See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/275885916

3D SIMULATION OF DUST EXPLOSION IN THE

INDUSTRIAL DRYING TOWER

Conference Paper June 2011

CITATIONS READS

0 77

6 authors, including:

Pawel Wawrzyniak Maciej Jaskulski

Lodz University of Technology Lodz University of Technology
32 PUBLICATIONS 118 CITATIONS 35 PUBLICATIONS 39 CITATIONS

SEE PROFILE SEE PROFILE

Ireneusz Zbicinski
Lodz University of Technology
119 PUBLICATIONS 869 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Sofware Modelling of thermal separation processes View project

ENTHALPY - "Enabling the drying process to save energy and water, realising process efficiency in the
dairy chain" View project

All content following this page was uploaded by Maciej Jaskulski on 05 May 2015.

The user has requested enhancement of the downloaded file.

 3D SIMULATION OF DUST EXPLOSION IN THE INDUSTRIAL DRYING
TOWER

P. Wawrzyniak, M. Jaskulski, M. Podyma, A. Polaczyk, J. Rabaeva,

I. Zbiciski

Faculty of Process and Environmental Engineering,

Technical University of Lodz, 213 Wolczanska Str., 93-005 Lodz, Poland
ireneusz.zbicinski@p.lodz.pl

Keywords: CFD, dust explosions, spray drying, safety, combustion

ABSTRACT

The three-dimensional simulation of dust explosion in a closed volume is still one

of the most challenging applications of CFD technique. Standard ANSYS
software was selected to model dust explosion inside the spray drying tower. A
typical approach replacing dust with a flammable gas-air mixture in the CFD
calculations was used. Two sets of CFD calculations of dust explosion in the
tower were carried out at uniform gas concentration, and variable concentration in
the dryer which reflect distribution of wet and dry powder zone in spray dryer.
The CFD calculation showed that dust explosion develops from the ignition point
mostly to the upper part of the tower. The temperature, velocity and reaction
profiles followed the same upward pattern. Results of calculations allow to
determine position and parameters of venting devices which should be installed to
minimize the risk of damages of the dryer construction.

INTRODUCTION
Dust explosion technical parameters

Explosive atmosphere is characterized by its fire or explosion parameters and qualities. These
parameters are typical of each matter, but may change because of process or explosive atmosphere
conditions. Explosion parameters are established in the laboratory according to the European
standards and they should be interpreted in normal conditions (ambient temperature 20C,
atmospheric pressure 1013 hPa, zero humidity, extremely low particle size (dust).
If the explosive atmosphere contains multicomponent dusts, the worst characteristic is always
analyzed and the mixture is considered as if there is 100% of the matter with the worst
characteristics.
Explosion limits

The Lower Explosion Limit (LEL) is the amount of dust in grams per cubic meter. Above that value
the dust-air mixture can be ignited with a strong enough ignition source. If the concentration is
below LEL, the explosion is not possible because the amount of dust particles in the mixture is too
small to spread combustion. Typical values of LEL are from tens to hundreds of grams per cubic
meter.
The Upper Explosion Limit (UEL) is a theoretical value defined as the highest dust concentration in
the mixture with air (in grams per cubic meter) where the mixture is not explosive because there is
not enough air for combustion.
Above the lower explosion limit and below the upper explosion limit there is a region of
explosibility. Roughly in the middle of this region, there is a point of optimal concentration where
energy needed to initiate the explosion is the lowest and during the explosion pressure reaches a
maximum value (the ignition energy and explosion pressure depend on the concentration).

Minimum ignition energy

The lowest energy needed to ignite dust at the optimal concentration is called minimum ignition
energy. A lower amount of energy is not able to ignite the dust. When the released energy grows,
even mixtures at non-optimal concentrations within the explosibility region can be ignited.
The values of minimum ignition energy of dusts are very diverse (from several miliJoules up to tens
of Joules). Some kinds of dust (fine plastic or dry sugar) can be ignited even by a strong
electrostatic spark and by all kinds of mechanical sparks, the other kinds of dusts (like carbon
black) are hard to ignite even by very strong ignition sources. The minimum ignition energy
depends on the concentration of the mixture.

