-
Contemporary Engineering Sciences, Vol. 11, 2018, no. 48, 2357 - 2369
HIKARI Ltd, www.m-hikari.com
https://doi.org/10.12988/ces.2018.85248
Numerical Prediction of Erosion Zones
in the Aero-Mixture Channel
Amel Mešić
EPBIH Concern, Thermal Power Plant Tuzla, 21. april br.4
Tuzla, Bosnia & Herzegovina
Izudin Delić
Faculty of Mechanical Engineering, University of Tuzla
Univerzitetska br.4, Tuzla, Bosnia & Herzegovina
Nedim Ganibegović
EPBIH Concern, Thermal Power Plant Tuzla, 21. april br.4
Tuzla, Bosnia & Herzegovina
Midhat Osmić
Faculty of Mechanical Engineering, University of Tuzla
Univerzitetska br.4, Tuzla, Bosnia & Herzegovina
Sead Delalić
Faculty of Mechanical Engineering, University of Tuzla
Univerzitetska br.4, Tuzla, Bosnia & Herzegovina
Copyright © 2018 Amel Mešić et al. This article is distributed under the Creative Commons
Attribution License, which permits unrestricted use, distribution, and reproduction in any medium,
provided the original work is properly cited.
Abstract
Velocity of solid particles in aero-mixture channel has been recognized as the
most significant factor that has an influence on erosion. In this paper, numerical
study of particle erosion in aero-mixture channel was conducted due to the main
parameters that have the greatest impact. Pulverised coal boiler OB650 is supplied with an aero-mixture (pulverised coal and flue gases) through distribution channels,
-
2358 Amel Mešić et al.
where the distribution of the same aero-mixture is regulated with a tilt of
regulation dumper. These types of regulation cause the problems that have greater
impact on the internal structure of the same distribution channels. It is concluded
that the best way to determinate that problems is with CFD tool. In this case CFD
tool was used to simulate the flow field inside the channel, using Euler-
Lagrangian mathematical approach.
Keywords: CFD analysis, pulverised coal particles, erosion, aero-mixture
channels
1 Introduction
Pneumatic transport of solid particles represents significant type of multiphase
flow in industrial applications. Majority of these systems was affected with
erosion problems [1-3] and these problems are even more expressed in systems
that include more curved sections. Since the early 1990s, computational fluid
dynamics has been widely used for solid particle erosion prediction in curved
pipes and ducts [4-6], with various analytical, semi-empirical and empirical
models which have been developed. Meng and Ludema [7] provided a critical
review of some of the erosion models that has been developed and found, and
they concluded that there were 28 models that were specifically made for solid
particle–wall erosion. The authors reported that 33 parameters were used in these
models, with an average number of five parameters per model that were used for
description of erosion process and erosion mechanism. Erosion of metals occurs
when solid particles are carried by a stream of gas or liquid impinge against the
metal surface and removed microscopic particles from its surface. As a result of
such process, destruction and failure of industrial equipment have occurred.
Therefore, present study is focused on the erosion analysis of aero-mixture
channel walls. From previous papers, it can easily be seen how CFD is actually
very useful in determination and prediction of various variables of multiphase
flow, which cannot be determinate and predicted with classical methods.
The pulverized coal fired boiler unit type OB 650 was selected as the subject
of investigation. Due to the fact that pulverized coal fired boilers require
homogeneous coal mass distribution between burners and sufficiently low particle
moisture content, in order to maintain plant performance and operating efficiency,
pulverised coal boiler has to have good coal preparation system.
The preparation system was constructed from three main elements low-
emission burners, aero-mixture channels and ventilation mills (type S-36.50),
which are used for drying, milling and spraying of coal dust particles. The feed
rate of coal, amount of flue warm gasses and temperature are variable and depend
from boiler load. The raw coal is fed into ventilation mill parallel with warm flue
gasses recirculated from the outlet of the furnace. As the coal gets milled and
dried, ventilation mill supresses that aero-mixture through aero-mixture channel
and low emission burner to be used as fuel. Under normal operating conditions,
-
Numerical prediction of erosion zones 2359
the heat in the combustion zone will cause ignition of the all incoming fuel. To
achieve better understanding of distribution of the coal and gas mixture through
complicated mill-ducts, various experimental and numerical research have been
conducted (see, for example, Shah et al. [8], Dodds et al. [9], Vuthaluru et al.
