Dust explosion sizing comparison

Comparison of experimental dust and gas
explosion measurements with published vent
sizing correlations
Christopher Bell

- 2 -
1 Introduction
An explosion may occur if a flammable gas or vapour, or a finely divided
combustible dust is dispersed into the atmosphere in the presence of an energy
source that has sufficient energy to cause ignition. The flame front will then travel
through the flammable gas or dust cloud. If the flammable gas or dust is present
within an enclosure, such as an item of process plant equipment i.e. a vessel, the
flame propagation will generate pressure, due to expansion of the burned fuel within
the enclosure. This may result in the catastrophic equipment failure producing an
external explosion when the pressure is released to atmosphere. Hence, it is
important that process plant that is a risk of an internal deflagration has a developed
basis of safety. The basis of safety may comprise control and/or mitigation
measures, and often equipment basis of safety is a combination of both
preventative and mitigation measures.
Examples of preventative measures are:-
• Avoidance of flammable atmospheres
• Control of ignition sources
However, an adequate basis of safety may not be achieved by the reliance of the
above preventative measures alone. Hence, the preventative measures are often
supplemented by additional mitigation measures, which include:
• Explosion containment, where the equipment design pressure, or shock
resistant strength, exceed the maximum explosion over-pressure generated
based on the fuel and the likely initial conditions;
• Explosion suppression; and
• Explosion venting.
Explosion venting is often relied on within industry as the basis of safety for process
plant equipment. The basic premise of explosion venting is to provide a vent of
sufficient area that upon opening will release unburnt gas or dust, and products of
combustion to escape from the vessel. The size of the vent should be capable of
limiting the developed explosion pressure to within the safe limits of the equipment,
such that rupture does not occur.
The sizing of the explosion vents is the subject of several industry guides and
international standards. However, the various methods are typically correlations
based on experimental data, and hence their use outside the published limits of
applicability may lead to either impractically large or conversely inadequate vent
sizes that comprise the selected equipment basis of safety.
This report aims to review some published experimental explosion data for both
gases and dusts and compare the results obtained with several of the currently
available vent sizing methods.

- 3 -
2 Explosion Vent Sizing
2.1 Gases
2.1.1 NFPA 68 1994 Edition
The 1994 edition of the NFPA 68 standard provided the following equation for the
estimation of explosion vent area for high strength enclosures (i.e. capable of
withstanding greater that 100 mbarg)
d
red
cPb
v PeaVA stat
=
Where V is the vessel volume, m3
e is the base of natural logarithm
Pstat is the vent opening pressure, barg
Pred is the maximum pressure developed during venting within the enclosure, or the
reduced explosion pressure, barg
a, b, and c are constants that are dependant on the fuels reactivity, and are shown
in the table below:
Table 1: Constants for use in explosion vent sizing [3]
a b c d
Methane 0.105 0.770 1.230 -0.823
Propane 0.148 0.703 0.942 -0.671
Hydrogen 0.279 0.680 0.755 -0.393
Coke Gas 0.150 0.695 1.380 -0.707
This equation was developed based on the explosion nomographs that were
published within the standard, with the use of the equation limited to enclosures
having a length to diameter ratio of less than 5. For fuels other than those listed in
the above table if the fundamental burning velocity is less than 60 cm/sec i.e. 1.3
times that of propane, then the propane constants are used. If the fundamental
burning velocity is greater than 60 cm/secs then the hydrogen equation is used.
However, it should be noted that this method is no longer considered appropriate as
it does not take sufficient account of the fuels reactivity, for example hydrogen is ten
times as reactive as methane yet use of the NFPA 68:1994 edition nomographs, on
which the above equation is based, will yield similar results.
2.1.2 NFPA 68: 2007 Edition
Subsequent revisions of the NFPA 68 standard used correlations based on VDI
guidelines [10] and [11], which also remains unaltered in the latest 2007 edition of
the standard.
For explosion vent sizing of enclosures having a length to diameter ratio of less
than 2, the vent size can be estimated from the equation:-
( ) ( ) ( )[ ]{ } ( ) ( )[ ] 3
2
572.03
2
582.0
10 1.0175.00567.0log127.0 VPPVPKA statredredGv −+−= −−
Where KG is the gas deflagration index, = (dP/dt)max V(1/3)

- 4 -
(dP/dt)max is the maximum rate of pressure rise obtained from standardised
experimental test equipment, bar/s
For enclosures having length to diameter ratios between 2 and 5 an additional vent
area should be added to the vent area estimated from the above equation.
750
2
2






−
=∆
D
L
KA
A
Gv
The limits of applicability for the above method are:-
KG ≤ 550 bar.m/sec
Pred ≤ 2 bar and at least 0.05 bar > Pstat
Pstat ≤ 0.5 bar
V ≤ 1000 m³
2.1.3 BS EN 14994:2006
The harmonised European norm standard EN 14994 also utilises the VDI
correlation
( ) ( ) ( )[ ]{ } ( ) ( )[ ] 3
2
5722.03
2
5817.0
10 1.01754.00567.0log1265.0 VPPVPKA statredredGv −+−= −−
However, this standard also provides an alternative simple vent sizing method,
which is based on the turbulent Bradley number:-
( ) ( )
( ) ( )
25.0
5.25.2
5.2
5.25.2
8.59.7:1if
65.5:1if
t
i
istat
i
red
i
istat
i
red
t
i
istat
i
red
i
istat
i
red
Br
P
PP
P
P
P
PP
P
P
Br
P
PP
P
P
P
PP
P
P
−=





