CH-99-08-1 -- Smoke Movement and Detector Activation in High Bay Spaces PDF

A series of fire experiments were conducted in two aircraft hangars with ceiling heights of 15 m (50 ft) and 22 m (73 ft). The purpose of the experiments was to analyze the activation characteristics of smoke and heat detectors in response to JP-5 and JP-8 pool fires. The 15 m (50 ft) hangar was located in Hawaii, where ambient temperatures were approximately 30°C (86°F). The 15 m (50 ft) experiments used fire sizes that ranged from 100 kW (95 Btu/s) to 7.7 MW (7300 Btu/s). Experiments were conducted with and without draft curtains in the 15 m (50 ft) hangar. The 22 m (73 ft) hangar was located in Iceland, where ambient temperatures were approximately 12°C (54°F). The 22 m (73 ft) experiments used fire sizes that ranged from 100 kW (95 Btu/s) to 33 MW (31000 Btu/s). Draft curtains were present for all the 22 m (73 ft) experiments. Open- and closed-door fire experiments were conducted in both hangars.

Commercial detectors used in the series of experiments included spot smoke and heat detectors, bulb and fusible link elements, projected beam smoke detectors, UV/IR optical flame detectors, and a line-type heat detector. Other instrumentation included thermocouples, mass flow meters, and radiometers.

The analysis of these experiments has led to the following observations:

1. Draft curtains improved the response time of heat detectors and sprinklers at these ceiling heights and reduced the size of the threshold fire needed for activation. Both the plume centerline temperature and the ceiling jet temperature increased in response to the growing smoke layer.
2. Standard response sprinklers were either activated substantially slower or not at all when compared to the activation of quick-response sprinklers at these heights.
3. Trouble windows used for beam-type smoke detectors gave false trouble signals in the presence of dense smoke from JP-5 fires.
4. Tests conducted in the presence of wind and open hangar doors showed that ceiling jet temperatures were substantially reduced but that downwind smoke detectors continued to activate for small fire sizes. Wind speeds inside the hangar ranged from 2 km/h to 32 km/h (1 mph to 20 mph).

Based on the observed detector activation, spacing for both spot smoke and heat detectors at these heights was analyzed.

Units: Dual