[10,11] and Arakaki et al. [12]). Previous papers have shown that CFD analysis
was used for analysing flow profiles and for designs or redesigns of mill-ducts
with aim to improve the overall performance.
In this case aero-mixture distribution in upper and lower part of burner was
regulated with the tilt of the regulation damper, which allows distribution of aero-
mixture in required ratio. Regulation damper moves manually and fixes at certain
position. The whole system has 21 position. The distance between two positions is
3.14°, so the possible tilt of dumper is 33° on both sides of aero-mixture channel
from the axis (Figure 1). The entire channel, above the separator is divided by a
partition that is located exactly in the middle of the channel. All parts of the
channel are made of steel sheets of the thickness 16mm, but in the place where the
aero-mixture changes its flow, the thickness is adopted to 32mm. The boiler has
six lowemission burners and every burner is consisted of two aero-mixture jets
and three jets of secondary air. Every aero-mixture jets has thirteen tubes of core
air that are sorted like a cross. This type of construction allows deep reduction of
NOx and better regulation and distribution of air and pulverised coal. It is
important to emphasize that plant uses brown coal as a fuel, with the
characteristics given in Table 1. Coal granulation, speed, mass and volume flows
of individual phase’s data, that were used for numerical simulations, were taken
from the submitted Report of the optimization of the boiler unit.
Table 1 - Basic characteristics of used coal.
Characteristics of coal Project values Limit values
Low. heating value 15.000 [kJ/kg] 14.000-18.000 [kJ/kg]
Ash content 25 [%] 15-23.8 [%]
Moisture content 20 [%] 14-22 [%]
Figure 1 shows better the construction of air and aero-mixture channels with
the position of regulation damper. Due to the this kind of regulation, the observed
system has a problem with erosion of the internal structure, and the size of this
problem is generally best explained with the situation in Figure 2, that is taken
during the control of the aero-mixture channels in one of the boilers in Thermal
Power Plant.
-
2360 Amel Mešić et al.
Fig.1: Secondary air and aero-mixture channels with regulation damper.
Velocity of the solid particles has been recognized by researchers as the
significant factor that has influence on erosion. Experimental results have shown
that the erosion rate is exponentially proportional to the velocity of the solid
particles or the velocity of the carrier gas that surround the particles [13].
Therefore special attention in this paper is given to this parameters.
Fig. 2: The consequences of the erosive effect in the aeromixture channals.
2 Numerical modeling
Modelling of the geometry for discretization process represents very important
and first step in whole process of numerical modeling and this step isn’t simple at
all. During this process, it is very important to realize the construction from every
unimportant constituents of the construction. Of course, this step requires special
knowledge about that area, and if every important part isn’t considered properly,
as a result these different situations may appear:
₋ Too simplified geometric configuration, where excessive simplifying can undermined the validity of the results
₋ Extremely complex geometric configuration that requires extremely large computing resources for the simulation process.
-
Numerical prediction of erosion zones 2361
Of course optimal geometric configuration is the target that will allow us good
discretization process and further analysis. In this case, the whole geometry was
modeled (Figure 5) in CAD package Catia V5, and as it can be seen that the
whole geometry isn’t very demanding in terms of complexity, but the overall
dimensions of the geometry are problem. Best convergence in this case can be
achieved with tetrahedral discretisation, so entire area was discretised with
1.522.007 tetrahedral cells, in case of existing dumper and 1.522.007 tehtrahedral
cells in case of modified dumper. Important thing is that volume discretization of
domain for every simulation is done with same initial conditions and that for
reference basic value is taken 0.05 [m].
Fig. 3: 3D model, mesh and boundary condition of aero mixture channel
Better view of fluid and discretized domain with inlet and outlet regions is
shown in figure 3. The observed problematics is solved with Euler-Lagrangian
approach, in which primary phase is solved by applying the Eulerian model and
secondary or dispersed phase by applying the Lagrangian model. In this case
Eulerian equations are solved for the fluid phase (the carrier gas – in this case air)
and Newtonian equations of motions are solved for dispersed phase (pulverised
coal particles). Each particle can have different properties. Mathematical
definition of the particle size as the injected material in the continuous phase has
an important role in this process due to the fact that velocity of every reaction
depends from the surface contact, so several assumptions needs to be adopted. It
is assumed that all injected particles are correct geometrical bodies of the
spherical shapes and that particle sizes in the injector area are generated with
cumulative distribution function (CDF). In this case Rosin-Rammler's distribution
function is used for description of granuleometric characteristics of pulverised
coal [Eq.1], where the whole particle size interval is divided into the set particles
of certain dimensions.
n
d
dR exp100 (1)
Where is:
n of uniformity, for dust this value ranges between 0.4 and 1.2.
d a representative diameter is defined as the size at which 63.2% of particles
-
2362 Amel Mešić et al.
are less then representative diameter.
