Page 424 - Dust Explosions in the Process Industries

Ignition of Dust Clouds and Dust Deposits 39 7

5.2.2.2
Dust Deposit on a Hot Surface at Constant Temperature

Miron and Lazzara (1 988) determined minimum ignition temperatures for dust layers
on a hot surface, for several dust types, using the method recommended by the
International Electrotechnical Commission and described in Chapter 7. The materials
tested included dusts of coal and three oil shales, lycopodium spores, corn starch, grain
dust, and brass powder. For a few of the dusts, the effects of particle size and layer thick-
ness on the minimum ignition temperatures were examined. The minimum hot-surface
ignition temperatures of 12.7 mm thick layers of these dusts, except grain dust and corn
starch, ranged from 160°C for brass to 190°C for oil shale. Flaming combustion was
observed only with the brass powder. The minimum ignition temperatures decreasedwith
thicker layers and with smaller particle sizes. Some difficulties were encountered with
the corn starch and grain dusts. During heating, the starch charred and expanded,whereas
the grain dust swelled and distorted. The test was found acceptable for determiningthe
minimum layer ignition temperature of a variety of dusts.
Tyler and Henderson (1987) conducted a laboratory-scalestudy to determinethe min-
imum hot-plate temperatures for inducing self-ignition in 5-40 mm thick layers of
sodium dithionitehnert mixtures. The kinetic parameters for the various mixing ratios
were determined independently using differential scanning calorimetry (DSC) in both
scanning and isothermal modes and by isothermal decomposition tests. This allowed
measured minimum hot-plate temperatures for ignitionto be compared with correspond-
ing values calculated from theory, using a modified version of the Tyler-Jones computer
simulation code. The code required no approximation of the temperature dependence,
and reactant consumption was accounted for assuming first-order kinetics.
Tyler and Henderson found that the minimum hot-platetemperatures for ignition were
significantly affected by the airflow conditions at the upper boundary, as predicted by
theory. This must be allowed for when interpreting or extrapolating experimental data.
It was further found that the simple model of Thomas and Bowes can be used to inter-
pret experimental results, even when appreciablereactant consumption occurs.
Henderson and Tyler (1988) observed that, for certain types of dust, different exper-
imental routes for the determinationof the minimum ignition temperature of a dust layer
can lead to widely differingexperimentalvalues. For sodiumdithionite,experimentsstart-
ing at a high temperature and working down led to an apparent minimum ignition tem-
perature of nearly 400"C, compared to a value of about 190°Cwhen experimentsstarted
at a low temperature, working up. The cause of this behavior was the two-stage decom-
position of sodium dithionite and the problems with preparing the dust layer on the hot-
plate fast enough for the first-stage temperature rise to be observable at high plate
temperatures,in the range 350400°C. Similarbehavior may be expectedfrom some other
materials.

5.2.2.3
Constant Heat Flux Ignition Source in the Dust Deposit

As pointed out by Beever (1984), situations may arise in industry where hot surfaces
on which dust accumulates should be described as constant heat flux surfaces rather
than as surfaces at constant temperature. Beever mentioned casings of electric motors,