Page 510 - Dust Explosions in the Process Industries

Assessment of lgnitability 477

A third, more-fundamental approach would be to characterize the electric discharge
ignition sensitivity of the dust in more basic terms, for example, as a function of the
distribution of spark energy in space and time, as discussed in Chapter 5. However,
in this case, the theory needed to relate the test result to real system behavior would
have to be considerably more detailed and complex, perhaps too complex to be man-
ageable at present. Furthermore, the measurements would be very demanding in
themselves.
Therefore, whenever a test method is designed to identify real, specific industrial haz-
ards, one has to ask the basic strategic question: What is the most suitable level of res-
olution and generality of experiment and associated theory? Figure 7.2 gives an
introductory overview of the various test methods to be considered in the following
sections.

I ORGINAL DUST SAMPLE 1
1
IDENTIFICATION*
r
I
PARTICLE SIZE DISTRIBUTION I 1 SAMPLE MODIFICATION
MOISTURE CONTENT
MELTING POINT
OVERALLTEST SAMPLE 1
J
i
.____..._______.._
SAMPLE SPLITTING
DEPOSITED DUST 1 DUST IN CLOUD FORM SMOLDERINGGASES 1
4 EXPLOSIBLE/NONEXPLOSIBLE? 1
4MIN. EXPLOS. DUST CONCENTR. I
d MIN. IGNITIONENERGY 1
-1 MAX. 0, CONC. FOR INERTING TEMPERATURE OF DUST
MAX. EXPLOSION PRESSURE
MAX. RATE OF PRESSURE RISE
PROPERTIESOF HYBRID TAKE PROPER PRECAUTIONS
CLOUDS (EXPLOSIBLE GAS IF SUBSTANCE IS TOXIC OR
MIXED WITH DUST) *I EXPLOSIVE
d ELECTROSTATICPROPERTIES 1

Figure 7.2 Diagram of possible tests for assessing ignitability and explosibility properties of dusts
(Slightly modified, translated version of original by Verein deutscher ingenieurn 1988).

7.
SAMPLING OF DUSTS FORTESTING

As part of a general philosophy of testing, a few words must also be said about the need
for representativedust samples. Chemistry,including moisture content and particle size
and shape distributions,have a vital influence on both ignition sensitivity and explosi-
bility. Therefore, if the dust sample tested is not representativeof the dust or powder in