 +
≥





 +
=





 +
<





 +
−
The turbulent Bradley number is subsequently used to solve the following equation:-
Where cui is the speed of sound at initial conditions of explosion, m/s
Ei is the expansion ratio of the combustion products
A is the vent area, m²
Pi is the initial enclosure pressure, bar
Sui is the burning velocity at the initial conditions, m/s
β is an empirical constant = 0.5 for hydrocarbons, and 0.8 for hydrogen

- 5 -
α is an empirical constant = 1.75 for hydrocarbons and 1 for hydrogen
γu is the ratio of specific heats of the unburned mixture
πv = (Pstat + Pi)/ Pi
π0 = 3.14
πi# is initial pressure expressed in bar i.e. (Pi/ 1, bar)
The quoted limits of applicability for the above simple method are:-
L/D ≤ 3
V ≤ 8000 m³
0.09 < A/V2/3
< 1.23
0 ≤ Pstat ≤ several bar
0 ≤ Pi ≤ 6 bar overpressure
2.2 Dusts
2.2.1 NFPA 68: 1994
The NFPA 68 1994 edition provided two methods for the estimation of dust
explosion vent sizes. The Radandt methodology is based on the use of
nomographs, with equations provided as an alternative. The Radandt method did
not require the use of the dust deflagration index or KST value, (KST = (dP/dt)max
V(1/3)
), but used the St grouping of the dust instead, which is a classification of the
dusts reactivity based on the KST value. The Radandt nomograph equations are:-
For St-1 dusts
Log Av = 0.77957 log V – 0.42945 log Pred – 1.24669
For St-2 dusts
For V= 1 to 10 m³
Log Av = 0.64256 log V – 0.46527 log Pred – 0.99241
For V = 10 to 1000m³
Log Av = 0.74461 log V – 0.50017 log (Pred + 0.18522) – 1.02406
In addition to the Radandt methodology NFPA provided the Simpson nomographs
and an equation developed to reproduce values obtained from their use. The
Simpson equation is:-
Av = a V2/3
KST
b
Pred
c
Where a = 0.000571 e(2 Pstat)
b = 0.978 e (-0.105 Pstat)
c = -0.687 e (0.226 Pstat)
The Radandt method will give different results to those obtained using Simpson's
correlation above. However, for all practical purposes they are sufficiently close. If
the KST value is known, then the Simpson correlation is preferable to Radandt
method.
2.2.2 VDI 3673 Part 1: 1995
The German VDI 3673:1995 [10] standard published the correlation developed by
Scholl for cubic enclosures:-

- 6 -
[ ][ ] 753.05.0
max,
569.0
max,max
5
1.027.010264.3 VPPPKPxA redstatredST
−−−
−+=
The Scholl equation is valid for:-
Vessel volumes between 0.1m³ and 10000m³
Static opening pressure, Pstat of between 0.1 and 1 barg
Maximum reduced explosion over-pressure of between 0.1 and 2 barg
Maximum explosion over-pressure, Pmax, of between 5 and 10 barg for a dust with a
deflagration index (KST) between 10 bar.m/s and 300 bar.m/s, or a Pmax of 5 to 12
barg for a KST value between 300 bar.m/s and 800 bar.m/s.
For enclosures that were elongated the VDI guideline modified the Scholl equation:-
( )( ) )/log(758.0log305.4 max, DLPAA redL +−=∆
Where the additional vent area is added to that vent area estimated for an
enclosure with an L/D ratio of below 2. The use of this equation results in a step
change in vent area for vessels with an L/D ratio of greater than 2. The above
equations were retained in the 2002 edition of the VDI 3673 guide [11].
2.2.3 NFPA 68:2002
Editions of the NFPA 68 standard after 1994 incorporated the Scholl equation from
the VDI guidelines. However, the NFPA 68 guide ceased to use the Scholl equation
in the 2002 edition, which published a vent sizing equation that removed the vent
sizing step change that was inherent in the use of the Scholl equation for elongated
vessels. The NFPA 68 2002 correlation was:-
( )( )























 −
+= −
max
max75.05
1
75.1110535.8
P
P
P
P
VKPxA
red
red
STstatv
For L/D ratios greater than 2 and less than 6 the vent area estimated by the above
equation is increased by adding the incremental vent area estimated by:-






−





−=∆ 1log
11
56.1
65.0
max D
L
PP
AA
red
v
2.2.4 NFPA 68:2007
The latest edition of the NFPA 68, which has now changed from a guide to a
standard, uses the following equation for dust explosion vent sizing:-
154.11101 max4
3
5
4
4
0, −



 += −
red
STstatv
P
P
VKPxA
For enclosures with an L/D ratio greater than 2 and less than 6 the vent area is
again increased by adding an incremental area estimated by:-
( )








−





−+= 2
75.0
01 95.0exp26.01 redvv P
D
L
AA
The limits of the above equation are:-
5 ≤ Pmax ≤ 12 bar
10 bar.m/sec≤ KST ≤ 800 bar.m/sec

- 7 -
0.1 m³ ≤ V ≤ 10000 m³
Pstat ≤ 0.75 bar
2.2.5 BS EN 14491:2006
The current harmonised European standard EN 14491 retains the use of the Scholl
equation used by the VDI 3673 guidelines [10] and [11].

