Modeling of Coal Spontaneous Fire in A Large-Scale Stockpile
Corresponding email: yulianto.nugroho@ui.ac.id
Published at : 04 Apr 2023
Volume : IJtech
Vol 14, No 2 (2023)
DOI : https://doi.org/10.14716/ijtech.v14i2.5367
Thabari, J.A., Auzani, A.S., Nirbito, W., Muharam, Y., Nugroho, Y.S., 2023. Modeling of Coal Spontaneous Fire in A Large-Scale Stockpile. International Journal of Technology. Volume 14(2), pp. 257-266
| Jeri At Thabari | Department of Mechanical Engineering, Faculty of Engineering, Universitas Indonesia, Depok 16424, Indonesia |
| Ahmad Syihan Auzani | Department of Mechanical Engineering, Faculty of Engineering, Universitas Indonesia, Depok 16424, Indonesia |
| Wahyu Nirbito | Department of Mechanical Engineering, Faculty of Engineering, Universitas Indonesia, Depok 16424, Indonesia |
| Yuswan Muharam | Department of Chemical Engineering, Faculty of Engineering, Universitas Indonesia, Depok 16424, Indonesia |
| Yulianto Sulistyo Nugroho | Department of Mechanical Engineering, Faculty of Engineering, Universitas Indonesia, Depok 16424, Indonesia |
The increasing need for energy consumption has
resulted in the use of energy sources in coal continuing to increase. The
transportation and distribution activities of coal also cause the pile to be
exposed to heat when it is in a pile. Due to the kinetic characteristics of
low-rank coal, the pile is very susceptible to spontaneous fire processes. Of
course, this spontaneous fire phenomenon harms the safety and economic aspects
of the coal pile. This study aims to model finite elements using Multiphysics simulation
to determine the effect of the relative humidity of the pile on the temperature
distribution of large-scale coal piles. Thus, handling methods and things that
must be considered in storing and transporting coal piles can be formulated.
Thermal phenomena modelling in coal piles is modeled using COMSOL Multiphysics
software. The simulation is carried out by varying relative humidity of the
environmental conditions (ambient). The simulation results show that this
parameter can change the level of vulnerability of the pile to burn at an
earlier time.
Coal; COMSOL; Modeling; Relative Humidity; Spontaneous Combustion
Coal spontaneous fires have received a lot of attention, especially in terms of safety. There have been catastrophic events in the coal industry due to the susceptibility of coal piles to burn themselves (Kong et al., 2019). Although much of the spontaneous combustion of coal is related to laboratory scale, in reality, this problem often occurs in large-scale industries. Recently, studies on the tendency of coal piles to experience spontaneous fires continue to concern some researchers worldwide. Since 1924, testing methods have been introduced to determine the factors that play a role in the tendency of coal to burn on its own (Wang et al., 2018; Davis and Byrne, 1924). However, the problems faced in industrial conditions are very different from what occurred on a laboratory scale; namely, coal piles have a vast dimension. Therefore, the approach or method of testing used in the laboratory must reflect physical and chemical phenomena at large scales. This statement has been discussed alongside overviews regarding typical industrial spontaneous combustion characteristics in ref (Shi et al., 2022). Due to laboratory limitations in conducting large-scale testing, researchers continue to strive to study the phenomenon of spontaneous fires through numerical studies.
The development of numerical studies related to spontaneous coal fires has entered the domain of CFDs (Taraba et al., 2014) and multiphysics simulations (Thabari et al., 2021; Saleh et al., 2017). Large-scale simulations with stockpile conditions exposed to prolonged ambient thermal exposure have been done before (Zhang et al., 2016; Taraba et al., 2014). However, these studies ignore the influence of water content, both in the air in the form of relative humidity (RH) and the moisture concentration of coal in its simulation. In reality, RH is a vital parameter determining coal's susceptibility to spontaneous combustion phenomena. Another study (Wang et al., 2018) provides that RH is vital in initiating spontaneous fires in coal piles.
This study seeks to improve previous studies by involving the influence of RH from the air in a simulated process. The simulation was conducted using COMSOL Multiphysics which proved capable of modeling spontaneous coal fire simulations on a large scale (Li et al., 2021). The model is validated by comparing the simulation results with similar references, especially temperature and flow development.
2.1. Chemical Kinetics
The process of spontaneous combustion in coal is initiated by an oxidation reaction that is influenced by temperature and other factors. Despite its complexity, previous research conducted by Carras and Young (1994) and Yuan and Smith (2008) have determined that the rate of oxidation reactions in coal piles can be simplified to
is
the amount of heat generated during the coal oxidation process. The value used
in this study is 300 kJ/g-mol O2 (Zhang et al.,
2016)2.2. Momentum Conservation
2.3. Energy Conservation
The heat transfer that occurs from forced convection in an environment with ambient temperature occurs with the following equations 12 and 13 as follows:
The heat generated from the oxidation reaction in a pile is modeled according to equation (1) and equation (3). However, equation (3) shall be modified by including porous fraction so that it becomes equation 14 as follows:
In the simulation with RH, the gas flow that enters the pile is
considered air with certain humidity.
