Page 132 - Dust Explosions in the Process Industries

Dust Explosions: An Overview 105

-

TIME, t TIME, t TIME, t
RE-IABLE SUPPRESSION CRITICAL SUPPRESSION FAILED SUPPRESSION

Figure 1.I 22 Mass of suppressant required and delivered as functions of time, for reliable sup-
pression, critical suppression, and failed suppression (From Moore 198 1, 1987).

REQUIRED

TIME. t TIME, t TIME, t
TOO LATESTART TOO LOW RATE TOO SMALL QUANTITY

Figure 1.I 23 How failed suppression can result from too late a start of suppressant injection, too
low an injection rate, and too small a quantity of suppressantinjected (From Moore, 198 I, 7987).

Automatic dust explosion suppressionhas proven feasible for organic dusts of the max-
imum rate of pressure rise in the standard 1 m3closed ISO-vessel of up to 300 bar/s (i.e.,
Ks,= 300 bar m/s, see Chapter 4). It remained uncertain, however, whether the method
could also be used for aluminum dusts of Ks, in the range 300-600 bar ds. Moore and
Cooke (1988) investigated this experimentally in a 18.5 m3 vessel, using aluminum
flakes of Ks, = 600 bar ds. A special powdered suppressant, consisting essentially of
NaHC03 (IC1DessicarbTM),proved to be the most effective for suppressing aluminum
dust explosions and was therefore used in all experiments.
However, they found that, for aluminum flakes of Ks, = 320 bar ds, even under opti-
mum conditions for suppression, it was difficult to ensure lower suppressed explosion
pressures than about 2 bar(g). In the case of dusts of natural organic materials and plas-
tics of Ks, up to 300 bar ds, the corresponding suppressed explosion pressures would
have been 0.2-0.4 bar(g).
Moore and Cooke (1988) concluded that reliable suppressionof aluminumflake explo-
sions is difficult. However, they showed that a combination of explosion suppression and
venting can reduce the maximum explosionpressure to a level significantlylower than the
level from venting only. For an aluminumflake cloud of Ks,=600 bar m/s and a staticopen-
ing pressure of the vent cover of 0.5 bar&), venting only (about 1 m2 vent area) yielded
8.2 bar(g). When combined with optimal suppression, the maximum pressure was 3.8
bar(g). However, althoughthis is considerablylower than 8.2 bar, it is still a high pressure.
Et should be mentioned that Seneca1(1989), over the range 240 to 340 bar ds inves-
tigated, found that the correlation between Ks, and reduced explosion pressure in simi-
lar suppression experiments, was rather poor.