Page 59 - Dust Explosions in the Process Industries
P. 59

32  Dust Explosions in the Process Industries






















                                                   Figure 1.29  Influence of mean particle diame-
                                                   te; on minimum explosibleconcentrationfor three
                  0   20   40   60   80  100  120  140   different dusts in a 20 liter  USBM vessel (From
                      MEAN PARTICLE  DIAMETER  [pml   Hertzberg and Cashdollar, 1987).

            (1987) interpreted the data in Figure 1.29 in terms of the existence of a critical particle
            size, above which the devolatilization process becomes the critical factor in the flame
            propagation process. Below this size, devolatilization is so fast that the combustion is
            controlled by gas mixing and gas combustion only. Note that the limiting particle size
            at the minimum explosible dust concentration is not necessarily the same as at higher
            concentrations, where the explosions are more violent.
              Figure 1.30 shows how particle size influencesthe minimum ignition energy for three
            differentdusts. The vertical scale is logarithmic,and it is seen that the effect is very strong.
              Kalkert and Schecker (1979) developed a theory indicating that the MIE is proportional
            to the cube of the particle diameter, as illustrated in Figure 1.30by their theoretical pre-
            diction of the relationship for polyethylene.
              Investigations at the Chr. Michelsen Institute (CMI) showed that a 50-150  pm frac-
            tion of atomized aluminum powder could not be ignited as a cloud in air, even with a
            welding torch. This contradicts somewhat with the data in Figure 1.30.The discrepancy
            could be due to the presence of  a fine particle size fraction in the powders used by
            Bartknecht (1978). This emphasizesthe need for consideringthe entire size distribution
            rather than just a mean size.
              Figure 1.31 gives some independent experimentalresults for MIE as a function of par-
            ticle size for methyl cellulose, confirming the trends in Figure 1.30.


            1.3.3
            DEGREE OF DUST DISPERSION EFFECTIVE PARTICLE SIZE

            In his experimental studies of  burning times of pulverized fuels, Bryant (1973) found
            that persistent agglomeration was the reason for comparatively long burning times for
            apparently small particles. The situation is illustrated in Figure 1.32.A stable agglom-
            erate behaves as a large single particle of the size of the agglomerate.
   54   55   56   57   58   59   60   61   62   63   64