SMOKE DATA DETERMINATION FOR VARIOUS TYPES OF FUEL

Fire protection assessment is of major interest in industrial environm ents. Among the resulting threats, smoke production is the object of g rowing attention. In this context, a proper analysis of smoke generate d by four of the most commonly used oils in the Electricite de France production units, as well as heating oil, often implicated in everyday life accidental fires, is reported. The convenient relative measure o f smoke emission, the specific extinction area, is identified as the v ariable to be used. Results show that all these fuels have a high prop ensity to generate soot particles. The evolution of extinction coeffic ient and wavelength, modelled as K(e)similar to lambda(-n), reveals sm all-size, mainly absorbing particles. However, the dispersion coeffici ent n appears not spectrally constant. Despite the few data points, gi ving an exploratory nature to the work, a slope discontinuity of the c urves, more or less marked as a function of the atomic ratio H/C of th e fuel, appears between the visible and near infrared range of the spe ctrum. Theoretical predictions using appropriate values of soot optica l constants show a satisfactory agreement with experiment. Beyond a si mple classification of the different fuels, the results are valuable a s input values in radiative exchange modelling. (C) 1998 Published by Elsevier Science Ltd. All rights reserved.