based carbon monoxide sensor met Stan Blachman in 1983 and asked him about the need for carbon monoxide safety shutoff device for the gas industry.  Mr. Blachman encouraged Dr. Goldstein to participate in the Z21.20 Standard committee. In addition, Mr. Blachman introduced Dr. Goldstein to Al Chamberlain of Atlanta Stove Works (ASW) and the Tom Altpeter of the Gas Research Institute who later jointly funded part of the R&D and application engineering.  Both Mr. Chamberlain and Dr. Altpeter contributed to the application of the safety shut off device to gas appliances. During that period of time AGA Laboratories developed requirements for carbon monoxide safety shut off devices and tested prototypes for application with ASW gas heaters and wall furnaces. 

Mr. Bachman also encouraged Dr. Goldstein to work with the furnace committee on a CO alarm standard.  At Mr. Bachman’s urging Dr. Goldstein participated in that Z21.20 standards making committee and related working groups for over 5 years.  Once the committee had completed the standard it was turned over to UL in 1992.  UL made a number of changes to make the standard very sensitive to CO. This would turn out to be a serious mistake.  UL did not listen to the RV or gas industry recommendations.  Later in 1994 they were very embarrassed by too many nuisance complaints and had to change the standard back closer to the way the AGA Labs had originally specified: going from a threshold of 15 PPM in 8 hours to 30 PPM in 30 days. This was a compromise between the gas industry which wanted a 50 PPM threshold and the medical community that wanted a 15 PPM threshold.

This UL 2034 carbon monoxide alarm standard provides alarm requirements for all single station carbon monoxide alarms that are based on the carboxyhemoglobin biomarker as explained above.  Today the standard is much better than when it started and great credit is given to all stakeholders who work hard to perfect this standard.

History of Carbon Monoxide Alarm Standard

Dr. Mark K. Goldstein wants to give Stan Blachman the most credit for the CO alarm standard.  Mr. Blachman was formally head of Standards at the American Gas Association (AGA) Laboratory near Cleveland, OH and after retiring from AGA, he became Executive Director of COSHA in early 1990s.   Dr. Goldstein, one of the inventors of the biotechnology based carbon monoxide sensor, met Mr. Stan Blachman in 1983 and asked him about the need for carbon monoxide safety shut-off devices for the gas industry.  Mr. Blachman encouraged Dr. Goldstein and introduced Dr. Goldstein to Al Chamberlain of Atlanta Stove Works (ASW) as well as Tom Altpeter of the Gas Research Institute (GRI), who later jointly funded part of the CO R&D and application engineering.  Both Mr. Chamberlain and Mr. Altpeter contributed to the application of the safety shut off device to gas appliances. During that period of time AGA Laboratory developed requirements for carbon monoxide safety shut off devices and tested prototypes on Atlantic Stove Works’ for application with ASW gas heaters and wall furnaces.  These product were designed and constructed to the A.G.A, Requirements for Gas Appliances Equipped with Carbon Monoxide Safety Shutoff System No. 3-86, published November 10, 1986 and A.G.A Requirement for Carbon Monoxide Safety Shutoff Systems No. 2-86, published November 10, !986.  The CO test levels in these A.G.A Requirements were 200 ppm for 2 hours and 400 ppm for 30 minutes.  There was also a false alarm level at below 50 ppm.  Many of these products were field tested successfully in 1987 and 1988. 

In 1992, the committee completed the standard and it was turned it over to the Underwriter’s Laboratories (UL).  UL made numerous changes to the standard requiring that CO alarms be very sensitive to low CO concentrations. This would turn out to be a serious mistake.  UL would not listen to the RV industry or the gas industry.  Later in 1994, UL was embarrassed by too many nuisance CO alarm complaints and had to change the standard back closer to the way the AGA Labs had originally specified.

In 1992and 1993, UL modified the AGA draft carbon monoxide alarm standard (UL Standard 2034) and revised it again in 1995 to be closer to the original AGA Laboratories draft. This standard provides alarm requirements for carbon monoxide alarms that are based on the % carboxyhemoglobin biomarker as explained above.

History of UL 2034 Standard

The first carbon monoxide test documents were written by AGA Laboratories and the ANSIZ21.20 furnace committee during the 1980s with major technical inputs from Dr. Mark K. Goldstein President of Quantum Group Inc, Mr. Al Chamberlain, Mr. Stan Blackman and Mr. Donald Switzer of CPSC.  There were several other contributors for editing.  The documents published as a result of then work were published in 1986 by A.G.A Laboratories: A.G.A Requirements for Gas Appliances Equipped with Carbon Monoxide Safety Shutoff System No. 3-86, published November 10, 1986 and A.G.A Requirement for Carbon Monoxide Safety Shutoff Systems No. 2-86, published November 10, !986.  Later the ANSI Z21.20 furnace committee developed a draft standard for carbon monoxide alarm.  This took about 5 years from 1997 to 1991.  Then in 1991/2 time frame the document was given to UL.  UL made modification to the carbon monoxide test levels dropping the 50 ppm.

The original UL 2034 Standard published in 1993 had a false alarm resistance test requirement of at 15 ppm CO for 8 hours, meaning that an alarm signal was allowed after 8 hours at this low concentration.  The alarm levels were set between 2.5% COHb and 10% COHb.  The first CO alarm test level was to signal an alarm pattern between 10 and 90 minutes of exposure to 100 ppm. The second test point was between 10 minutes and 35 minutes of exposure to 200 ppm or with no minimum within 15 min of exposure to 400 ppm. . 