Dust explosion index Kst

The maximum explosion pressure depends on the volume of the dust cloud, capacity of the vessel
and dust concentration. The dust explosion index reflects the speed of pressure growth in the
volume of 1 m3, filled with the dust cloud at an optimal concentration. Kst is specific for each dust
and illustrates shuttering potential of the dust cloud. The value of Kst is vital for the determination
of dust explosibility and it is considered to be a more important parameter than Pmax for the design
of all dust explosion protective systems. It is more important how fast the explosion runs than how
high the explosion pressure can be.
According to Kst, dusts are classified into four explosibility groups (Table 1.). Here are the values to
divide dusts into these groups:

Table 1. Definition of dust explosion classes

Dust explosion class Kst, bar m/s Characteristics

St 0 0 non-explosible
St 1 0 < Kst 200 week to moderately explosible
St 2 200 < Kst 300 strongly explosible
St 3 300 < Kst very strongly explosible

The dust class (or its value of Kst) for which the explosion protection system should be designed
and certified, is most essential for it evaluation and selection. The St 3 class has no upper level of
Kst (some dusts can have the values of Kst higher than 800 bar*m/s), as a result there is no
protective system certified for the whole class St 3. The protective systems suitable for the values of
Kst exceeding 300 bar m/s, must have the upper value of Kst declared in the certificate.
Dust explosion index measuring

Procedures to measure the main parameters which describe dust explosibility are given in the
European Standard EN 14034. A typical vessel for the determinations is 1 m3 in volume, but the
standard also allows the use of alternative vessels provided it can be shown that they can give
comparable results. The criteria for conformity are standard.
During the test, dust is dispersed and ignites. The highest explosion pressure developed during the
enclosed dust explosion is measured at the optimum dust concentration.
The maximum rate of pressure rise generated by the enclosed dust explosion is also measured at the
optimum dust concentration.
The peak value of the maximum rate of pressure rise (dP/dt)max is used to calculate dust explosion
index Kst (Equation 1). The value is given by:

dP
K st V1 3 (1)
dt
max
where:
(dP/dt)max maximum rate of pressure rise, [bar/s]
V total volume of the vessel, [m3]
The units for Kst are barm/s.
For the dust with higher Kst, the pressure grows faster than in the case of dust with lower Kst. Also
the dust with higher Kst usually reaches a higher maximum pressure (Pmax).
Violence of the explosion falls as moisture content of the dust increases. Eventually, the dust is no
longer explosible. The rule of thumb says that moisture content below 5% has little effect, from 5-
25% decreases sensitivity and above 25% the dust is unlikely to be explosible (Barton, 2002).

EXPERIMENTAL
Dust explosion CFD experiment

To foresee the results of explosion in the closed space the dedicated CFD software developed for
description of gas explosions in congested environments is offered (the most popular in EU is
FLACS). Also standard CFD software for flow modeling and analysis, equipped with combustion
models, can be applied. Dust particles are usually assumed to be infinitely small and dust
combustion is described by a general theory based on such parameters as the laminar burning
velocity, turbulence intensity and length scale.
Standard ANSYS software was selected to model dust explosion inside the spray drying tower. The
key step in such an approach is selection of proper gas and its concentration to model dust behavior.
Hydrogen and methane were checked. preliminary tests proved that methane better simulated dust
parameters.
To calibrate the parameters of gas mixture satisfactorily reflecting the dust behavior, CFD
experiments imitating a standard 1 m diameter sphere procedure (EU 14034) were performed.

