Effectiveness of Smoke Alarms in Saving Lives

(Posting Date: 01/12/10)

CTV's "W5" program on January 4, 2000 presented "tests" of eleven models of smoke alarms, typically sold to consumers through retail stores for use in homes. The "tests" appeared to indicate that  "ionization" type smoke alarms are not effective in detecting smoldering type fires, and do not meet Canadian standards.

The Canadian Fire Alarm Association (CFAA) takes great issue with the fact that an impression may have been conveyed to the public that ionization-type smoke alarms are ineffective, and that U.L.C. may not have properly tested the model types that they approved. We do not believe such conclusions are appropriate, based on the investigation W5 undertook.

Questions have been raised by various experts and industry participants, regarding the "test" conditions and methods. Many criticisms have been leveled at the assumptions made in the testing, and the lack of testing of photoelectric detectors, to compare their performance with that of ionization types.

The evidence is that both ionization and photoelectric type smoke alarms work, and do save lives. We believe that "W5" may have undermined the public's confidence in  the reliability and performance both of ionization smoke alarms and of the Canadian consensus standards/certification process.  The limited information we have about their "test" does not lead us to conclude that this undermining is justified.

We understand that W5 have for quite some time declined  written requests from John Roberts, President of ULC, and from Ontario Fire Marshal Bernard Moyle, (on behalf of the Council of Canadian Fire Marshals and Fire Commissioners), asking that the W-5 fire test data and tested smoke alarms be provided to ULC for evaluation and validation to the appropriate Canadian Standard, "CAN/ULC-S531". We are concerned that W5, with an objective of public education and serving the public interest, would not want to provide the data requested so that it could be reviewed and any issues resolved as quickly as possible in a co-operative, constructive way.

Re-testing of three randomly purchased samples of each of the eleven smoke alarms has been undertaken by U.L.C., for and with representatives of Health Canada, and other government and industry officials. More than 192 individual fire tests were involved. U.L.C., on February 4 reported that 100% of the smoke alarms tested passed the CAN/ULC-S531 tests. It would therefore appear that the integrity of U.L.C.'s testing practices remains sound notwithstanding W5's findings.

BACKGROUND:   Standards Development and Testing in Canada

Our Canadian standards and certification processes have earned a worldwide reputation for reliability, consensus, expertise and independence.  They represent a vital element within Canada's safety system.  The Canadian standard relating to Smoke Alarms, CAN/ULC-S531, incorporates a series of comprehensive tests in 20 construction categories, involving 110 individual tests relating to audibility; sensitivity; performance under various environmental conditions; interference; accelerated aging; mechanical and false alarm reduction concerns.

Underwriters Laboratories of Canada, (U.L.C.), is accredited by the Standards Council of Canada, as a testing and certification organization.  In this capacity ULC tests smoke alarms in accordance with S-531 and certifies them for ongoing compliance with this standard.

This includes ongoing unannounced visits to the smoke alarms manufacturing facilities to check quality control procedures, production records and to select samples for counter-checking testing at U.L.C.'s laboratories.

 The U.L.C. mark is a symbol of compliance with the official Canadian standard, and we believe the public can have confidence in both ionization and photoelectric smoke alarms bearing this mark, provided that these devices are regularly maintained, are properly installed, are operational, and have not exceeded their expected life span.

CAN/ULC-S531 came into effect in 1987, and has undergone three amendments since then.  The standard was developed by a committee of Canadian professionals from across Canada, under the secretariat of the Underwriters Laboratories of Canada, with extensive input from consumer groups, fire safety experts, government agencies, manufacturers, insurers, fire authorities and others involved in the consensus standards process.

It is our belief that members of the CAN/ULC-S531 sub-committee are dedicated, professional, focused and very objective. They are volunteers from diverse disciplines across Canada, and function under strict guidelines from the Standards Council of Canada and U.L.C. The committee's "bias", if any, is toward making Canada a safer place to live, work and play.

Indiana Dunes Test Project Conducted in 1975 and 1976

The fire test parameters ultimately Incorporated into CAN/ULC-S531 were developed to a great extent based on the findings of a comprehensive "real world fire test project" known as the Indiana Dunes fire tests. This was a costly US Government sponsored program designed to study fire phenomena and develop fire safety standards for residential applications.

Conducted in 1975 and 1976, this project involved 76 different real world fire scenarios using real world combustibles and ignition sources in real world residences.  These fire tests covered every conceivable type of fire encountered in residential applications.  Utilizing the latest scientific equipment Richard Bukowski, Richard Bright and a number of the top fire protection engineers and researchers in North America thoroughly studied, evaluated and profiled the fire growth, smoke density, temperature development, aging properties, alarms responses, smoke spread and related fire and smoke parameters. They tested a variety of fire "signatures", combustible materials, construction configurations and ignition sources.

All combustibles and ignition sources used in these fires were authentic. Throughout these tests the performance and response capabilities of thermal, photoelectric and ionization detectors were measured and recorded.

Each fire profile was completely documented and thoroughly evaluated by international fire officials and was segmented into common types for verification.