- 8 -
3 Experimental data
3.1 Gases
The experimental data for gas explosions have been taken from G A Lunn –
Venting Gas and Dust Explosions – A review [1].
3.1.1 Methane
The table below summarises methane vented explosion experimental results from
Buckland, taken from Table 10 [1]
Table 2: Methane vented explosion test results [1]
Enclosure
Volume
Vent
coefficient Vent area
Vent
opening
pressure
Reduced
explosion
pressure
V K Av Pstat Pred
m3
V^(2/3)/Av m2
barg barg
26.64 8.04 1.11 0.007 0.083
26.64 8.04 1.11 0.004 0.055
26.64 2.00 4.46 0.017 0.066
26.64 2.00 4.46 0.066 0.062
26.64 2.00 4.46 0.057 0.109
26.64 2.00 4.46 0.063 0.05
26.64 4.00 2.23 0.019 0.109
26.64 4.00 2.23 0.076 0.101
26.64 4.00 2.23 0.079 0.102
26.64 2.50 3.57 0.072 0.11
26.64 2.50 3.57 0.039 0.11
26.64 5.01 1.78 0.115 0.219
26.64 5.01 1.78 0.086 0.221
26.64 4.00 2.23 0.017 0.098
26.64 4.00 2.23 0.09 0.066
26.64 4.00 2.23 0.07 0.07
26.64 4.00 2.23 0.075 0.13
3.1.2 Propane
The table below summarises propane explosion data results from Bromma, taken
from Table 8 [1]
Table 3: Propane vented explosion test results [1]
Enclosure
Volume
Vent
Vent opening
pressure
Reduced
explosion
pressure
V K Av Pstat Pred
m3
V^(2/3)/Av m2
barg barg
200 1.11 30.81 0.0549 0.0588
200 1.11 30.81 0.0294 0.0333
200 1.11 30.81 0.0098 0.0181
200 1.11 30.81 0.0289 0.0343
200 1.11 30.81 0.0549 0.0588
200 1.38 24.78 0.049 0.0637
200 1.38 24.78 0.0137 0.0299
200 1.38 24.78 0.0196 0.0295
200 1.38 24.78 0.0196 0.0348
3.1.3 Pentane
The table below summarises pentane vented explosion experimental results from
Harris and Briscoe, taken from Table 4 [1]. Note that the vent opening pressure is 0
barg.

- 10 -
3.3 Factory Mutual Standard 7-76
The Factory Mutual Global standard 7-76 includes the paper published by Tamanini
and Valiulis [2], and includes the experimental test results from a 10m³ vessel
containing powders with a deflagration index, KST, of 190 bar.m/s and 290 bar.m/s.
The vent opening pressure was 0.2 barg. The test results are detailed below:-
Table 6: FM Global vented explosion test results [2]
Enclosure
Volume Vent coefficient Vent area
Vent opening
pressure
Reduced
explosion
pressure
Deflagration
index
V K Av Pstat Pred KST
m
3
V^(2/3)/Av m
2
barg barg bar.m/s
10 7.25 0.64 0.2 0.45 190
12.21 0.38 0.2 1.4 190
16.58 0.28 0.2 2.1 190
12.21 0.38 1.86 3 190
7.25 0.64 0.2 0.75 290
12.21 0.38 0.2 2.2 290
16.58 0.28 0.2 3.6 290
7.25 0.64 1.4 1.65 290
7.25 0.64 2.5 3 290
3.3.1 Wheat dust
The table below summarises the wheat dust vented explosion results from vented
explosions within a 500m³ silo at Boge, Norway, and are taken from Figure 49 [1].
Additional data for vented wheat dust explosions within a 20m³ elongated silo (L/D
ratio of 6.25) obtained by Radandt, were taken from Figure 50 [1].
Table 7: Wheat grain dust vented explosion test results [1]
Enclosure
Volume
Vent
Reduced
explosion
pressure
V K Av Pred
m
3
V^(2/3)/Av m
2
barg
500 7.87 8 0.025
7.87 8 0.03
7.87 8 0.05
7.87 8 0.06
7.87 8 0.12
4.44 14.2 0.015
4.44 14.2 0.03
12.60 5 0.3
31.50 2 0.4
20 4.91 1.5 0.3
6.41 1.15 0.4
9.82 0.75 0.7
14.74 0.5 1.1
24.56 0.3 1.8
36.84 0.2 1.9
3.3.2 Dextrin
The table below summarises the dextrin dust vented explosion results from Donat,
Figure 40 [1].

- 11 -
Table 8: Dextrin vented explosion test results [1]
Enclosure
Volume
Vent
Reduced
explosion
pressure
V K Av Pred
m
3
V^(2/3)/Av m
2
barg
30 2.54 3.8 0.1
3.22 3 0.15
4.83 2 0.3
6.44 1.5 0.5
9.65 1 1
12.87 0.75 1.3
19.31 0.5 2
32.18 0.3 2.8
1 2.50 0.4 0.1
3.33 0.3 0.15
5.00 0.2 0.3
6.67 0.15 0.4
10.00 0.1 0.6
2.00 0.5 1.1
3.33 0.3 2

- 12 -
4 Comparison of experimental results with vent sizing correlations
A spreadsheet was used to calculate and compare the results obtained for the
reduced explosion pressures predicted by the various vent sizing correlation. An
example output from the spreadsheet is shown in Appendix A.
4.1 Gases
4.1.1 NFPA 68:1994 Calculation:-
For methane and propane, the constants used in the correlation are listed within the
standard. For Pentane, table C.1 of the NFPA 68:1994 [3] edition gives the
fundamental burning velocity of pentane as 46 cm/s, which is the same as the
burning velocity for propane, based on the NFPA 68 quoted data. Therefore, for
pentane the vent sizing calculation used the propane constants.
4.1.2 NFPA 68:2007 Calculation:-
To enable vent sizing using this standard it is necessary to know the deflagration
index, KG, the vent opening pressure, and the length to diameter ratio. The values
for the deflagration index, KG, where obtained from the NFPA 68 Standard 2007
edition table E.1 [6] and are 55, 100, and 104 bar.m/sec for methane, propane, and
pentane respectively. For the purpose of the calculation it is assumed that the vent
opening pressure is 0.1 barg and that the enclosure has a length to diameter ratio
of less than 2. It should be noted that for methane and propane the majority of the
test results were obtained using a vent panel that opened at pressures less than 0.1
barg, and the L/D ratio was not stated. For pentane, the vent panel had a negligible
opening pressure.
4.1.3 BS EN 14994:2006 Calculation:-
The alternative simple vent sizing method detailed in the above standard requires
knowledge of various thermodynamic, and combustion properties of the fuels. This
information was obtained from table A.1 of the above standard [8], and is detailed
below:-
Table 9:- Thermodynamic data and burning velocity for some fuel-air mixtures [8]
Ratio of
specific heats,
γu
Expansion
ratio of
combustion
products, Ei
Speed of
sound at initial
conditions, cui,
m/s
Fundamental
burning
velocity, Sui,
cm/s
Methane 1.39 7.52 353 43
Propane 1.37 7.98 339 45
Pentane 1.36 8.07 335 43

