Practical comparison of smoke detector sensitivity standards

In 1995 it was proposed that the British Standard for domestic smoke alarms be revised in order to bring the fire sensitivity tests established in the British Standard, BS 5446 : Part 1 : 1990, into line with those in the British/European standard for commercial point-type smoke sensors, BS 5445/EN 54 : Part 9 : 1984 and to reduce nuisance alarms which smoke detector manufacturers considered to be a problem. The UK Home Office wished to be reassured that a reduction in the sensitivity would not constitute a significant reduction in the level of life safety provided by the devices. A series of eleven tests were conducted to examine the performance of a range of smoke detectors. The fires were selected to closely match the fire sensitivity tests specified in the BS/EN/ISO stand-ards or as representative of realistic fire scenarios. The tests were conducted in a detached three-bedroom dwelling of typical 1970’s UK design and construction. Optical and ionisation domestic smoke alarms and commercial analogue addressable smoke sensors, complying with UK and US standards, were installed in four locations within the dwelling; in the lounge, in the hall, on the landing and in a bedroom. The optical density of smoke close to the detectors was also measured. In each location, the times to alarm for each detector and the time at which the density of the smoke reached a hazardous level were recorded. The time between an alarm and the onset of the smoke density hazard was calculated for each smoke detector. These data were analysed by com-paring the performance of individual smoke detectors, by using standard statistical methods for groups of detectors and by ranking the relative performance of each smoke detector across a range of tests. For detectors complying with the UK domestic and commercial standards, analogue sensors responded as well as or, in some cases, better than the domestic smoke alarms depending on the type of fire. The overall performance of the analogue and domestic detectors were statistically similar. The ranking analysis indicated that the analogue optical sensors performed better than the domestic optical smoke alarms whereas the ionisation detectors exhibited a similar performance. Therefore, revising the domestic smoke alarm requirements to match the commercial smoke sensor standard would not reduce the life safety effectiveness of domestic smoke alarms. Analysis of the performance of the detectors complying with the US standards found that the domestic ionisation smoke alarm and the analogue ionisation smoke sensor showed a statistically significant variability. The ranking analysis indicated that the domestic ionisation smoke alarm detected smoke earlier than the analogue ionisation smoke sensor. In the case of the optical devices, the overall performance of the two types was statistically similar. However, the ranking analysis indicated that the analogue optical sensor performed better in some types of fire scenario