Page 467 - Dust Explosions in the Process Industries

434 Dust Explosions in the Process Industries

6.1.4
THE NORWEGIAN METHOD (MODIFIED DONAT METHOD)

The method most often used in Norway and described by Eckhoff (1988) is a slightly
modified version of the method of Donat (1971), based on Hartmann bomb assessment
of (dPIdt)-. The modification consists in the use of continuous graphs,obtainedby inter-
polation and extrapolation of Donat’s tabulated data. For elongated enclosures of length-
to-equivalent-diameter ratios exceeding 4, the enclosure should be divided into the
number of fictitious subvolumes required for an LID of each subvolume to be 14. The
vent area for each subvolume is assessedindividually,and the sum of all the areas is taken
as the total area required for venting the enclosure.
After the enforcement of the European Union “Atex 1OOa” Directive (see Chapter 8)
in 2003, continued use of the traditional Norwegian venting guidelines will be harmo-
nized with common European requirements.

6.1.5
THE RADANDT SCALING LAW FORVENTED SILO EXPLOSIONS

Bartknecht (1987) indicated that equation (6.3), derived by Radandt,

PIed = (bV“)I (A -aV“) (6.3)

could be used for scaling vent areas for silos. In this equation,which was also presented
by Radandt (1989),A (m2)is the vent area, Pred(bar(g)) is the maximum explosion pres-
sure in the vented silo, V (m3)is the silo volume, and a, b, and c are empirical constants
depending on the K,, value of the dust. P,,,, is assumed to equal 0.1 bar(&, and Predmust
not exceed 2 bar(g). For Kst, =200 bar m/s, that is, the upper limit of the dust explosion
class St 1, the constants are a = 0.011, b = 0.069, and c = 0.776, based on results from
experimentsin a 20 m3 silo with direct injection of dust from a conventionalpneumatic
transport line. It is not clear, however, how the volume scaling constant was obtained.
Eckhoff (1991) investigatedequation (6.3)by comparing data from silo explosions of
twice the linear scale used by Radandt with Radandt’s data. As shown by Eckhoff (1987),
the violence of explosionsin vented large-scale silos of LID = 6 are strongly dependent
on the location of the ignition point. For this and other reasons, it appears that Radandt’s
equation (6.3) may not be entirely satisfactoryas a general scalinglaw for silo vent areas.

6.1.6
OTHER VENT SIZING METHODS

Somefurthermethods that have been suggestedfor sizing dust explosion vents are also dis-
cussedby Schofield(1984)and Lunn (1984).They includethe K-factor method investigated
by Gibson and Harris (1976), the Schwab and Othmer Nomograph, the equivalence coef-
ficient method by Maisey (1965a), and the method by Rust (1979), the latter three based on
(dPIdt),, from the 1.2liter Hartmannbomb. Areview of literature on further developments
in design of dust explosion venting systems is given in Section 9.3.7.5 in Chapter 9.