Page 490 - Dust Explosions in the Process Industries

Sizing of Dust Explosion Vents 457

the dust explosion, or the burning rate of the dust cloud, depends on the geometry of the
enclosurein which the dust cloud burns. Turbulence and dust dispersion induced by flow
is a key mechanism for increasing the dust cloud burning rate.

6,4
INFLUENCE OF ACTUAL TURBULENCE INTENSITY
QFTHE BURNING DUST CLOUD ON THE MAXIMUM
PRESSURE IN A VENTED DUST EXPLOSION

This problem was studied specifically by Tamanini (1989), who conducted vented dust
explosionexperimentsin a 64 m3rectangularenclosure of base 4.6 m x 4.6 m and height
3.0 m. The vent was a 5.6 m2square opening in one of the four 14m2walls of the enclo-
sure. Details of the experiments were given by Tamanini and Chaffee (1989).
The dust injection system essentially was of the same type as illustrated in Figure
4.39 and discussed in Section 4.4.3.1 in Chapter 4. It consisted of four pressurized-air
containers, each of 0.33 m3capacity and 8.3 bar(g) initial pressure, connected to four
perforated dust dispersion nozzles. Two nozzle sets (i.e., eight nozzles) were mounted
on each of two opposite walls inside the chamber. The dust was placed in four canis-
ters, one for each of the pressurized-aircontainers,located in the lines between the pres-
surized containers and the dispersion nozzles. On activation of high-speed valves, the
pressurized air was released from the containers, entrained the dust, and dispersed into
a cloud in the 64 m3 chamber via the 16 nozzles. The high-speed valves were closed
again when the pressure in the pressurized containers had dropped to a preset value
of 1.4 bar(g).
As illustrated in Figures 4.40-4.42 in Chapter 4, this type of experimentgenerates tran-
sient dust clouds characterized by a comparativelyhigh turbulence intensity during the
early stages of dust dispersion and subsequent marked fall-off of the turbulence inten-
sity with increasing time from the start of the dispersion.This means that the turbulence
level of such a dust cloud at the moment of ignition can be controlled by controlling the
delay between start of dust dispersion and activation of the ignition source.
Tamanini (1989) and Tamanini and Chaffee (1989) used this effect to study the influ-
ence of the turbulenceintensity at the moment of ignition on the maximum pressure gen-
erated by explosion of a given dust at a given concentration in their 64 m3 vented
chamber. The actual turbulence intensity in the large-scale dust cloud at any given time
was measured by a bidirectional fast-response gas velocity probe, in terms of the rms
(root mean square) of the instantaneous velocity.
However, Tammini and Chaffee (1989) also found that, during the dispersion air
injection into the 64 m3 chamber, a strong mean flow accompanied the turbulent fluctu-
ations, at least in certain regions of the chamber. Furthermore, despite the injection of
the air charge through a large number of distributed points, the flow field in the cham-
ber was highly nonuniform, with the nonuniformitycontinuing during the decay part of
the transient turbulence when the discharge of the air containers was complete. However,
it was pointed out that the observed deviation of the flow field from uniformity is prob-
ably representative of the situation in actual process equipment and complicates the