Page 662 - Dust Explosions in the Process Industries

Research and Development 629

where the expected frequency of explosions is comparatively high. Emde and Penno
(1996) discussed further improvements of the design of Q-pipes for flame-free dust
explosion venting, allowing a significant reduction of the size of the Q-pipe, for main-
taining a given Predwith a given vent area, compared to the Q-pipe size required with
the previous design concept.
The influence of vent ducts on the maximum explosion pressure in the vented vessel
has been studied experimentally by several workers. Ural (1993) presented a theoreti-
cal model for vented gas explosions by which he was able to calculate pressure-versus-
time characteristicsin the vented vessel that agreed well with correspondingexperimental
data. It remains to be investigated,however, whether this theory can reproduce existing
experimental data for dust explosions. Tamanini (1995a, 1995b) formulated a set of
model equations of the dynamics of the processes taking part in the duct during vent-
ing. The equations rested on the assumption that inertia and friction in the duct are the
main flow controlling processes. By analyzing the equations Tamanini identified three
dimensionlessgroups of variablesthat characterizethe venting process.Finally, he fitted
his model to the experimental data of Lunn et al. (1988) (see Section 1.4.6.5in Chapter
1). Good agreement was found between theoretical predictions and the conserva-
tive envelopes of experimental data, over a range of vented-enclosure volumes from
0.2 m3to 20 m3.Griesche (1996) discussedthe special case where the vent cover is placed
at the duct exit, as opposed to the more common location at the duct entrance.
relating data from lignite dust explosionexperimentsin 1.2and 10m3vessels fitted with
ducts of various lengths and diameters, he found an empirical relationship for estimat-
ing reduced explosion pressures for cases where the vent cover is located at the vent duct
exit. Lunn et al. (1998) discussed the effects of vent ducts on the maximum pressu
generatedin vented dust explosionsin dust collectors,whereas Lunn (2001)used the U
Institution of Chemical Engineers guidance for design of vent ducts to illustrate the
increase of the reduced explosionpressure that inclusion of bends in a vent duct can pro-
duce. He also suggested a consistent method for quantifyingthis effect in the context of
the current European standard for design of venting systems.
Ponizy and Leyer (1999a, 199913) conducted an in-depth experimental study of the
flame dynamics in a vented vessel connected to a duct, using propanehr as the explo-
sive mixture. A main conclusion was that the change of the combustion regime inside
the vessel that causes the increased maximum explosion pressure experienced with ducts
is driven by an impulse generated in the initial part of the duct shortly after the flame
front has entered it.
Tamanini and ValiuLis (1998a,2000) presented a new theoretical approach for predicting
the resultant reaction impulse acting on a process structure during a vented explosion.
This applies to situations where the peak of the explosion pressure pulse cannot be
regarded as quasi-static with respect to the response of the actual vented structure.The
equations obtained by Tamanini and Valiulis were extensively validated by comparison
with data from actual vented dust explosion experiments in vessels of volumes ranging
from 0.64 m3to 95 m3.It was concluded that some previously published methods grossly
overestimate the expected resultant impulse.
Nasr and Eibl(2001) described a complete system for predicting the course as well as
the consequencesof vented dust explosions in silos, comprisingboth a numericaldust explo-
sion model and a finite-element-basedmodel for the structural response. The model pre-
dictions were supported by results from experiments in a 50 m3reinforced concrete silo.