- 13 -
4.2 Dusts
To enable use of the dust explosion vent sizing correlations the following data was
used:-
Table 10: Explosion data for selected combustible dusts [2], [6], and [7]
Deflagration index, KST,
bar.m/sec
Maximum explosion
pressure, Pmax, barg
Aluminium 415 12.4
Cork dust 202 9.6
FM Global dust 1 190 8.5
FM Global dust 2 290 8.5
Wheat dust 112 9.3
Dextrin 106 8.8
The above data was taken from the NFPA 68: 2007 edition Tables E.1 (a) to (e) [6],
with the exception of the data from the Factory Mutual Global tests, which were
taken from Tamanini and Valiulis [2], and wheat grain dust data which is taken from
Table A.1 R K Eckhoff Dust explosions in the process industries, 2nd
Edition [7].
Data for the deflagration index and maximum explosion pressure is also available in
the G A Lunn Venting Gas and Explosions – A review [1]. However, this data was
not used as it was obtained on the Hartmann apparatus, which will not yield similar
results to explosion data obtained from the 20 litre sphere or 1m³ iso standard test
vessel. The published explosion correlations are based on data not obtained from
the Hartmann apparatus, and hence values published by Lunn [1] have not been
used.
For the purpose of this report all vent opening pressures, Pstat, were assumed to be
0.1 barg, and the length to diameter ratio is assumed to be less than 2, with the
exception of the wheat dust 20m³ enclosure comparison.
4.3 Results
The graphs below show the experimental test results for the reduced explosion
pressure within particular test equipment equipped with a defined vent area. In
addition, the graphs show the results of the various vent sizing correlations, using
information related to the enclosure, and data either obtained from referenced texts
or assumed as detailed above.

- 14 -
4.3.1 Gases
Figure 1: Methane gas explosion vent sizing results
Comparison of methane experimental data with published
vent sizing methods (vessel volume = 26.64m
3
, Pstat below 0.1
barg)
0.01
0.1
1
10
100
0.1 1 10
Vent area, Av, m2
Reducedpressure,Pred,bar
NFPA 68:2002 prEN14994 prEN14994 alt Table 10 Page 46 Buckland NFPA 68:1994
Figure 2: Propane gas explosion vent sizing results
Comparison of propane experimental data with published vent
sizing methods (vessel volume= 200m
3
, Pstat under 0.1 barg)
0.01
0.1
1
10
100
0.1 1 10 100
Vent area, Av , m2
Reducedexplosionpressure,Pred,bar
NFPA 68:2002 prEN14994 prEN14994 alt
Table 8 Page 42 Bromma NFPA 68:1994

- 15 -
Figure 3: Pentane gas explosion vent sizing results
Comparison of experimental data for Pentane with published
vent sizing methods (vessel volume 1.7m
3
)
0.01
0.1
1
10
100
0.01 0.1 1 10
Vent area, Av, m
2
NFPA 68:2002 prEN14994 prEN14994 alt Table 4 Page 33 Harris & Briscoe NFPA 68:1994
4.3.2 Dusts
Figure 4: Aluminium dust explosion vent sizing results
Comparison of experimental cork dust explosion with
published vent sizing methods (vessel volume 1.21m
3
)
0.001
0.01
0.1
1
10
100
0.01 0.1 1
Vent area, Av, m2
Simpson Radandt Scholl NFPA 68:2002 NFPA 68:2007 Exp. Data Fig 46 Lunn

- 17 -
Figure 7: FM Global vented explosion test data and comparative vent sizing
results
Comparison of FM Global test data from FM std 7-76 with published vent sizing
methods (vessel volume 10m3
, Kst=290 bar.m/s, Pstat = 0.2 barg)
0.001
0.01
0.1
1
10
100
0.01 0.1 1 10
Vent area, Av, m2
Simpson Radandt Scholl NFPA 68:2002 NFPA 68:2007 FM Global Std 7-76
Figure 8: Wheat dust vent sizing results in an elongated 20m3
silo
Comparison of wheat dust experimental data with vent sizing
methods (vessel volume 20m
3
and L/D = 6.25)
0.001
0.01
0.1
1
10
100
0.1 1 10
Vent area, Av, m2
Simpson Radandt Scholl NFPA 68:2002 NFPA 68:2007 Wheat grain dust Elongated

- 18 -
Figure 9: Wheat dust vented explosion sizing results
Comparison of wheat dust experimental data with vent sizing
methods (vessel volume 500m
3
)
0.001
0.01
0.1
1
10
100
1 10 100
Vent area, Av, m2
Simpson Radandt Scholl NFPA 68:2002 NFPA 68:2007 Wheat grain dust
Figure 10: Dextrin vented explosion and vent sizing results
Comparison of dextrin experimental data with published vent sizing
methods (vessel volume = 1m3
)
0.001
0.01
0.1
1
10
100
0.01 0.1 1
Vent area, Av, m2
Simpson Radandt Scholl
NFPA 68:2002 NFPA 68:2007 Dextrin Figure 40 Page 117

