Page 334 - Dust Explosions in the Process Industries

Propagation of Flames in Dust Clouds 303

necessary to include some other important factors in addition to those considered by
Contknello (1988), in particular turbulence and aspects of entrainment and dispersion of
dust particles, as discussed in Chapter 3 (see also Section 4.4.8).

4.2.6
~ I N I ~ U ~MAXIMUM EXPLOSIBLE DUST CONCENTRATIONS
AND
See also Section 9.2.4.3 in Chapter 9.

4.2.6.,1
The Problem

The existenceof well-defined minimum and maximum explosibleconcentrations of fuel
in air is well establishedfor various gases and vapors. At the outset, it would be reasonable
to expect that such limits of explosiblefuel concentrationsalso exist for combustible dusts.
However, as shown by Makris and Lee (1988), who considered the minimum explosi-
ble concentration,there is substantialdisagreement among experimentaldata for a given
dust. For example,reported values for corn starchin air range from 8 g/m3to 400 g/m3.The
disagrelement arises from considerable differences in apparatus and interpretation of
data. Because of the extremely energetic pyrotechnical igniter used, it is not surprising
that the exceptionallylow value of 8 g/m3was determinedby Siwek (1977)using a 20 liter
spherical bomb. On the other hand, it cannot be excluded that there were some real dif-
ferences among the dusts used. Although the primary grains of corn starch have a fairly
uniform size of 10-15 pm diameter,commercial corn starch qualities often contain con-
siderable fractions of stable agglomeratesthat behave as large single particles, as shown
by Eckhoff and Mathisen (1977/1978). Furthermore, the moisture contents of the corn
starches investigated were often not reported and may have varied.
One basic problem in all experimentaldeterminationof explosibilitylimits is the def-
inition of an explosion.It has been customaryto relate this definition to either direct obser-
vation of a self-sustained flame through the dust cloud, at constant pressure, or to the
increase in pressure that results if flame propagation occurs at constant volume in a
closed vessel. If the dust concentrations are in the middle of the explosible range, the
observation of explosion is simple, irrespective of the criterion chosen. Both extensive
flame propagation and extensive pressure buildup result. Problems arise when the dust
concentrationapproaches the lower or upper explosibilitylimits, and flame propagation
and pressure rise become marginal. Because of the inherent inhomogeneityof real dust
clouds and the correspondingcomparatively poor reproducibility of repeated, apparently
identical experiments, it is necessary to choose some arbitrary criterion of a minimal
explosion, either in terms of a minimal extent of flame propagation or a minimal mag-
nitude of pressure rise. Unfortunately, there seems to be no really basic scientific crite-
rion that specifies the “right” choice.
In their analysis, Makris and Lee (1988) concluded that any meaningful criterion of a
minimum explosible dust concentrationmust be related to a distinct flame propagation in
the dust/air mixture at constant pressure. They claimed that it is not possible to decide
whether or not such flame propagation occurs in constant volume experiments,and they
therefore did not considerthat results from closed bombs had any fundamental significance.