Dust explosion 1m diameter sphere

The main characteristics of dust explosion potential are dust explosion index Kst and Pmax. For
considered powder these parameters are 50 bar m/s and 5.5 bar, respectively.
The combustion is modeled with a burning velocity model. Conservation equations are solved for
the following properties; mass, momentum, enthalpy, mixture fraction and the turbulent parameters
k and . Taking advantage of the perfect symmetry of the sphere two dimensional calculations were
performed.
Initial condition for the CFD calculation was atmospheric pressure, temperature 20C, and uniform
concentration of methane (0.052 mass). Ignition has been started in the center of the defined
geometry in the spherical volume of 10 mm diameter.
Figure 1 shows total pressure changes inside the 1m diameter sphere during the CFD experiment.
The peak value of the maximum rate of pressure rise was calculated on the basis of depicted
pressure vs. time relation (Equation 2).

dP dt max 70.8 (2)

Knowing the sphere volume V=0.7854m3 dust explosion index Kst can be calculated (Equation 3):

dP bar m
K st V1 3 57 (3)
dt max s

Figure 1. Total pressure changes during CFD calculation of explosion in 1m3 sphere.

The calculated value of explosion index Kst is close to the value for dust and checked gas mixture
can be used for simulation of dust explosion in the tower. The value of maximum pressure
calculated in the experiment is higher than Pmax for powder. The value of Kst is considered to be a
more important parameter than Pmax for the design of all dust explosion protective systems (Barton,
2002).

Dust explosion a laboratory experiment

The concentration of methane used for the CFD calculation of explosion in a sphere proved to
simulate well total pressure changes. In order to ultimately confirm accuracy of that simulation, a
laboratory test was performed in a 20 dm3 sphere according to EN 14034. The mixture of air and
methane at the concentration 0.052 mass exploded in the sphere.
The peak value of the measured maximum rate of pressure rise was (Equation 4):

dP dt max 192 bar s (4)

Knowing the sphere volume V=0.020 m3, simulated dust explosion index Kst can be calculated
(Equation 5):
 dP bar m
K st V1 3 52 (5)
dt max s

The concentration of methane used for the CFD calculation of explosion in the sphere satisfactorily
simulate the explosion of dust characterized with Kst=50 (bar m)/s.

SIMULATION
Dust explosion in the drying Tower

An extensive and comprehensive analysis of dust explosion process in a spray drying tower was
performed at this stage of the project.
There were the following parameters of dust air explosion mixture:

St = 1
m
Kst = 50 [bar ]
s
Pmax = up to 5.5 [bar]

The above parameters situates the analyzed powder in the 1st class of dust explosion materials
which means that the probability of explosion is small (Barton, 2002).
The three-dimensional simulation of dust explosion in a closed volume still represents one of the
most challenging applications of CFD technique. A typical approach to solve the problem of dust-
air explosion is to replace dust with a flammable gas-air mixture (Zimont et al, 1998).
This approach requires determination of the content and proportions of a flammable gas-air mixture
to reflect changes of pressure, temperature and velocity like for a dust-air mixture (Skjold et. al,
2006). At the first step of this part of the project, described earlier, a set of experiments was carried
out to find properties of the explosive gas-air mixture. Finally, the mixture of methane-air,
containing 0.052 mass fraction of methane, was found appropriate to simulate the behavior of the
dust and to be applied in the calculations.

The geometry

3D calculations were carried in a spray dryer shown in Figure 2. Lower part of the tower was cone-
shaped with a diameter increasing from 0.5m to 5.2m. The dryer height was 25m, diameter in the
bottom part 0.5m, in the cylindrical part 4.5m and at the top 5.3m.
Air flow field in the tower was previously found and described in the first part of the project
(Podyma et al, 2010).
According to the procedure typically applied in the calculation of gas explosions, all air inlets and
outlets were "virtually" closed (Figure 2) giving the closed system of solid walls in which the
explosion took place. This assumption does not affect the accuracy of calculations as the explosion
process is quick and air flow does not influence the propagation of pressure and temperature waves.
The above assumption reduces significantly the time of calculation and first of all makes the
calculations more stabile and easy to converge.
Additionally, as the main goal of the calculations was to predict the most extreme conditions which
can occur in the dryer, i.e. maximum pressure, maximum and average air temperature and velocity,
when the above parameters achieved maximum values, the calculations were stopped because
further continuation would not bring any valuable information useful for safety analysis of the
system performance.