- 19 -
Figure 11: Dextrin vented explosion test and vent sizing results
Comparison of dextrin experimental data with published
vent sizing methods (vessel volume = 30m
3
)
0.001
0.01
0.1
1
10
100
0.1 1 10
Vent area, Av , m2
Reducedexplosionpressure,Pred
,bar
Simpson Radandt Scholl
NFPA 68:2002 NFPA 68:2007 Dextrin Figure 40 Page 117

- 20 -
5 Conclusions
For methane it is apparent that all the vent sizing correlations will provide sufficient
area to adequately an internal deflagration. The current NFPA 68 and European
standard utilise the same correlation and offer an improvement in accuracy when
compared to the NFPA 68 1994 methodology. The current European standard
‘simple’ calculation will significantly over-estimate the required vent area for a given
enclosure configuration. This situation is also apparent when considering the vent
sizing results obtained for the propane vented explosion data.
For pentane, the European standard simple calculation yields vent sizing results
lower than those obtained from NFPA 68 and the preferred method in EN 14994.
However, the pentane vented explosion test results are enveloped by the NFPA
and EN standard methods, whilst half of the test data points lie outside of the
published correlations limits of applicability, i.e. reduced explosion pressure greater
than 2 barg. Hence, based on the test results assessed, the current industry
standards (NFPA and BS EN 14994) will over-estimate the required vent area that
limits the reduced explosion over-pressure to within acceptable limits. Therefore,
the current gas vent sizing correlation offers a margin of safety when estimating
required vent areas.
For dust explosion vent sizing the results obtained show a greater degree of
variation when compared to actual test results, than the gas explosion venting
correlations. For aluminium, the published maximum explosion pressure, Pmax, is
outside the limits of applicability of both the Scholl and NFPA correlations. However,
it is the Scholl equation that provides better results as the degree of under-
estimation of the required vent area is less than that of the NFPA 68 equations. For
cork dust, all the correlations significantly over-estimate the required vent area.
However, the current correlations provide an improvement in vent sizing when
compared to the previous Simpson and Radandt equations.
For the dusts used in the FM Global tests, the Scholl equation correlates very well,
whereas the current NFPA standard consistently under-estimates the required vent
area. For the elongated enclosure with vented wheat dust explosions both the
elongated correlations published in the harmonised European standard and the
current NFPA 68 standard provide good agreement. For the larger volume vented
explosion of 500m3 all the correlations over-estimate the required vent area. This
over-estimate is considered to be attributable to the reduced degree of turbulence
likely to be present within the large silo, when compared to smaller enclosure
volumes. The degree of turbulence would be reflected in a reduced deflagration
index, or KST value. However, in the standard laboratory equipment (20litre sphere)
for measuring KST, there is a high degree of turbulence, and hence the test yields a
higher deflagration index than that which would be obtained if there was a reduced
degree of turbulence within the enclosure.
The results obtained for the dextrin vented explosions show poor correlation with
the test data. However, without exact explosion property data for the dextrin i.e. KST
and Pmax, or enclosure and vent information it is difficult to attribute the reasons for
the poor correlation. For this reason, it is important that when vent sizing, that the
vent is calculated using actual dust explosion test parameters. This is because the
deflagration index will vary with various factors such as particle size, and moisture
content, which may be altered by the actual processing being undertaken e.g.
attrition of particles due to pneumatic conveying. Hence, the reliance on published
explosion data for KST and other parameters is not recommended.

- 21 -
However, it is evident from the above graphs that the current NFPA 68 correlations
will yield a smaller vent size that that obtained from the Scholl equation adopted by
the EU standard. From the above graphs, it is evident that the NFPA 68 standard
correlation may not be conservative when compared to actual vented explosion
results. Hence, it is considered that the Scholl equation represents a more
appropriate correlation on which to base equipment safety.

References
1. Venting Gas and Dust Explosions – A review, G A Lunn
2. FM Global Property Loss Prevention Data Sheets – Prevention and mitigation of
combustible dust explosions and fire 7-76 May 2006
3. NFPA 68 Guide for venting of deflagrations 1994 edition
6. NFPA 68 Standard on explosion protection by deflagration venting 2007 edition
7. Dust Explosions in the Process Industries, R K Eckhoff, 2nd
Edition
8. BS EN 14994:2006 Gas explosion venting protective systems
9. BS EN 14491:2006 Dust explosion venting protective systems
10.VDI 3673 Part 1:July 1995 Pressure venting of dust explosions
11.VDI 3673 Part 1:November 2002 Pressure venting of dust explosions