Representative diameter and parameter of distribution were determinated by
data taken from the sieve analysis (80μm, 200μm, 500μm) of coal dust of the mill
with seventy working hours (Table 2).
Table 2 - Calculated parameter of distribution and representative diameter
Number of working
hours of the coal mill [h] Parameter of
propagation n [/] Representative
diameter d̅ [μm]
70 1,4544 154,554
Diagram on Figure 4 was constructed from sieve analysis of the coal that was
taken after mills with different working hours.
Fig. 4: Sieve analysis for different working hours of ventilation mill [14]
From the Figure 4 it is indicative that deterioration process of striking plates in
ventilation mills have a large impact to the quality of the milling process, and so
on the calculated parameters in Table 3. Boundary conditions for this generated
domain were determinated from the experimental data, and overview of all
essential boundary conditions is given in Table 3.
Table 3 - Overview of boundary conditions
Type
magnitude
Aer
o-m
ixtu
re v
elo
city
[m/s
]
Pa
rtic
le v
elo
city
[m
/s]
Ma
ss f
low
of
pa
rtic
les[
kg
/s]
Av
era
ge
dia
met
er o
f
pa
rtic
les
[μm
]
Pa
ram
eter
dis
trib
uti
on
of
pa
rtic
les
[/]
Inle
t fl
uid
tem
per
atu
re
[oC
]
Ta
ng
enti
al
coef
fici
ent.
rest
itu
tio
n o
f th
e
pa
rtic
les
[/]
No
rmal
coef
fici
ent
of
rest
itu
tio
n p
art
icle
s [/
]
Inlet bound. 23 23 6,83 154,554 1,4544 195 / / Wall bound. / / / / / / 0,9 1
-
Numerical prediction of erosion zones 2363
3 Results and discussion
In this part of the paper, simulation results of the flow aero-mixture through the
channel are shown in the two different views. Secondary phase (solid) is shown
in the form of a 3D trajectory, describing the coal pulverised particles in the feed
channel, while the primary phase (fluid) is shown in the form of the scalar
velocity distribution in the plane of the longitudinal section channel (Fig. 5).
Fig. 5: Flow of primary and secondary phases in the channel
Vector velocity field (Figure 6) shows the impact of the regulation dumper on
the aero-mixture dispersion in the channel. The regulation dumper separates the
flow into two parts where a larger proportion of 70% is in the upper part, and less
than about 30% in the lower part of the channel. By increasing the inclination of
the regulation dumper, the flow through the channel is reduced, and the fluid flow
is concentrated along the channel walls. This increases the flow rate and the
particles are directed towards the zone along the channel wall.
Fig. 6: Distribution of the vector field velocity in the channel
-
2364 Amel Mešić et al.
Figure 7 shows particle trajectories of a smaller diameter and the particles of a
larger diameter. Different diameters represent different efficiency of mill
operation, the time after the overhaul and the time before the overhaul of the mill.
During the exploration the efficiency of coal pulveriser in the mill decreases,
Figure 2.
Fig. 7: Particle trajectories in the channel (fine granulation and coarse granulation)
It can be noticed that the flow of the particles with a smaller diameter (Dp =
103μm) has a tendency to follow the dominant flow of fluid that follow the
curvature of the channel, while the particles of a larger diameter (Dp = 310μm)
particles do not follow the dominant fluid flow and strike the channel wall. So, by
the increase in particle diameter, the concentration of particles, that have more
aggressive contact with the wall, increases. Further movement in the inclined part
of the particle takes different positions. Fine particles continue to follow the
geometry of the channel, while particles with a larger diameter due to the impact
on the wall of the channel change the direction of movement. From the aspect of
the durability of the channel construction, this effect is negative due to the angle
under which the particles strike from the walls of the channel. From the enclosed
pictures, two distinct types of erosion can be clearly seen; it is impact (acute angle
erosion) and abrasive (erosion of obtuse angle) erosion.
The size of the erosive damage depends on the influence parameters and the
mechanical characteristics of the channel wall material. Influential parameters are
primarily angle of impact particles on the wall, particle velocity, particle size as
well as the shape and density of the particles.