The results of the tests indicated that different kinds of fires, in different circumstances, produced different patterns of smoke and heat development. This was taken into account as standards and tests were developed for smoke alarms, which is why a variety of fires must be detected, both flaming and smoldering type, according to present

It should be remembered that Richard Bukowski, Senior Engineer at the National Institute of Standards and Technology, a leading member of the Indiana Dunes tests and one of the World's leading authorities and researcher on smoke detector and smoke alarm technology, made the following statement on the W-5 show on January 4th, 2000;

"Our fire death rate in the United States has declined by 50% over the last 25 years.  More than 90% of the smoke alarms out there are ionization type, so that would seem to indicate that they are doing a pretty good job in most cases."

Internationally, CAN/ULC-S531 is widely recognized as a reliable, consistent fire test and performance standard for smoke alarms.  It is closely harmonized with the equivalent smoke alarm standard in the USA and is periodically verified through statistical analysis, research, fire modeling and testing performed by the leading fire protection engineers, academia, fire investigators, educators, researchers and related specialists from across Canada and around the world.

S-531 Fire Tests for Ionization and Photoelectric Smoke Alarms - Overview

a) Flaming Fire Tests

Three smoke alarms set at minimum sensitivity settings are installed 460mm apart on the ceiling of the draft-free fire test suite, with all heating, humidification and air conditioning closed off.  All three samples must respond to each test of the following fire "signatures" in the designated time cycle:

"Paper Fire" (flaming fire)
Newsprint, approximately 28g, black print.

"Flammable Liquid" (black fire)
A mixture of 20mL of 25% toluene and 75% heptane.

b) Smoldering Fire Tests

Three smoke alarms set at minimum sensitivity settings are installed 460mm apart on the ceiling of the draft free fire test suite.  The ambient temperature of the suite is between 16°C and 27°C and the relative humidity is 50 + 20%.

"White Smoke"
Six pieces of white pine on a hot plate covered with a steel plate.

The rate of smoke build-up is controlled to be approximately uniform and maintained.  All three samples shall respond before the smoke density exceeds 6.0% / 304mm obscuration (6% / ft).

Both Ionization and Photoelectric Smoke Alarms are Subjected to the Same S531 Standard
Both technologies are subjected to the identical standard and testing criteria, and both are capable of detecting all types of fires that commonly occur in the home.

 Ionization Smoke Alarms
 (  Fastest type to respond to flaming fires
 (  Generally faster to respond to black smoldering fires involving "man-made"
        materials such as PVC wiring, wall coverings and carpeting
 (  Generally slightly slower to respond to white or gray smoky fires involving
    natural materials such as wood, cotton, etc.
 (  Lowest cost and most commonly sold (more than 90% market share)
 (  Some models have a "hush" or temporary silence feature
 (  Some models are available with a long life (10 year) battery

 Photoelectric Smoke Alarms
 (  Generally faster to respond to white/gray slow smoldering fires
 (  Less prone to nuisance alarms from cooking

Installing both types of smoke alarms in your home can enhance fire safety.

Leading causes of fire deaths in Canada are:

* Residential fires caused by smoking materials that ignite upholstered furniture in the living or family room at night,
* Smoking materials igniting bedding in sleeping areas at night, and
* Cooking appliances left unattended at night.

Although some of these incidents may initiate smoldering fires, they often develop into flaming fires. According to sources at the National Fire Protection Association, most people do not die from extended smoldering fire conditions. They die from the lethal conditions related to flaming fires.

Ionization Smoke Alarms - They Do Save Lives

The evidence is that ionization smoke alarms really work!  They do save lives. Thousands of people around the globe owe their lives to these devices. We believe it is unfortunate that W5 presented the results of their tests in such a way that they may have undermined public confidence in smoke alarms in general. Our perspective is that they should have more fully considered historical evidence, and undertaken more extensive, balanced research.

We conclude with some interesting statistics, and some Life safety tips!

In 1999, Ontario recorded its lowest fire death rate in the province's history.  This reassuring fact continues a 20-year trend that has seen the fire death rate in this province drop by more than 60 percent.

The Fire Marshal for Ontario, Bernard Moyle, attributes this decrease to enhanced codes / standards; fire safety education and most importantly to the installation of smoke alarms (mostly ionization type) in Ontario homes.

This trend is repeated from shore to shore in Canada and throughout the United States.

According to the National Safety Council's "Injury Facts" report, for the period from 1968 to 1998, the annual fire death rate in North America fell approximately 60%- - - thanks in no small measure to ionization smoke alarms.

Fire Safety Tips

* Install at least one operating smoke alarm on each level of your home plus one outside each sleeping area (preferably one in every room of the house).
* Test each smoke alarm at least monthly.
* Replace the battery twice annually.
* Replace all smoke alarms older than 10 years.

The best defense against fire is not to have one in the first place.  Fire and smoke travel very quickly, often with tragic consequences. Therefore, you must also develop a fire escape plan and practice it with your family.  If fire does occur, get out and stay out and phone 911 or your local fire department.