Appendix A
Typical spreadsheet output for gas and dust vent sizing

- 24 -
REF REV
1 Gas
2 KG Parameter KG 55 bar.m/s
3 Maximum Pressure Pmax 7.1 barg
4 Vessel L/D ratio L/D 1.00
5 Vessel Volume V 26.64 m³
6 Vessel Surface Area As 53.52 m²
7 Initial Pressure Pinit 0 barg
8 Vent open pressure Pstat 0.1 barg
9 Reduced pressure Pred 0.35 barg OR Av 0.178396 m²
10
11
12 m² valid barg Valid
13 2.700 OK 37.23 ERR
14 2.700 ERR 42.04 ERR
15 2.685 OK 36.96 ERR
16 3.166 ERR
17 3.572 OK 4.74 OK
18 3.528 13.12
19
20
21 Limits: lower upper
22
23 KG Parameter KG 55 bar.m/s KG 0 550 bar.m/s OK
24 Maximum Pressure Pmax 7.1 barg Pmax 5 10 barg OK
25 Vessel L/D ratio L/D 1 L/D 0 2 OK
26 Vessel Volume V 26.63981 m³ V 0.1 1000 m³ OK
27 Vent open pressure Pstat 0.1 barg Pstat 0.1 0.5 barg OK
28 Pred 0.15 2 barg OK
29
30 Reduced pressure Pred 0.35 barg Av 2.700 m²
31
32 OR
33
34 Reduced pressure Pred 37.22826 barg 0.000 Av 0.179 m² ERR
35 Note:- alter Pred to obtain desired Av
36
38
41 Vessel L/D ratio L/D 1 L/D 2 5 ERR
43 Vent open pressure Pstat 0.1 barg Pstat 0.1 1 barg OK
45
47 delta A 0.000 m²
48
49 OR Av 2.700 m²
50
51 Reduced pressure Pred 42.0418 barg 0.000 Av + dA 0.179 m² ERR
53
54
55
61 Pred 0.15 0.1 barg ERR
62
63
64 C 0.035
65 Av 3.166 m² Using NFPA 68 2002 constants
66
67
68
69
70
71
72
73
74
75
82
84
85 OR
86
87 Reduced pressure Pred 36.95866 barg 0.000176 Av 0.179 m² ERR
88 Note:- alter Pred to obtain desired Av Pred adj 36.959 barg for high initial P
89
90
91
92
93
94
Methane
0.045
Gases with Su<1.3Su
propane
prEN 14994 Gas Explosion Venting Protective systems 2004 (Same method as NFPA 68 2002)
0.013
Methane 0.035
Fuel
C (bar½
) mixtures in
air only
Anhydrous NH3
Approved by
Checked by
Prepared by
Date
NFPA 68:2002 Low Strength Enclosures
ELONGATED NFPA 68 2007 Edition)
Revision A B C D E F
Fuel Characteristic Constant for Venting
Equation
pr14994:2004 Alternative
NFPA 68:1994
NFPA 68: 2007 (same as 2002, and 1998 editions)
NFPA 68:2007
NFPA 68:2007 ELONGATED
pr14994:2004
NFPA 68 Low Strength
INPUT DATA
Gas Explosion Calculation
Calculation Sheet
RESULTS
Vent Area (Av) Red Pressure
Table 6.2.2 from NFPA
68:2002
No dimesional limit to
the size. However,
panels should be
evenly spaced for
elongated enclosures
ie for L/D>3
Note that data on
which this correlation
is based gives a
maximum L/D of 2.
For high initial Pressure (up to 3
barg) adjust Pred guess until
adjusted Pred is within required
limits
Calculate

- 25 -
REF REV
1 Pred 4.735262 Limits
2 Dimensionless Reduced Pressure πred 0.35 4.735262 Lower Upper
3 Dimensionless Static Pressure πv 1.1 Av/V^(2/3) 0.400513 0.09 1.23 OK
4 Dimensionless pressure complex 0.275795 3.731318 Pinit 0 0 6 OK
5 Turbulent Bradley Number Brt 3.346454 0.26686 Pstat 0.1 0 3 OK
6 Specific heat ratio for unburned mixture kui 1.39 V 26.63981 0 8000 OK
7 Expansion ratio of combustion products Ei 7.52 L/D 1 0 3 OK
8 Speed of sound at initial conditions cui 353 m/s
9 Burning velocity at initial conditions sui 43 cm/s
10 Empirical constant α 1.75
11 Empirical constant β 0.5
12
13 Vent area Av 3.572488 0.178396 m²
14
15 RHS of Transcendental equation 0.175811 0.01402
16 LHS of transcendental equation 0.176634 0.013071
17 Error 0.0008 -0.0009
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
61 Reduced pressure Pred 0.35 barg a b c d
62 Burning velocity at initial conditions sui 43 cm/s Methane 0.105 0.77 1.23 -0.823
63 Constant a 0.105 Propane 0.148 0.703 0.942 -0.671
64 Constant b 0.77 Hydrogen 0.279 0.68 0.755 -0.393
65 Constant c 1.23 Coke gas 0.15 0.695 1.38 -0.707
66 Constant d -0.823
67
68 Av 3.528 m²
69
70
71 Reduced pressure Pred 13.11597 barg 0.000 Av 0.179 m²
72
73
74
75
76
77
78
79
80
Alternative Vent Sizing pr14994:2004 Annex A Method
Calculation Sheet
Gas Explosion Calculation
NFPA 68:1994
If burning velocity is greater than 60 cm/s i.e. greater than 1.3 x that of propane,
then hydrogen constants are used. Otherwise propane data is used.