Figure 2. Spray dryer geometry (left) and part of the
tower geometry after discretization (right).
At the first step of CFD calculations
a numerical mesh filling volume of the analyzed
object must be developed.
Test calculations prove that a mesh consisting
of tetrahedral elements will be the most suitable
to solve this problem (Figure 2). Finally, if dust
ignition starts in the axis of the tower,
propagation of the explosion will be
axisymmetrical which allows us to divide the
spray dryer into 4 quarters and perform
calculations in one fourth of the space of the
tower. Due to this assumption we may increase
a number of elements in the mesh which makes
the calculations more accurate in a reasonable
calculation time. During data postprocessing the
quarters were "glued" and a 3D picture of dust
explosion development was presented.
As it was mentioned before, the calculation
were carried out to determine the maximum
values of pressure, temperature and velocity.
Smooth, no slip on the walls (Equation 6) and
adiabatic (Equation 7) boundary conditions
were assumed.

U wall 0 (6)

qw 0 (7)

The point of ignition was fixed to be in the axis of a spray dryer on the level of hot air inlets, as this
is the area with the highest air temperature and low material moisture content encountered in the
dryer.

The model of dust explosion

The explosion process was calculated with the application of two combustion models: Spark
Ignition Model (SIM) and Burning Velocity Model (BVM). Ignition of the process was described
with the use of the Spark Ignition Model. (Zimont et al, 2001)

Gas mixture distribution in a drying Tower

As it follows from the experiment and theoretical calculations the methane-air mixture was most
suitable to simulate dust explosion of powder.
CFD calculations were carried out for two different methane distributions in the tower:

Spray dryer filled with uniform, constant mass fraction of methane in the gas mixture
Spray dryer filled with variable mass fraction of methane in the gas mixture

As it was described earlier, mass fraction value of the methane-air mixture obtained from
experiments which reflected explosive properties of the analyzed dust in the best way, was equal to
y=0.052.

a) b)
Figure 3. Distribution of gas mixture concentration along the spray
dryer: a) uniform methane-air gas mixture content, b) variable
methane-air gas mixture content
The first case reflects a classical
approach to solve explosion
problems and potentially describes
the most drastic damages in the
system occurring in the explosion.
We should remember, however,
that in the spray dryer we may
observe an area with reduced gas
temperature and high material
moisture content which
substantially decreases the
probability of explosion. To reflect
this case methane content
distribution in the gas mixture
along the dryer height was
assumed.
The methane content in the gas
mixture along the height of the
dryer changed from 0.052 to 0.031
was assumed arbitrarily. Figures 3.
show the distribution of methane-
air mixture in the spray dryer for
constant and variable methane-air
gas mixture content.

RESULTS
Distribution of pressure in the spray dryer

Results of CFD calculations of the maximum pressures during the explosion of detergent in the
spray dryer for constant and variable gas mixture concentration are presented in Figure 4.
For constant gas mixture concentration
in the spray dryer (y=0.052) maximum
value of pressure during process of
methane-air combustion achieved 6.3
bar after about 1.7 sec (Figure 4) and
was similar to the average pressure
which was equal to 6.1 bar.
For the case of variable methane mass
fraction, the maximum pressure
achieved 5.3 bar (Figure 4) was similar
to the average pressure which was
equal to 5.1 bar. There is a clear
difference between both cases in terms
of maximum pressure which could be
achieved during the combustion process
(6.3 bar and 5.3 bar) as well as in terms
of time for completion of the
Figure 4. Maximum pressure inside the spray dryer for two combustion process 1.7 sec and 2.5 sec
cases: constant gas mixture concentration and variable gas respectively, which is a result of
concentration. different methane content in the tower.
Like for the reaction progress after about 0.5 sec from the beginning of the combustion process to
1.4 sec, a faster increase of pressure was reported for variable gas mixture concentration than for
constant gas mixture concentration despite a lower methane content which is a result of lower
turbulent burning velocity in this time period. After 1.4 sec the pressure decreases for both
maximum and average values due to a smaller content of methane.
Analysis of the results show that in the whole dryer maximum pressure is achieved at any point due
to a quick propagation of shock wave in the system.
Calculation results confirm a quick nature of the process, pressure is practically identical in the
whole dryer for any time step.

