Page 509 - Dust Explosions in the Process Industries

476 Dust Explosions in the Process Industries

FORMANCEOF
THEORY FOR
IGNlTABlLlWOR REALSYSTEM
EXPLOSIBILITY BEHAVIOR OF (IGNITION/
TEST DATA INDUSTRIAL NO IGNITION,
SYSTEM EXPLOSION

Figure 7.1 Overall context of ignitability and explosibility testing.

can be made about the real hazard, the test result must be passed through an adequate
theory of the industrial system and transformed to a useful statement on the behavior of
the system.
Figure 7.1 is a “philosophical” model, which becomes useful only when the contents
of the boxes are adequately specified. There are two extremes for the testing box to the
left: The first is full-scale realistic testing in true copies of industrial plants; the other is
measurements of basic behavior of particles and molecules. In the first case, there is no
need for the coupling theory, because what is measured in the left-hand box is, by defi-
nition, what happens in the box to the right. In the second case, however, a very detailed,
comprehensivetheory is required to transform the fundamental test data to real system
performance.
It could be argued that one should generally aim at testing on a fairly basic level and
develop corresponding,complex theories. However, the rational approach seems to be
to take a more balanced view. To make an optimal choice on the level of resolution, some
questions need to be answered: How good are the available measurement techniques?
How good are the theories? How much resolution is really needed for adequate design
in practice?
Consider for example the ignition of dust clouds by electric sparks. In practice, there
are many kinds of sparks, as discussed in Section 1.1.4.6 and in Chapter 5. When elec-
trically conducting wires are broken, break flashes occur and the spark energy is deter-
mined by the self-inductionof the system and the current. In other situations,the spark
arises from capacitive discharge from nongrounded, electrically conductive bodies.
Further, there are brush discharges from nonconducting surfaces, corona discharges,
propagating brush discharges, lightning discharges, and discharges from powder heaps.
So, how should one assess the electric spark ignition hazard?
The actual measurements, symbolizedby the left-hand box in Figure 7.1, can take many
forms. For example, one could construct a full-scale copy of the industrial plant, intro-
duce the powder or dust in a realistic way, and see whether ignition results. However,
as a general approach to hazard identification, this would not be very practical.
A more-realisticapproach would be to design a range of separate laboratory tests, one
exposing the dust cloud to capacitive sparks from nongrounded electrical conductors,
another to break flashes, and further special tests to other kinds of electrostatic dis-
charges. In addition, one would have to visit the industrial plant and measure the rele-
vant parameters, such as capacities, voltages, and inductivities, and estimate likely
dischargeenergy levels from theory (intermediatebox in Figure 7.1). By comparing these
theoreticalenergies with the minimum ignition energies measured in the various test appa-
ratus, one could determine whether the electric discharge ignition hazard in the plant
would be significant.