- 26 -
REF REV
1 Combustible Dust
2 Dust Explosion Class St 2
3 Kst Parameter Kst 290 bar.m/s
4 Maximum Pressure Pmax 8.5 barg
5 Vessel L/D ratio L/D 1.00
7 Vessel Surface Area As 27.85 m²
8 Duct Length LD 0.00 m
10 Reduced pressure Pred 0.866 barg OR Av 1 m²
11
12
13
14 m² valid m² valid barg Valid
15 1.000 OK 2.537 OK 0.866 OK
16 0.475 ERR - - 0.175 ERR
17 0.659 OK 1.149 OK 0.406445 OK
18 0.659 ERR 1.149 ERR 0.406445 ERR
19 0.558 OK 0.648 OK 0.289662 OK
20 0.571 OK 0.303356 OK
21 0.659 OK 0.701 OK 0.406445 OK
22 0.898 ERR - -
23 1.287 ERR - -
24
26 Dust Explosion Class St 2 St 1 3 OK
27 Kst Parameter Kst 290 bar.m/s Kst 10 600 bar.m/s OK
30 Vessel Volume V 9.999517 m³ V 1 1000 m³ OK
33
34 Reduced pressure Pred 0.866 barg Av 1.000 m² OK
35 a 0.00085183
36 OR b 0.95767615
37 c -0.71876488
38 Vent Area Av 1 m² Pred 0.866 barg OK
39 EFFECT OF DUCT
41 Duct Length LD 0.00 m LD 0 6 m OK
42
43 Reduced pressure P'red 0.2371986 barg Av 2.537 m² OK
44 Note : Simpson equation published in NFPA 68 1994 ed section 7-1.1.1
45
48 Rate pressure rise Kst 290 bar.m/s Kst 0 300 bar.m/s OK
51 Vessel Volume V 9.999517 m³ V 1 1000 m³ OK
52 Vent open pressure Pstat 0.2 barg Pstat 0.1 0.1 barg ERR
54
55 Reduced pressure Pred 0.866 barg Av 0.475 m² OK
56 St 1 0.36282
57 OR St 2 0.47534
58
59 Vent Area Av 1 m² Pred 0.175 barg ERR
60 St 1 0.08170
61 St 2 0.17511
62 Note : Radandt equation published in NFPA 68 1994 ed section 7-2.3.1
63
72
74
75 OR
76
77 Reduced pressure Pred 0.406445 barg 0.000 Av 1.000 m² OK
79 EFFECT OF DUCT
82
83 Reduced pressure P'red 0.3161994 barg Av 1.149 m²
84 Note: Scholl equation published in NFPA 68 1998 ed section 7-2.2
85
86
87
88
89
FM Global dust 2
NFPA 68:2007
Approved by
Checked by
Prepared by
NFPA 68:2002
NFPA 68:1998 Low Strength
Date
BS EN 14491:2006
Modified Swift eqn
SIMPSON (VDI 3673:1979 & NFPA 68 1994 Section 7-1.1.1)
RADANDT (NFPA 68:1988)
SCHOLL (VDI 3673:1995 & NFPA 68 1998 Edition)
Simpson (VDI 3673:1979)
Radandt (NFPA 68:1988)
Scholl (VDI 3673:1995)
Elongated (VDI 3673:1995)
RESULTS
Red Pressure
Vent Area (Av) with
duct
Vent Area (Av)
INPUT DATA
Dust Explosion Calculation
Calculation Sheet
Pmax upper limit is
11 bara for St1 and 2
13 bara for St 3
ref Dust Explosion Prevention and
Protection Part 1 page 73
This is a correlation to the Radandt
nomographs, which are dependent
only on the St group, and not the Kst
parameter. This method will give
different results to those obtained
using Simpson's correlation above.
However, for all practical purposes
they are sufficiently close. If the Kst
value is known, then Simpson is
preferable to Radandt.
Straight duct of
maximum length 6m
Equations are derived from the
Figure 5-4(b) in NFPA 68 1994
page 68-18
Upper limit assumed to be 6m based
on subsequent NFPA 68 issues
Calculate Pred

- 27 -
REF REV
1
6 Vessel L/D ratio L/D 1 L/D 2 6 ERR
9 Pred 0.1 1.5 barg OK
10
12 delta A 0.000 m²
13
14 OR Av 0.659 m²
15
16 Reduced pressure Pred 0.406445 barg 0.000 Av + dA 1.000 m² OK
18
19 EFFECT OF DUCT
22
24 delta A 0.000 m²
25
26 OR Av 1.149 m²
27 Note: Elongated Scholl equation published in NFPA 68 1998 ed section 7-2.3, for homgenous dust clouds only
28
37
39 delta A 0.000 m²
40
41 OR Av 0.558 m²
42
43 Reduced pressure Pred 0.2896624 barg 0.000 Av + dA 1.000 m² OK
45 EFFECT OF DUCT
48
50 delta A 0.000 m²
51
52 OR Av 0.648 m²
53
54
61 Vent open pressure Pstat 0.2 barg Pstat 0 0.75 barg OK
62
64 delta A 0.000 m²
65
66 OR Av 0.571 m²
67
68 Reduced pressure Pred 0.3033561 barg 0.000 Av + dA 1.000 m²
69
70 EFFECT OF DUCT
71 1 Limits: lower upper
72 Duct Length LD 0.00 m LD 0 6 m E1
73 E2
75 delta A #NUM! m²
76
77 OR Av 0.648 m²
78
79
88
90 OR
91
92 Reduced pressure Pred 0.406445 barg 0.000176 Av 1.000 m²
93
94 EFFECT OF DUCT
95 Max duct length that needs to be considered Ls 5.0020351 m
96 1 0.06489196
97 Pred max with duct and an L/D ratio = 1 P'red max 1.24 barg Av for L/D=1 0.087 m²
98 Pred max without duct and an L/D ratio = 1 Pred max 0.7727633 barg
99 Pred max for a L/D ratio = 6 P'red max 0.866 barg Av for L/D=6 1.227 m²
100 Pred max 0.8438026 barg
101
102 Pred max for a L/D ratio between 1 and 6 Pred max 0.7727633 barg Av for 1 ≤L/D ≤ 6 0.701 m²
103
104
105
106
107
NFPA 68:2007
ELONGATED (VDI 3673:1995 & NFPA 68 1998 Edition)
BS EN 14491:2006
Checked by
NFPA 68:2002
Approved by
Prepared by
D E F
Date
Revision A B C
Calculation Sheet
This equation is sensitive to Pred.
For low values of Pred the additional
area is relatively large.
For Pred values of 1.5 bar and above
the dAv equation should not be used,
and only use the eqn for Av.
This equation is sensitive to Pred.
For low values of Pred the additional
area is relatively large.
For Pred values of 1.5 bar and above
the dAv equation should not be used,
and only use the eqn for Av.
Vent pipes with a length of L>Ls have no
additional effect upon the pressure
increase, as flow reaches sonic velocity
NOT VALID FOR METAL DUSTS
Pmax upper limit is
10 barg for St1 and 2
12 barg for St 3
ref WinmVent Handbook April 2001
page 42
Calculate for L/D =1

