Page 624 - Dust Explosions in the Process Industries

Research and Qeveiopment 59 1

of spontaneous ignition in an auto-oxidative powder bed and found that the critical state
for ignition could be classifiedinto two categories:the Frank-Kamenetskiiand the oxygen-
deficient types. The first implies an irreversible transient process at the critical point, lead-
ing to ignition. Its thermal stabilityis rather frail against the changes in ambient conditions.
In the second type, oxygen diffusion controls the rate of heat generation in the bed, and
the thermal stability is comparatively robust. The two types can be largely discri
by considering the nature of the combustible material and the porosity of the bed.
Scheidemann and Adomeit (1996) presented a mathematical analysis of the transfor-
mation of a slow smoldering$re inside a heap of carbon dust into an openjire, by the
impact of an airflow on the side of the heap. The model does not consider the transfor-
mation of the fire into a dust explosion should a sufficiently strong airflow disperse the
dust in the heap into an explosive dust cloud. In their studies of the efect ofadmixed inert
material on the minimum hot-plate ignition temperature of coal dust, Reddy, Amyotte,
and Pegg (1998) successfully applied the classic ignition model developed by Thomas
and Bowes.
Li and Xiao (1999) developed a mathematical and numerical simulation model to
predict the self-heating behavior of milk powder deposits of low moisture contents. The
numerical scheme solves the mass and energy balances simultaneously. Model predic-
tions agreed well with experimental data, and it was foreseen that the model could also
be applied to other exothermally reactive solids. In a subsequent investigation Li, Xiao,
and Mackereth (1999) studied the effect of aging of milk powders and their fat content
on the self-heating and smoldering properties. It was found that the kinetic parameters
changed with the aging temperature and the tendency to self-heal or smolder increased
with the fat content.
Arisoy and Akgiin (2000) developed a non-steady-state mathematical model for pre-
dicting the safe storage height of coal stockpiles, below which significant self-heating
is not initiated during the finite storage time of the stockpile. The numerical solution pro-
vided by the model is in terms of the maximum temperature within the stockpile as a
function of time.
Anderson, Sleight, and Torero (2000), in their experimental investigation, identified
specific “ignition signatures” that indicate the onset of a self-sustained downward smol-
dering process in a porous material. Polyurethane foam was used as the porous mate-
rial, but the findings are probably valid also for powder and dust deposits prone to
smoldering. The test samples were exposed to a constant heat flux imposed by a cone
heater for different periods of time. Three stages were observed during the ignition
process: (I) warming up and (2) unsteady smoldering, both controlled by the heat flux
supplied from the outside, and (3) self-sustained smoldering, supported by the heat gen-
erated by the smoldering process itself. Each stage was characterized by specific changes
in temperature and mass loss rates (ignition signatures).
Glinka, Klemens, and Wolanski (1993) conducted an experimental, theoretical study
of ignition ofdust layers by thermal radiation. Important features of the ignition process
were resolved in detail by means of high-speed Mach-Zehnder interferometry.
In a series of unique experiments carried out in a space shuttle, Walther et al. (1999)
studied smoldering combustion processes under microgravity conditions. The objective
of the study was to achievea betterunderstanding of the variousmechanisms that control smol-
dering processes, to provide improved means to prevent and control such processes. The
data from the microgravity experiments were compared with corresponding experiments