Page 648 - Dust Explosions in the Process Industries

Research and Development 6 75

of an intial smoldering nest (see also Section 9.2.3). Krause and Schmidt (2002) studied
self-ignition in deposits of five dusts in oxygednitrogen atmospheres containing less
oxygen than air. They found that self-ignitioncould occur even with oxygen contents as
low as 3-6 ~01%.This is in agreement with the findings of Walther shown in Figure 1.67
in Chapter 1.
Sweis (1998) and Reddy et al. (1998) studied the effect of admixed inert material on the
minimumignitiontemperaturesof oil shale and tar sand (Sweis)and cod dust (Reddyet d.1.
The hot-plate configurations used in the two investigations were different. El-Sayed and
Abdel-Latif (2000)investigatedthe criticalhot-plate temperature as well as the criticalheat
flux for ignition of layers of corn flour and a 80/20 mixture of wheat Born and corn flour.
The effect of sample diameter and height was studied. Empirical correlations of the criti-
cal temperature for ignition, time to ignition, thermal properties, and geometrical dimen-
sions of the powder sample, were given. Lebecki et al. (2002) investigatedthe influence of
the thickness of the dust layer on the minimum hot-plate temperature for ignition.
Experiments with both constanthot-plate temperatureand constantheat flux were conducted.
Tuomisaari, Baroudi, and Latva (1998) investigated suitable methods for fighting
smoldering fires in silos, both experimentally and theoretically. By means of the rele-
vant heat and mass balance equations, combined with dimensional analysis, critical
parameters were identified, including particle size and moisture content, but the theo-
retical analysis also revealed the great complexity of the problem. Some qualitative
guidelines for fighting smoldering silo fires were given. To obtain quantitative guidelines,
further studies would be required.
Based 011 practical experience,Broechann (2001) discussedthe limitationsof infrared
radiation detectors and low-concentrationCO monitoring systems for detecting “sparks”
and open or smoldering fires in the process industries.Aspects to be taken into acount
to ensure adequate performance of such systems were outlined.
Garvalho, Gurgel-Veras, and Carvalho (2002) presented the results of an experimen-
tal study of smoldering processes in wood logs, both on a laboratory scale and in pre-
scribed forest burns. The main objective was to analyze the parameters that control
initiation and sustained propagation of the smolderingprocess.
Xu, Cui, and Xu (1993) found that very weak electric spark discharges, in the range
0.1-1 .O mJ, can probably initiate smoldering combustion in layers of linen flax. Hesby
(2000) ignited Layers of combustible dusts by showers of metal sparks generated by
forcing rods of various metals against a grinding wheel. Parameters controlling the
number and size of the sparks impacting on the dust layer included the contact pressure
between metal rod and grinding wheel and the duration of the contact.A conclusionwas
that it seems highly unlikely that steel spark showers from single accidentalimpacts can
initate smoldering combustion in layers of dusts of natural organic materials.

9.3.5.3
Ignition of Dust Clouds by Hot Surfaces and Smoldering and Flaming Nests

Work up to 1990 is reviewed in Section 7.8 in Chapter 7.
Zockoll and Wiemann (1996) found that minimum ignition temperatures of dust
clouds determined in the standard Godbert-Greenwald furnace were systematically
about 100K lower than those determinedwith plane hot surfaces of areas up to 144cm2.
Therefore, current requirements for maximum permissible surface temperatures of