- 28 -
REF REV
1
9 Pred 0.1 0.2 barg ERR
10
11
12 Av 0.898 m² Using NFPA 68 1998 constants
13
14 C 0.043 bar½
15 Av 1.287 m² Using Lunn data for C
16
17
18
19
20 C (psi½
) C (bar½
)
21
22 10 0.005 0.001
23 20 0.01 0.003
24 30 0.015 0.004
25 40 0.021 0.006
26 50 0.027 0.007
27 75 0.041 0.011
28 100 0.055 0.014
29 150 0.084 0.022
30 200 0.105 0.028
31 250 0.127 0.033
32 300 0.163 0.043
33 400 0.21 0.055
34 500 0.248 0.065
35 600 0.3 0.079
36
37
38
39 Limits: lower upper For flame length equations
46 Reduced explosion pressure Pred 0.406445 Pred 0.1 2 barg OK
47 Vent area Av 1 m²
48 Method of evaluating Pred Elongated
49 Pressure venting orientation Horizontal
50 For Horizontal pressure venting Limits: lower upper For VDI 3673 Pmax,a and Pr (at distance) equations
51 Flame Length LF 21.544 m St 1 1 ERR
52 For Vertical pressure venting Kst 0 200 bar.m/s ERR
53 Flame Length LF 17.235 m Pmax 0 9 barg OK
54 Flame Width WF 6.034 m L/D 0 2 OK
55 Maximum external peak overpressure Pmax,a 0.123 barg V 0 250 m³ OK
56 Distance to peak external overpressure RS 5.386 m Pstat 0 0.1 barg ERR
57 Distance to peak external overpressure RS 4.309 m Pred 0.1 1 barg OK
58
59 barg psig barg psig Barg psig Limits: lower upper For EU CREDIT FORMULAS
60 5.386 0.123 1.784 4.3088 0.123 1.784 1 0.690243 10.011 St 1 1 ERR
61 6 0.105 1.518 5 0.106 1.538 2 0.345121 5.006 Kst 10 200 bar.m/s ERR
62 7 0.083 1.204 6 0.088 1.282 3 0.230081 3.337 Pmax 5 10 barg OK
63 8 0.068 0.986 7 0.076 1.098 4 0.172561 2.503 L/D 0 6 OK
64 9 0.057 0.826 8 0.066 0.961 5 0.138049 2.002 V 0 1000 m³ OK
65 10 0.049 0.705 9 0.059 0.854 6 0.11504 1.669 Pstat 0.1 0.2 barg OK
66 11 0.042 0.611 10 0.053 0.769 8 0.08628 1.251 Pred 0.1 2 barg OK
67 12 0.037 0.537 11 0.048 0.699 10 0.069024 1.001
68 13 0.033 0.476 12 0.044 0.641 12 0.05752 0.834 Distance to struc./obstacle 15.36 m
69 14 0.029 0.426 13 0.041 0.591 14 0.049303 0.715 Maximum pressure at robs 0.069028 barg
70 15 0.026 0.384 14 0.038 0.549 16 0.04314 0.626 Lateral flame spread 6.378725 m
71 16 0.024 0.349 15 0.035 0.513 18 0.038347 0.556
72 17 0.022 0.318 16 0.033 0.481 20 0.034512 0.501
73 18 0.020 0.292 17 0.031 0.452 22 0.031375 0.455
74 19 0.019 0.269 18 0.029 0.427 24 0.02876 0.417
75 21.544 0.015 0.223 21.544 0.025 0.357 26 0.026548 0.385
76
77
78
79
80
81
Distance,
m
Distance,
m
Distance,
m
Pressure HattwigPressure EU CREDITPressure VDI 2002
mixtures in air only
Equation
Equation
Fuel
St-2 dusts
St-3 dusts
NFPA 68:1998 LOW STRENGTH ENCLOSURE (SWIFT EQUATION)
Calculation Sheet
Checked by
FLAME PROPAGATION - VDI 3673 Part 1 2002
Approved by
Prepared by
Date
0.026
0.03
0.051
Fuel
C (bar½
) mixtures in
air only
St-1 dusts
Table 4 from Venting Gas
and Dust Explosions 2nd
Edition GA Lunn
Table 4-3.1 from NFPA
68:1998
Taken from the EU CREDIT project. The equation
is only valid if Kst<= 200 bar.m/s
EU Credit report formula:-
For venting directed vertically
Rs = 0.25 LF ,
For venting directed horizontally
Rs = 0.2 LF
Hattwig method uses equation:-
Pblast = Pred C1 C2 / r
log C1 = -0.26/Av + 0.49
VDI and CREDIT eqns
Psmax = 0.2 Predmax A0.1
V0.18
HATTWIG FOR COMPARISON ONLY
Venting towards an obstruction
Method uses EU CREDIT project equations only and not VDI for
estimation of distance to peak external overpressure, Rs.
Pr,obs = 2 (Rs / robs) Ps,max
For CURRENT VDI Guidelines
2002 use Scholl ONLY
VDI 3673 Part 1 2002
Max external peak
pressure occurs at a
distance,
Rs = 0.25 LF

Dust explosion sizing comparison

More Related Content

What's hot (20)

Similar to Dust explosion sizing comparison (17)

Dust explosion sizing comparison