Investigation of an approach to fuel identification for non-flaming sources using light-scattering and ionization smoke detector response

Smoke detector performance depends on the detector and smoke characteristics as two decoupled phenomena. In this work, smoke properties on which various smoke detection principles are based were simultaneously measured in an enclosure using existing reference smoke detection systems. The measured smoke properties were scattered light intensity, response of measuring ionization chamber (MIG), obscuration, smoke particle density, and fuel mass loss. Data were collected for conventional fuels such as wood, paper, nylon, weal, polyurethane and PMMA, and also for unconventional fuels such as cooking oil and bread. The fuels were ranked nearly consistently, but in reverse order, based on ionization and light scattering reference detector responses. Discrimination of fuels using a simplified Mie scattering theory was examined. The analysis indicates that the Mie scattering coefficients are fuel-dependent and may be essentially unique to each fuel; however, an overall classification of different fuels using these coefficients may not be practical and requires more precise measurements. (C) 1997 Elsevier Science Ltd.