Page 343 - Dust Explosions in the Process Industries

3 72 Dust Explosions in the Process Industries

4.2.6.3
Experimental Determination of Maximum Explosible Dust Concentration

The results of Palmer and Tonkin (1971) from the large-scale apparatus shown in Figure
4.25 give an indication of the maximum explosibleconcentrationof a coal dust containing
36.4% volatiles on a dry,ash-free basis. Extrapolation from their data for mixtures of
coal and sodium chloride to zero content of the latter indicates a value of 2000-3000 g/m3.
This is on the same order as the value indicated by extrapolatingthe data from stabilized
burner experiments with a similar coal dust (Pittsburgh)in air, presented by Smoot et al.
(1977). These workers measured laminar burning velocities of more than 0.15 m/s even
at 1800 g/m3.
Slezak, Buckius, and Krier (1986), using a tumbling horizontal explosion cylinder of
0.3 m diameter and 4.5 m length, estimated the maximum explosible concentration of
Pittsburgh coal dust in air to be about 1500 g/m3.
However,Cashdollaret al. (1988),using their closed 20 liter explosionvessel, were unable
to detect any maximum explosive dust concentration for Pittsburgh coal up to 4000 g/m3.
They refer to other laboratory and large-scale experiments that confirm this result.
On the other hand, Ishihama, Enomoto, and Sekimoto (1982) could determine maxi-
mum explosibleconcentrations of different noncohesivecoal dust fractions using a rotat-
ing drum apparatus in which the dust cloud was generated continually by being lifted
along the drum wall, subsequentlyfalling freely under gravity. For the particle size frac-
tion 35-50 pm, the maximum explosible concentration in air was 2700 g/m3for a 45%
volatiles coal, 2200 g/m3for 33% volatiles, and 1400 g/m3for 22% volatiles. The max-
imum explosible concentration decreased with increasing particle size, and for the 45%
volatiles coal, it was 2400 g/m3for 50-75 pm and 1800 g/m3for 100-150 pm.
Ishihama et al. also investigated potato starch of mean particle size 50 pm and found
a very high maximum explosible concentration of about 8000 g/m3.It seems probable
that the cohesive potato starch, as opposed to the free-flowing coal dust fractions, only
dispersed partly in the rotating drum apparatus, yielding a lower real concentration of
dispersed dust than the nominal value.
Other data on maximum explosible dust concentration, from more-direct experimen-
tal determination than these rather scattered and partly contradictory results, have not
been traced. It is therefore of interest to consider the more indirect determinations by
Zehr (1959). He made the first-order assumption that the conditions for flame propaga-
tion in a dusdair mixture depends only on the mass ratio of dust to air and is independ-
ent of air pressure and mean distance between particles. He then constructed the
cylindrical combustion bomb illustrated in Figure 4.30 to determine the maximum
explosible concentration of dusts.
The central 25 cm long glass tube of about 1 cm2cross section and one end closed is
first filled completely with the dust to be tested, loosely packed. The glass tube is then
inserted into the combustion bomb and the air pressure raised to the desired level.
Because the bulk densities of loosely packed organic dusts are typically on the order of
500kg/m3and maximum explosibleconcentrationson the order of 1kg/m3,air pressures
up to the order of 500 bar were required to obtain the same dust/air mass ratio in Zehr’s
combustion tube as in a dust cloud at the maximum explosible concentration at atmo-
spheric pressure. At these high pressures, the equation of state has to be corrected for non-
ideal gas behavior. Zehr (1959) gives a detailed description of the computational