Next UL introduced an option for a warning signal at 2.5% COHb to 7.5% COHb; however, the alarm signal requirement was still between 5% and 10% COHb.  In 1998, after extensive lobbying by the gas industry the carbon monoxide alarm standard was revised again, to dropping all warning signals including any digital readings before alarm levels were reached.  In 2009, the Carbon Monoxide Alarm Standard UL 2034 was revised again to add end-of- life signals.  Please note from 1990 to 1997, Quantum was the only Company that developed, licensed and marketed CO alarms with sensors, which had end-of-life features.  Quantum’s replaceable sensors had end-of-life function tied to a battery life of one to two years.  Then under pressure from customers Quantum agreed to make its CO alarms for 6 year- life because Quantum believed that its biomimetic CO sensors, under extreme conditions, had a limited life of just over 7 years with proper sealing from the atmosphere during storage and an ammonia getter removal system.  However, in 1997-1999 timeframe, one customer extended their advertised life to ten years.  Therefore, Dr. Goldstein began an effort to educate the U.S. Consumer Products Safety Commission (CPSC) and UL about the limited life of all sensing technologies and a need for limited life requirement on all life safety products. 

Quantum followed both Canadian Standard Association (CSA) and UL Standards for carbon monoxide alarms. Just after the turn of the century CSA purchased International Approval Service (IAS) (formally AGA Laboratories) and added an important requirement to its CSA 6.190-01 Carbon Monoxide Alarm Standard.  CSA also added manufacturing quality testing requirement known as Time of Manufacturing (TOMs) and In Service Reliability Test (Shelf Testing).  In addition, CSA added the requirement for a “replace by date” on the side of the carbon monoxide alarms so it can be easily seen by consumers but UL did not adapt this important measure at that time. As a result people may have lost their lives by having alarms with expired sensors in them. Many are still out there so if you do not have a “replace by date” on your CO alarm, please buy one that has that feature.

TOM Reliability Testing is used to sample recently manufactured product over a limited time period.  Product is exposed to CO for a specified ppm level and time exposure that is defined in UL 2034 or CSA 6.19-01. Each quarter of a year is normally defined as the minimum requirement for sampling. The failure rate observed each quarter is then compared to previous quarters and a decision made to proceed as being acceptable or to re-sample the product for additional confidence. This sampling can be done where the product is manufactured and the samples then can be returned to the main production lot.

In-Service Reliability testing is done on a sample of product by using the product, accumulating the time in service of many device hours of the same units originally powered up in an as used condition. The number of failures and the type of failures observed are tabulated and then compared to the failure rate predetermined by statistics to be at a 90% confidence level for compliance with a mean-time-between-failures (MTBF) estimate which is expressed in hours. This reliability procedure must be reinitiated whenever there is a significant change in the design or manufacturing process. In-Service Reliability samples are kept out of commerce as they represent an aged product.

History of Sensing Concentration of Specific Molecules 

In the beginning, canaries were used in the mines perhaps for many centuries.  This was the preferred method to detect poisonous gas such as carbon monoxide.

Next wet chemistry sensing of gas became an available although they were a complex method of a series of chemical reaction and gravimetric measurement.  Classical solutions to chemical sensing tasks have been dominated by complex, expensive laboratory equipment in more modern times.

In the late 18th century, Antoine-Laurent Lavoisier and others developed a new discipline to analyze and identify chemical species including gases.  The analytical methodology was further developed by Carl Fresenius and Karl Friedrich Mohr in the 19th century.  Carl Fresenius developed an extensive qualitative analysis scheme that became the first text book in Analytical Chemistry.  It was published in the about 1848. Fresenius trained students in gravimetric analysis methods that he had invented.  Karl Friedrich Mohr invented many laboratory devices such as the volumetric pipette and titration technology. He invented many colorimetric endpoint tests including the famous silver titrations and published this work in 1855.

In the 20th century, analytical chemistry techniques are based on instrumental methods involving photon spectroscopy instruments. The concentration of many elements and compounds can be measure accurately in gases by the amount of light absorbed or emitted by the molecules or atoms.   In addition, gas chromatographic methods of analytical chemistry separate the components of complex mixtures by using a column of silica or other high surface area material that are designed to interact with the surface in different ways because the chemicals have different interaction with that surface and therefore travel through the column at different times.  Thus we can separate and the identify the  substances coming out using standards and additional instruments such as ionization, flame, mass spectroscopy or ion mobility spectroscopy to determine the concentration of each component in the mixture.

These methods are expensive and require training to operate so there is a need for lower cost and simpler technology that has gradually evolved from analytical chemistry and micro-electronics based in silicon.

During World War II, there was a need for chemical sensing of carbon monoxide in aircraft.  The detection tube consisting of a chemical the reacts and changes color so that you can measure the length of the stain evolved. In the 1991, the first biotechnology based chemical sensors patent was published by Dr. Mark K. Goldstein of Quantum Group Inc. To date over 31 million sensors have been shipped.

There was a strong foundation for semiconductors in the 1950s with the commercialization of the transistor by Bell Labs. In the 1960s, Taguchi invented the semiconductor sensor using doped tin oxide.   These sensors are also used in carbon monoxide and other gas alarms.

In 1864, Wolcott Gibbs developed one of the first electro-analytical chemistry using gravimetric methods to obtain quantitative information based on Faraday's laws. In 1906 Max Cremer invented the glass electrode, which is the first electrochemical sensor in wide use. In the 1970’s electrochemical based industrial detectors and instruments based on electrochemical sensor proliferated.  They are used today in industry and residential applications such as carbon monoxide alarms.

These technologies continue to develop so that they are improvements in existing and the development of new biotechnology sensors.  These advances have led to low cost carbon monoxide alarms for home and other applications.