Summary of the results

Table 2. presents the most important process parameters: reaction progress, maximum and average
temperature, maximum and average velocity, maximum and average pressure and process time.

Table 2. Final parameters of the process

Parameter Constantmethanemassconcentration Variablemethanemassconcentration
Reactionprogress 100% 100%
Maximum 88.4m/s 61.0m/s
Velocity
Average 30.5m/s 25.0m/s
Maximum 6.3bar 5.3bar
Pressure
Average 6.1bar 5.1bar
Processtime 1.7sec 2.5sec

Two sets of CFD calculations of dust explosion in the tower were carried out at constant gas
concentration and variable concentration in the dryer. The variable gas concentration reflected the
situation of reduced gas temperature and high material moisture content encountered in the upper
part of the drying chamber which substantially reduced the probability of explosion. The CFD
calculation show that dust explosion is developing from the ignition point mostly to the upper part
of the tower. The temperature, velocity and reaction profiles follow the same upward pattern. This
indicates that a venting device should be installed above the hot air inlet which generates the highest
risk of ignition.
The drying chamber above the first nozzles level is filled with moist, hardly explosible dust. The
distance above the hot air inlet to the first nozzle level seems to be the best space to install the
venting device. However, in the case of explosion, the area close to venting will be exposed to
flames and parts of the construction could be lifted by blast.
The drying tower with the length/diameter ratio ~5.5 should be considered according to EN14994
and EN14797, as an elongated enclosure. For that type of enclosure the application of two venting
devices, one close to each end, is advised.

REFERENCES
Barton J. (2002), Guide to dust explosion prevention and protection, Parts 1-3, ISBN 08295 293
7/8/9, Institution of Chemical Engineers.
Podyma M., Zbiciski I., Wawrzyniak P., Bartczak Z., Rabaeva J. (2010), Modeling of air flow in
an industrial counter current spray drying tower, Proceedings of 17th International Drying
Symposium (IDS 2010) Magdeburg, Germany, Vol. B, pp.1209-1214.
Skjold T., Arntzenb B.J., Hansena O.R., Storvika I.E., Eckhoff R.K. (2006), Simulation of dust
explosions in complex geometries with experimental input from standardized tests, Journal of
Loss Prevention in Process Industries, Vol. 19, pp. 210-217.
 Zimont V.L., Polifke W., Bettelini M. and Weisenstein W. (1998), An Efficient Computational
Model for Premixed Turbulent Combustion at High Reynolds Numbers Based on a Turbulent
Flame Speed Closure, J. Engineering for Gas Turbines and Power, Vol. 120, pp. 526-532.
Zimont V.L., Biagioli F., Syed Khawar (2001), Modelling turbulent premixed combustion in the
intermediate steady propagation regime, Progress in Computational Fluid Dynamics, Vol. 1, pp.
14-28.
EN 14034-1:2004 Determination of explosion characteristics of dust clouds.Determination of the
maximum explosion pressure Pmax of dust clouds
EN 14034-2:2006 Determination of explosion characteristics of dust clouds. Determination of the
maximum rate of explosion pressure rise (dp/dt)max of dust clouds
EN 14797:2006 Explosion venting devices
EN 14994:2007 Gas explosion venting protective systems

View publication stats