2.4. Species Conservation
2.5. Boundary Condition
On
each side, there is heat transfer due to convection by external ambient
conditions with a convective heat transfer coefficient of 7.5 W/m2.K
(Krishnaswamy
et al., 1996). The moving wind flow is assumed to be the air
containing only O2 in a 21% mass fraction. Table 1 contains all set
parameters for the boundary conditions.
2.6. Model Geometry
Two coal pile geometries are depicted in two dimensions in this study. The geometry of the first pile refers to the following sources (Zhu et al., 2013) to validate the species transport equation (can be seen in Figure 2). The coal pile is depicted as a trapezoid with a base length of 12 m and a height of 6 m. The angle of the inclined part of the pile is 56.3°. As seen from Figure 1, the second pile geometry is chosen to facilitate the change (variation) of some test parameters. The second pile also has a similar shape, namely a trapezoid with a height of 10 m, an angle value of 45°, and the length maintained at 40 m.
Figure 1 Model
Geometry
It is
worth mentioning that the angle of pile correlates to the capacity of the
coal-handling facility. Even though it is not discussed further in this study,
this angle characteristic might be impactful because it correlates with the
drag force of the wind. The model is assumed to have an infinite size in the
z-axis direction. The simulation is carried out by taking a slice from the pile's
center. With a very long geometry in the z-axis direction, the effect of the
phenomenon in the z-direction is assumed to have no significant impact.
2.7. Model Assumption
The model in this study was developed by making assumptions in several
parts. It is done to reduce the computational load but still provide
representative results for the experiment. The assumptions are 1. The particle
size of the pile is assumed to be uniform; 2. The porosity of the pile is
assumed to be uniform for each point; 3. The spread of molecules in the stack
occurs by diffusion; 4. The heat source involved in the heat transfer process
is the convection process from the environment and the heat from the oxidation
process; 5. The phenomenon of movement of water particles (moisture) from coal
is not modeled; 6. Coal is considered a dry pile; 7. The intrinsic effect of
coal is neglected; 8. The characteristics of the simulated coal samples are
taken from (Chen et al., 2018) unless explicitly mentioned; 9. All simulations were transient
simulations with the time step of 1 day; 10. All simulations were assigned with
303 K ambient temperature.
The model is validated
using the following reference (Zhu
et al., 2013), including the input
physical and chemical parameters. This study selected a wind speed value of 0.05 m/s
and 1 m/s to match the results with the reference. This study attempts to apply
the transport of concentrated species interface to case studies. The Transport
of Concentrated Species (TCS) interface is selected based on COMSOL usage
guidelines that state that transport of diluted species is suitable for use
when the number of solvents reaches 90% (COMSOL,
2019). In Zhu's study, the
only species modelled was O2. In general, the composition of air is
21% O2 and 79% N2 by mass fractions 
Assuming that N2
is solvent, the interface is less precise to use. N2 is also
considered inert and only used as a mass train (Wu et al., 2016). It is known that the nitrogen content is still relatively stable to a
temperature of 600°C (Pels et al., 1995) and requires a temperature that can be reached up to 1200°C
to be able to react to the process of pyrolysis coal (Jiao et al., 2021) so the selection of N2 as inert in this study is still
entirely appropriate. In terms of validation, the isothermal contour
distribution of the study results is comparable to the exact parameters of
reference (Zhu et al., 2013) through Figure 2.
It is seen that the models that apply the
TCS interface can describe similar results with the reference. However, there
is a slight difference from the maximum temperature value achieved. In the
study with wind speeds of 0.05 m/s, the maximum temperature reached was 396 K.
Meanwhile, the maximum temperature recorded by this study on the same day and
conditions was 367 K. The same thing was also seen in the simulation results
with wind speed conditions of 1 m / s. The maximum temperature value recorded
from this study is 367 K. At the same time, the reference shows a value of 410
K. This difference is obtained due to the effort to use dynamic parameters in
the simulation model. The dynamic parameters used are the air's thermal
characteristics and the magnitude of the O2 flux that goes into the
pile. Thermal characteristics of air are obtained from the COMSOL Material
Library, while flux O2 is described through equation 27.
Although
discrepancies exist, it is worth noting that this study also reached a
temperature (80 – 90°C) where coal pile spontaneous combustion has occurred,
similarly like what was done in (Zhang et al., 2016).
