Page 484 - Dust Explosions in the Process Industries

Sizing of Dust Explosion Vents 45 1

1.2

1.0

- - 0.8 Figure 6.21 Maximum explosion pressures for four
-
in
L dusts in a vented 5.8 m3 filter at two vent areas, as
2 0.6 functions of (dP/dt),,, determined by the Hartmann
75 bomb:
n"
0.4 e = 0.2 mz vent area I
= 0.3 m2 vent area nitrocelhlose flame ignition
0.2 + = silicon dust flame ignition of silicon dust
P,,,( = 0.1 bar(g)
0 Comparison is made with VDI (1979 edition) pre-
0 100 200 300 400 dictions for 3.8 m3 volume (dusty section of filter)
ldP/dt) max IN HARTMANN BOMB [bar/sl (From Eckhoff, 7990).

The use of closed-bomb tests for predicting the violence of accidental dust explosions
in industrial plants was discussed by Eckhoff (1984/1985; see also Chapter 7).

6.2.6
OTI+ ER LARG E-SCALE EXPERI MENTS RELEVANT
TO INDUSTRIAL PRACTICE

Some quite early work that is still of considerable interest and practical value deserves
attention. The pioneering work of Greenwald and Wheeler (1925) on venting of coal dust
explosions in long galleries is discussed in Section 4.4.7 in Chapter 4.
A set of results from the comprehensive investigation by Brown and Hanson (1933)
on venting of dust explosions in volumes typical of the process industry were reproduced
in Figure 6.1. The paper by Brown and Hanson describes a number of interesting obser-
vations and considerations,including the effect of the location and distributionof the vents
and the influence of the size and type of ignition source.
Brown (1951) studied the venting of dust explosions in a 1.2 m diameter, 17 m long
horizontal tube with and without internal obstructions. The tube was either closed at one
end and vented at the other or vents were provided at both ends. In some experiments,
an additional vent 'was provided in the tube wall midway between the two ends. The loca-
tion of the ignition point was varied.
Brown and Wilde (1955) extended the work of Brown (1951)by investigating the per-
formance of a special hinged vent cover design on the explosion pressure development
in a 0.76 m diameter, 15 m long tube with one or more vents at the tube ends or in the
tube wall.
Pinea,u2Giltaire, and Dangreaux (1974, 1976), using geometrically similar vented
vessels off LID about 3.5 and volumes 1, 10, and 100 m3, investigated the validity of
the vent area scaling law A2=Al(V2/Vl)2'3.They concluded that this law, which implies