By analyzing the trajectories of particles in the previous figures we can deduce
where zones of erosion will appear. These are primarily zones with a pronounced
erosion speed as shown in Figure 8. The highest level of erosion is expected in the
area of the highest speed, right after the dumper on the outer wall of the channel
where his geometry changes, which is the location of the bends. Some of the most
-
Numerical prediction of erosion zones 2365
important influential parameters for the formation of erosion are the angle of the
impact, changes geometry, high speed and large diameter particles. The smaller
diameter particles are carried by fluid flow and follow the channel geometry
predominantly without touching the wall. The main ways of reducing erosive
wear are: eliminating particles with a larger diameter, changing the angle of
impact on the surface, reducing relative fluid velocity, choosing a suitable
material, and changing the surface of the material in the form of improving its
characteristics.
Fig. 8: Areas with possible erosive action Fig. 9: Modified geometry of the
of the particles regulation dumper
By changing the size and the position of the regulation dumper can be
influenced to the main parameters, what causes material erosion. Primarily, it was
considered to change the angle of impact of the particle on the surface and reduce
the relative velocity of the fluid. Geometric modification (Figure 9) is performed
under the condition that the ratio of the particle distribution to the upper and lower
channels remains as in the first case. Increasing the flow section to the upper
channel would reduce the local resistance generated by turning the manual flap, so
that it would significantly improve the current image. The most obvious influence
of the proposed modification of the control surfaces is obtained by analyzing the
velocity field of particles, with speeds greater than 30 m/s (Figure 10), where
there is an evident reduction in the zone of the increased speeds.
This increases the number of particles without the impact of the blade and
decreases the number of those that hit the channel wall. It allows keeping the
speed below the critical speed level to a much greater extent. In this way, zones
with high speeds are reduced or completely prevented.
-
2366 Amel Mešić et al.
Fig. 10: Impact of changing the geometry of the control blade to the fluid flow
Figure 11 better represents velocity values on the cross sectional axis immediately
after regulation blade, before and after modification. In the first case, the
maximum value of the speed is 43.2 m/s along the outer wall of the channel, while
after the change the maximum speed was reduced to 33.1 m/s along the inner part
of the channel.
Fig. 11: Distribution of the velocity in the cross-section of the channel before (V1)
and after the modification (V2)
The post-modification state has an area around the exterior wall that is fully
degraded at high speeds. The consequence of this is the smaller possibility of
erosion of particles on the canal wall in this area.
This velocity distribution allows the fluid to be distributed in the horizontal part of
the channel uniformly across the transverse section, primarily in the lower
channel, Figure 12. In the upper channel, the main flow in the center of the channel is formed away from the walls, which significantly reduces the possibility of abrasive erosion in the final part of the channel.
-
Numerical prediction of erosion zones 2367
Fig. 12: Distribution of fluid velocity in the cross section of the horizontal part of
the channel before and after modification
4 Conclusion
The erosion of the boiler plant is directly related to the quality of coal burning in
boilers of thermal power plants. The paper presents a CFD analysis of the
possibilitys of reducing particle erosion in the aero-mixture channel. Numerical
simulations represent a fast and cost-effective way of predicting erosion in aero-
mixture channels and could be used before their implementation. The approach
presented in this paper enables the detection of locations most exposed to the
erosive effect of coal particles as well as finding the reason for its occurrence.
Based on this, it was proposed to change the blade geometry in order to reduce the
two main factors (particle velocity and angle of impact of the particle) for erosion.
The numerical solution to the problem clearly indicated the significant
improvement of the flow conditions in aero-mixture channels. Once again,it is
proved that the numerical simulation can now be considered as a standard,
modern method for solving problems of this type and that the numerical
simulation does not represent a one-way process but, alternating and iterative. On
the one hand, the conducted numerical analysis clearly showed the extent and
complexity of the analyzed phenomena, on the other hand it has also pointed out
to the directions of future research with the aim of innovating and improving the
multiphase flow image and thus the life of the plant. Of course, the analysis of the
different granulometric and chemical composition of coal on erosion in the aero-
mixture channel is particularly emphasized.
Acknowledgements. The research was made with financial support from
Federation Bosnia and Herzegovina represented by Ministry of Education and
Science and Education of The Federation Bosnia and Herzegovina for the purpose
of implementation of the Federal Target Program “Support researches and
development on priority directions of the science for the 2017 year”.