Figure 4 Temperature distribution
(Top-Left); O2 consumption rate (Top-Right); CO2
production rate (Bottom-Left); CO-Production rate (Bottom-Right); on the 30th
day with the following configurations (1 cm particle size, 0.3 porosity, 1 m/s
velocity, 0.7 RH)
There
are two studies described the flow patterns in large-scale coal piles when they
are affected by wind (Taraba et al., 2014, Zhang et al., 2016). In both
studies, the computing domain was determined as very large to minimize the
outlet's influence of wind flow on coal piles. Based on Figure 3, the
simulation results from this study showed a flow pattern in a good resemblance
as reported in (Taraba et al., 2014, Zhang et al., 2016). Simulation
results show that the flow is experiencing circulation where air enters from
the side of the pile and goes towards the top end of the pile. In ref (Zhang
et al., 2016), this was described as
the chimney effect. From the study, it can also be observed that although a
stream of wind flows down the other side of the pile, the temperature
development on that side is not very significant. It can happen because the
flow does not provide the power to penetrate the cracks of the pile which
allows oxygen to enter and initiate the dominant oxidation process.
The flow rate tends upwards because the pressure recorded at
the top end of the pile is lower when compared to the pressure within the pile.
Therefore, O2 coming through the sides also experiences a tendency
to move towards the flow. Nevertheless, the reactivity of coal makes oxygen
consumed on its way to a low-pressure point. It is seen that the hot spot of
the pile is centered around the area where the rate of O2
consumption is very dominant. It gives the idea that the oxidation process
continues to occur in the lower third of the pile where the pressure is more
dominant, allowing O2 to enter and penetrate the cracks of the pile
and react before the oxygen makes it to the top of the pile. The region where
the highest temperatures are observed is also where the CO and CO2
production rate is seen highest.
3.3. Relative Humidity
Figure 5 Left: T-t graph on RH variation.
Right: T-t graph on velocity variation
The
simulation results
(Figure 5) show that with RH conditions of 0.5, the time required for the pile to
reach the thermal runaway (TR) phenomenon
is 33 days. Increasing the value to RH 0.7 shortens the time required for the
pile to reach thermal runaway to 25 days. The pile becomes increasingly
critical to experience thermal runaway when the RH value is at the highest
value, which is 0.9. Temperature development results were similar{Wang, 2018 #3} (Wang et
al., 2018). The
mentioned study observed that coal piles would experience an earlier
temperature increase when irrigated with air with a higher RH value. It was also explained that the heat of
rewetting occurs when extra heat is produced due to the
condensation and wetting process of the bulk material. Similar to (Wang et
al., 2018), this model
does not incorporate the
competition between
the evaporation of the pile’s moisture content and heat from the condensation process
because the pile is simulated as a dry pile. The heat (heat of rewetting) is caused by the
condensation of the water content carried by air at the pile's bottom. Alongside
the RH graph, the Temperature – time graph of the velocity effect shows that
the coal pile’s susceptibility toward spontaneous combustion increases as the
velocity (wind) increases. Future
developments can be made to simulate the competition process between moisture
evaporation and heat of rewetting in order to better analyze phenomena that may
occur in real-world conditions.
This study was done to observe how relative humidity could influence
the susceptibility of coal piles to undergo spontaneous combustion. Two coal
pile geometry were simulated, where one was done for model validation purposes
only. This study's result strengthens the fact that a higher relative humidity
value could make the coal pile more prone to spontaneous combustion. Given the result that ambient
condition is influential toward spontaneous coal combustion, adjustments to the
geographical conditions of the storage area or shipping lane to be traversed
can provide knowledge for managers in preparing appropriate plans/scenarios in
the storage or transportation process. Hence, it can increase the safety factor of the pile for the mentioned scenarios. Nevertheless, this study was done
in two-dimensional (2D) shape, where the assumption of very long geometry
toward one direction (here, it is z-axis direction). This assumption might fall
through for cases with intermediate sizes and might require three-dimensional
(3D) simulations. Therefore, future study with 3D simulations might help to
provide more understanding for coal spontaneous combustion in wider range of
large pile in storage transportation.
The authors would like to thank the financial support provided by
the Ministry of Education, Culture, Research, and Technology of the Republic of
Indonesia through Penelitian Dasar Unggulan Perguruan Tinggi (PDUPT) 2021
funding scheme under Grant No. 8/E1/KP.PTNBH/2021 dan
NKB-218/UN2.RST/HKP.05.00/2021 managed by the Directorate for Research and
Public Services (DRPM) Universitas Indonesia.
Chen, X., Li, H., Wang, Q., Zhang, Y., 2018.
