Page 675 - Dust Explosions in the Process Industries

642 Dust Explosions in the Process Industries

with special emphasis on the decade 1988-1997. The statistics showed that, on aver-
age, significant progress had been made during that decade in reducing both the annual
number of explosions and the severity of each explosion. In another paper, Schoeff
(1999b) discussed the grain dust explosion catastrophein Haysville, Kansas, in June 1998,
in which 7 men were killed and 10 injured and that caused substantialmaterial damage.
The whole event comprised a series of perhaps five successive distinct explosions.
A step-by-step account of successful extinction of a smoldering fire in a silo, using
nitrogen, was provided by Hoischen (1996). Fortunately, the extent of the fire, when
the extinction process was initiated, was limited to the part of the stored bulk material
close to its free upper surface. This may explain why extinction by applying nitrogen
was comparatively straightforward in this case (see also Section 2.8 in Chapter 2).
Hoischen also gave some general recommendations for advance inerting of silos by
applying nitrogen.
Masson (1999) discussed the cereal dust explosion catastropheat Blaye near Bordeaux
on August 20,1997. The explosion occurred when barley was transferred from a silo cell
to an open storageroom. Seven workers lost their lives. The plant, comprisinga total stor-
age capacity of about 50,000 m3,suffered extensive damage. Of the 44 silo cells, 28 were
completely demolished.The ignition source was probably either a smoldering fire in a
dust collector bin (self-heating) or a hot surface (friction heating) somewhere in the
headhouse.
Laar (1996, 1999) emphasized the necessity of basing the design of measures to pre-
vent and mitigate dust explosions on dust safety characteristicsrelevant for the actual
industrial situation. The point was illustrated by analyzingthe explosionhazard in a plant
for pneumatic transportation of a plastic granulate into a silo, which contrary to a priori
expectations, suffered a catastrophic dust explosion.
Mniszewski (1998) reviewed analytical techniques that have proven to be useful in
practical explosion accident investigation.Although the main focus of the paper is on
gas and vapor cloud explosions, much of the material covered also applies to dust explo-
sion investigation.
Broeckmann (1999) investigated an explosion case that occurred with a powder found
unable to produce dust explosions in standard tests. This special material gave rise to an
explosion in a mill via exothermal decomposition initiated by friction or impact, which
occurred even though the oxygen concentration in the atmosphere in the mill had been
reduced substantiallyby adding nitrogen.
Matsuda and Yamaguma (2000) analyzed a tantalum dust explosion that occurred in
a tantalum production plant in Japan in Spring 1997. One worker was killed and another
seriously injured. The explosion occurred while the two workers were discharging fine
tantalum dust, collected in a bag filter unit of volume about 10 m3, from the hopper at
the bottom of the filter unit. Matsuda and Yamaguma conducted a comprehensive inves-
tigation to identify possible ignition sources, and it was concluded that the explosion was
most probably initiated by an electrostatic discharge.
Alfert and Pistauer (2001) described a method used successfully to control and miti-
gate a fire in sugar residue pellets stored in a 30,000 m3silo. The novel element was the
application of a jelly layer on top of the burning pellets in the silo. In addition to reduc-
ing the convection inside the burning pellets mass, this cover also had a direct quench-
ing effect on the fire and a filtering effect on the smoke expelled from the silo into the
surroundings.However, the jelly layer did not prevent continued slow smolderinginside