In 1992, the Building Regulations were amended requiring every new build to allow for mains-wired, interconnected smoke alarms to be installed. With many alarms installed under this Regulation still in use and potentially approaching their twentieth year, it was necessary to research a recommendation as to when mains powered and battery operated smoke alarms should be replaced.
The majority of research found on this subject emanates from the US. The US National Fire Protection Association (NFPA) have issued NFPA Standard 72, National Fire Alarm and Signalling Code, which states that all smoke alarms (including alarms that use 10 year batteries, and those that are hard wired) should be replaced when they are ten years old, or sooner if they don't respond properly when tested.
US fire safety websites, along with those in the UK, Canada, New Zealand and Australia, tend to recommend the replacement of domestic smoke alarms, whether battery or mains-wired, when they:
Several reasons are provided to justify the replacement of smoke alarms after ten years.
The Minnesota Department of Public Safety cite a nationwide study undertaken by the Consumer Product Safety Commission (CPSC), which states that 97% of smoke alarms should still be functioning after one year, if supplied with power. After ten years it is 73%, whereas after 20 years, this figure stands at 54%. The study also indicated that 60% of the failures were due to flat or removed batteries or a disconnected power supply and the study offered possible reasons for this. Ageing alarms may experience sensitivity drifting, which may, in turn, result in an increased frequency of accidental activation and an increase in people removing the power supply. It was thought that newer alarms with a 'hush' feature may contribute to remedying this.
An NFPA report cites a study undertaken by Canada's Ontario Housing Corporation supporting the fact that 3% of smoke alarms will fail within one year. They also say that after 30 years, nearly all the alarms will have failed. They conclude that replacement after ten years, with roughly a 30% probability of failure, is an appropriate balance between safety and cost.
The South Australian Metropolitan Fire Service state that smoke alarm technology has improved significantly since legislation was introduced (similar requirements to the 1992 UK ones were introduced in Australia in 1995) and replacing old smoke alarms is an ideal opportunity to upgrade smoke alarm systems. The Australian Standard for smoke alarms (AS 3786) specifies an effective life of 10 years, suggesting that after that time, effectiveness may be compromised with accumulated dust, insects, airborne contaminants and corrosion of electrical circuitry.
In the early 1990s, the US Consumer Product Safety Commission undertook an extensive study, called the National Smoke Detector Project, to examine smoke alarm ownership and operability. Some failures were found in smoke alarms, but there were no large or systematic problems identified with detector designs or manufacturing practices that cast any doubt on their long term reliability. However, a variety of component failures, corroded battery clips and deterioration and corrosion of the horn element contacts were found in a few smoke alarms.
Other reports from New Zealand and Canada looked at operability of battery smoke alarms, but no further reports on the operability and longevity of mains-wired smoke alarms could be found.
It is also important to note that any smoke alarm which is certified to BS EN 14604 must have a "replace by" date.
It should be noted that the regular testing of smoke alarms should help identify inoperability, as testing a smoke alarm simulates smoke and does not simply test the power supply. This would indicate smoke alarm maintenance messages remain a priority. However, it was queried whether all smoke alarm test functions operate in this way, or whether some simply do test power supply, be it battery or mains-wired.
The general view from UK based smoke alarm manufacturers reflects the findings elsewhere. With contaminants such as dust, insects, grease and nicotine, the smoke alarm chamber is susceptible to becoming excessively sensitive or insensitive. This may lead to either an increase in nuisance false alarms, or to eventually becoming unable to detect smoke. One manufacturer reports that contamination is extremely variable, but that field experience indicated that 10 years is a reasonable compromise.
Evidence of smoke alarm longevity appears to be scarce and inconclusive. As with other electronic items, there will be failures in the units when they are produced and failures during their lifetime due to individual component faults. Similarly, as they get older more faults are likely to occur. Problems specific to smoke detectors include increased sensitivity, causing false alarms.
In all the work identified so far, none has specifically concentrated on the failure of smoke alarms when they age. Smoke alarms do fail, but the rate at which they do has not been accurately determined or related to their age.
Despite there not being much research, it does seem appropriate to replace smoke alarms after ten years (in line with manufacturers advice), unless individual alarm testing suggests earlier replacement.
Safelincs has developed products designed to help facilitate the process of replacing the smoke detector and heat detector heads for mains-wired interconnected alarms.
Reviewed: 05/02/2021 (doc:43 V1.1). Our articles are reviewed regularly. However, any changes made to standards or legislation following the review date will not have been considered. Please note that we provide abridged, easy-to-understand guidance. To make detailed decisions about your fire safety provisions, you might require further advice or need to consult the full standards and legislation.