Experimental investigation on the macroscopic characteristic parameters of coal
spontaneous combustion under adiabatic oxidation conditions with a mini
combustion furnace. Combustion Science
and Technology, Volume 190(6), pp. 1075–1095
COMSOL, 2019. Chemical reaction
engineering module user’s guide. In COMSOL, p. 122
Davis, J.D.,
Byrne, J.F., 1924. An adiabatic method for studying spontaneous heating
of coal1. Journal of the American
Ceramic Society, Volume 7(11), pp.
809–816
Jiao, T., Fan, H., Liu, S., Yang, S., Du, W.,
Shi, P., Yang, C., Wang, Y., Shangguan, J., 2021. A Review on nitrogen
transformation and conversion during coal pyrolysis and combustion based on
quantum chemical calculation and experimental study. Chinese Journal of Chemical Engineering, Volume 35, pp. 107–123
Kong, B., Wang, E., Lu, W., Li, Z., 2019.
Application of electromagnetic radiation detection in high-temperature
anomalous areas experiencing coalfield fires. Energy, Volume 189, p. 116144
Krishnaswamy, S., Agarwal, P.K., Gunn, R.D., 1996.
Low-temperature oxidation of coal 3. modelling spontaneous combustion in coal
stockpiles. Fuel, Volume 75(3), pp. 353–362
Li, Y., Shi, X., Zhang, Y., Xue, J., Chen, X.,
Zhang, Y., Ma, T., 2021. Numerical investigation on the gas and temperature
evolutions during the spontaneous combustion of coal in a large-scale furnace. Fuel,
Volume 287, p. 119557
Nugroho, Y.S., Rustam, R.R., Saleh, I., Saleh,
M., 2008. Effect of humidity on self-heating of a sub-bituminous coal under
adiabatic conditions. Fire Safety Science,
Volume 9, pp. 179–189
Pels, J.R., Kapteijn, F., Moulijn, J.A., Zhu,
Q.,Thomas, K.M., 1995. Evolution of nitrogen functionalities in carbonaceous
materials during pyrolysis. Carbon, Volume
33(11), pp. 1641–1653
Saleh, M., Muharam, Y., Nugroho, Y., 2017. modeling
of the crossing point temperature phenomenon in the low-temperature oxidation
of coal. International Journal of
Technology, Volume 8, pp. 104–113
Shi, X., Chen, X., Zhang, Y., Zhang, Y., Guo,
R., Zhao, T., Liu, R., 2022. Numerical simulation of coal dust self–ignition
and combustion under inclination conditions. Energy, Volume 239, p. 122227
Taraba, B., Michalec, Z., Michalcová, V.,
Blejcha?, T., Bojko, M., Kozubková, M., 2014. CFD simulations of the effect of
wind on the spontaneous heating of coal stockpiles. Fuel, Volume 118, pp. 107–112
Thabari, J.A., Supit, S., Nirbito, W.,
Muharam, Y., Nugroho, Y.S., 2021. Modeling on the effect of heat exchanger
submersion on controlling spontaneous combustion in a coal pile. Journal of Advanced Research in Fluid
Mechanics and Thermal Sciences, Volume 81(1), pp. 158–164
Wang, X., Luo, Y., Vieira, B., 2018.
Experimental technique and modeling for evaluating heat of rewetting effect on
coals' propensity of spontaneous combustion based on adiabatic oxidation
method. International Journal of Coal
Geology, Volume 187, pp. 1–10
Wen, H., Yu, Z., Deng, J., Zhai, X., 2017.
Spontaneous ignition characteristics of coal in a large-scale furnace: an
experimental and numerical investigation. Applied
Thermal Engineering, Volume 114, pp. 583–592
Wu, D., Vanierschot, M., Verplaetsen, F.,
Berghmans, J., Van den Bulck, E., 2016. Numerical study on the ignition
behavior of coal dust layers in air and O2/CO2 atmospheres. Applied Thermal Engineering, Volume
109, pp. 709–717
Yuan, L., Smith, A.C., 2008. Numerical study
on effects of coal properties on spontaneous heating in longwall gob areas. Fuel,
Volume 87(15), pp. 3409–3419
Zhang, J., Liang, Y., Ren, T., Wang, Z., Wang,
G., 2016. Transient CFD modelling of low-temperature spontaneous heating
behaviour in multiple coal stockpiles with wind forced convection. Fuel Processing Technology, Volume
149, pp. 55–74
Zhu, H.Q., Song, Z.Y., Tan, B., Hao, Y.Z., 2013. Numerical investigation
and theoretical prediction of self-ignition characteristics of coarse coal
stockpiles. Journal of Loss Prevention in
the Process Industries, Volume 26(1), pp. 236–244