-
2368 Amel Mešić et al.
References
[1] M. R. Sippola and W. W. Nazaroff, Particle deposition in ventilation ducts: connectors, bends and developing turbulent flow, Aerosol Science and
Technology, 39 (2005), 139–150. https://doi.org/10.1080/027868290908795
[2] L. Jianzhong, S. Xing and Y. Zhenjiang, Effects of the aspect ratio on the sedimentation of a fiber in Newtonian fluids, Journal of Aerosol Science, 34
(2003), 909–921. https://doi.org/10.1016/s0021-8502(03)00039-9
[3] M. Yu, A. J. Koivisto, K. Hämeri and M. Seipenbusch, Size dependence of the ratio of aerosol coagulation to deposition rates for indoor aerosols,
Aerosol Science and Technology, 47 (2013), 427–434.
https://doi.org/10.1080/02786826.2012.759640
[4] K. Sun and L. Lu, Particle flow behavior of distribution and deposition throughout 90° bends: analysis of influencing factors, Journal of Aerosol
Science, 65 (2013), 26–41. https://doi.org/10.1016/j.jaerosci.2013.07.002
[5] J. Lin, P. Lin and H. Chen, Research on the transport and deposition of nanoparticles in a rotating curved pipe, Physics of Fluids, 21 (2009), 122001.
https://doi.org/10.1063/1.3264110
[6] H. Hadziahmetovic, N. Hodzic, D. Kahrimanovic, E. Dzaferovic, Computational Fluid Dynamics (CFD) Based Erosion Prediction Model in
Elbows, Tehnicki Vjesnik-Technical Gazette, 21 (2014), 275-282.
[7] H. C. Meng, K. C. Ludema, Wear models and predictive equations: Their form and content, Wear, 181–183 (1995), 443–457.
https://doi.org/10.1016/0043-1648(95)90158-2
[8] K.V. Shah, R. Vuthaluru, H.B. Vuthaluru, CFD based investigations into optimization of coal pulveriser performance: effect of classifier vane settings,
Fuel Processing Technology, 90 (2009), 1135–1141.
https://doi.org/10.1016/j.fuproc.2009.05.009
[9] D. Dodds, J. Naser, J. Staples, C. Black, L. Marshall, V. Nightingale, Experimental and numerical study of the pulverised-fuel distribution in the
mill-duct system of the Loy Yang B lignite fuelled power station, Powder
Technology, 207 (2011), 257–269.
https://doi.org/10.1016/j.powtec.2010.11.007
[10] H.B. Vuthaluru, V.K. Pareek, R. Vuthaluru, Multiphase flow simulation of a simplified coal pulveriser, Fuel Processing Technology, 86 (2005), 1195–
1205. https://doi.org/10.1016/j.fuproc.2004.12.003
https://doi.org/10.1080/027868290908795https://doi.org/10.1016/s0021-8502(03)00039-9https://doi.org/10.1080/02786826.2012.759640https://doi.org/10.1016/j.jaerosci.2013.07.002https://doi.org/10.1063/1.3264110https://doi.org/10.1016/0043-1648(95)90158-2https://doi.org/10.1016/j.fuproc.2009.05.009https://doi.org/10.1016/j.powtec.2010.11.007https://doi.org/10.1016/j.fuproc.2004.12.003
-
Numerical prediction of erosion zones 2369
[11] R. Vuthaluru, O. Kruger, M. Abhishek, V.K. Pareek, H.B. Vuthaluru, Investigation of wear pattern in a complex coal pulveriser using CFD
modelling, Fuel Processing Technology, 87 (2006), 687–694.
https://doi.org/10.1016/j.fuproc.2005.10.011
[12] C. Arakaki, A. Ghaderi, A. Sæther, C. Ratnayake, G.G. Enstad, Air mass balance for mass flow rate calculation in pneumatic conveying, Powder
Technology, 202 (2010), 62–70.
https://doi.org/10.1016/j.powtec.2010.04.007
[13] M.S.A. Bradley, Pressure Loss Caused by Bends in Pneumatic Conveying Pipelines - Effects of Bend Geometry and Fittings, Powder Handling &
Processing, 2 (1990), 315-321
[14] A. Mešić, Numeričko Modeliranje Strujanja Aerosmjese U Kotlovskom Postrojenju, Diss, University of Tuzla, Tuzla, 2017.
Received: June 7, 2018; Published: June 27, 2018
https://doi.org/10.1016/j.fuproc.2005.10.011https://doi.org/10.1016/j.powtec.2010.04.007