A first approximation method for smoke detector placement based on design fire size, critical velocity, and detector aerosol entry lag time

An analog light-scattering-type smoke detector was tested in a wind tunnel at various low velocities. The air flow in the wind tunnel contained an aerosol concentration that resulted in a high ambient optical density, simulating smoke well above threshold optical detector density. The objective of this research was to determine the lag time to alarm, Δt, associated with difficulty of smoke entry into a detector. A "critical velocity" was identified for the smoke detector, below which the lag time increased exponentially with decreasing velocity. Increased lag time results in the detector responding unacceptably late-or not at all-even when ambient obscuration is well above limits defined in UL standard tests. A preliminary method for placing smoke detectors has been developed, based on a user-defined design fire size and the detector aerosol-entry lag time. The preliminary method applies only to flaming fires producing smoke, with the detector far from a wall and mounted on smooth ceilings. The critical velocity value used in the examples in this paper applies only to the smoke detector configuration tested in this work, at the evaluated sensitivity setting, with the optical densities reached using the generated artificial smoke. Any variation in detector housing, design, operation principle, or application with different aerosols requires specific tests to determine a different critical velocity. This paper provides the basis for further development of a smoke detector placement method based on a design fire size and the proposed detector critical velocity concept. It does not presume to report a unique "critical velocity" for all smoke detectors, but suggests that such a value may indeed exist, but with differing values among different smoke detectors.