Page 209 - Dust Explosions in the Process Industries

182 Dust Explosions in the Process Industries

Sunday 29th January, 0930: During its scheduled inspection round, the safety guard team discov-
ered fragments of shattered window panes spread over the entire yard. Inspection of the roof of the
silo building revealed that the roof of an intermediate star cell had blown up, as shown in Figure 2.23,
and dense smoke was emerging from the open cell top. The height of the cell involved was about 36
m and its cross-sectional area about 20 mz.Most of the silo was empty, the pelletized wheat bran occu-
pying only the first 7 m above the cell bottom.
Sunday 29th January, rest of day: Fire brigade and other personnel were called, and the entire
plant area was cordoned off. About 2000 kg of gaseous carbon dioxide was pumped into the burning
silo from above through a long vertical pipe extending right down to the surface of the smoldering
pellets.
Monday 30th January, early morning: The discharge valve at the cell bottom was removed, and
discharge of the pellets mass, using a mobile suction unit, was started. This gave rise to increased
smoke production, and at 03.30, more carbon dioxide was loaded into the silo cell from above.
Tuesday 31st January: The discharge operation was interrupted. Carbon dioxide was emerging
through the bottom silo exit, and more was loaded into the silo at the top.
Wednesday 1st February: More carbon dioxide was loaded into the silo at the top. From 03:OO to
1250, the smoke development was enhanced by vibrations due to operation of another silo cell. The
smoke temperature just above the pellets was 96°C and just above the silo top, 45°C.
Thursday 2nd February-Wednesday 8th February: Smoke development and temperature rise was
suppressed temporarily by loading several tonnes of carbon dioxide into the silo from the top, but
there was only slow permanent progress. Temperature rise was observed in the material stored in the
four larger adjacent silo cells.
Thursday 9th February-Saturday 11th February: Holes were drilled through the silo bottom and
at intervals a total of several tonnes of nitrogen were pumped into the pellets from below, while carbon
dioxide was charged from above.
Monday 13th February-Wednesday 15th February: Some 6000 kg of carbon dioxide and 3000 kg
of Nzwas injected into the burning pellets. Temperatures in the burning and adjoining cells and con-
tents of oxygen, CO, and COz in the gas above the pellets were monitored regularly.
Monday 20th February: The smoldering combustion in the wheat bran pellets had finally been
brought to an end.
This case history illustratesthat fighting smolderingcombustion in large silo complexes
is not only a matter of quenching, or terminating, the oxidation reaction but also indeed
a matter of cooling massive bulks of poor heat conductors to a temperature level at
which the combustion process will not start again once air is readmitted to the system.

2.9
LINEN FLAX DUST EXPLOSION IN HARBIN LINEN
TEXTILE PLANT, PEOPLES REPUBLIC OF CHINA,
IN MARCH 1987

2.9.1
GENERAL OUTLINE

In the middle of the night (2:39 A.M.), on March 15, 1987, the spinning section of the
large linen textile plant in Harbin, Peoples Republic of China, was afflicted with a cat-
astrophic dust explosion. The losses were substantial. Out of the 327 women and men
working the night shift in the spinning sectionwhen the explosion occurred,58 lost their
lives and 177 were injured; 13,000 m2 of factory area was demolished.
This explosion accident has been discussed in detail by Xu Bowen (1988) and Zhu
Hailin (1988). Xu Bowen et al. (1988) reconstructed a possible course of the explosion