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

net/publication/335401388

DUST REACTION MODEL IN WASTE HEAT BOILER OF COPPER SMELTING

Conference Paper · August 2019

CITATIONS READS
0 118

5 authors, including:

Seung Hwan Ahn Jongshin Chang

LS-Nikko Copper Kyungpook National University
4 PUBLICATIONS   13 CITATIONS    2 PUBLICATIONS   0 CITATIONS   

SEE PROFILE SEE PROFILE

Hosang Sohn
Kyungpook National University
65 PUBLICATIONS   306 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Seung Hwan Ahn on 26 August 2019.

The user has requested enhancement of the downloaded file.

 DUST REACTION MODEL IN WASTE HEAT BOILER OF COPPER SMELTING

*Seung Hwan Ahn, Moosang Lee, and Jongshin Chang

LS-Nikko Copper Inc. Extractive Metallurgy R&D Laboratory
148 Sanam-ro, Onsan-eup, Ulju-gun, Ulsan, 44997, Republic of Korea
(*Corresponding author: [email protected])

So-Yeong Lee and Ho-Sang Sohn

Kyungpook National University, School of Materials Science and Engineering
80, Daehak-ro, Buk-gu, Daegu, 41566, Republic of Korea

ABSTRACT

It is necessary to understand the characteristics, phase changes and chemical reactions of flue
dusts introduced from a furnace to an off-gas line in order to achieve the stable operation and the
improvement. This study is to develop the dust reaction model for chemical reactions of dusts in an uptake
shaft of smelting furnace and a waste heat boiler. Phase equilibria and reaction paths in this system were
determined by thermodynamic calculation with experimental and industrial data. The oxidation reaction
rate of dusts was determined by the gas film mass transfer controlling. And the sulfation reaction rate was
evaluated by the unreacted shrinking core model. Kinetic parameters were estimated by reference data
including industrial ones. The present study with the developed model shows various effects of
temperatures, dust particle sizes, and O2 gas partial pressures on dust reactions.

KEYWORDS

Dust, Reaction model, Unreacted shrinking core model, Waste heat boiler

INTRODUCTION

The flue dusts from a copper smelting furnace move along a furnace off-gas stream into an uptake
shaft and a waste heat boiler (WHB) and electrostatic precipitator(s). It is necessary to control the dust
quality in order to stabilize the operation because the sticking behaviors differ depending on the phase of
dusts. Phase changes of dusts are caused by chemical reactions at given conditions such as temperature and
partial pressure of gases. Therefore, it is required to understand the phase changes and chemical reactions
of dusts in order to perform the improvement and stable operation. The present study carried out the
investigation to understand their fundamental knowledge. Especially, this study was focused on the
kinetics of chemical reactions of flue dusts, developed the model. In particular, the scope of development
and utilization of the model was defined as the uptake shaft of the Mitsubishi S furnace and the S-WHB
radiation section in the Onsan smelter. Information needed for the modeling work referred to data from
these processes.

MODELING

Reaction paths of flue dusts

The temperature of the dusts decreases as it moves to the WHB. The reaction paths of flue dusts
were reviewed through thermodynamic equilibrium calculations with relevant temperatures using the
standard Gibbs free energy of formation of various reactions by Pankratz (1982, 1987, 1994) and Barin
(1995). The initial phases and compositions referred to the information of the Mitsubishi S-furnace dusts
which consist of mainly splashed matte (sulfides of Cu and Fe). Conditions such as temperatures and gas
partial pressures were applied based on the uptake shaft of the S-furnace and the S-WHB radiation section.
Potential diagrams for dusts with given conditions are shown in Figure 1. It was assumed that the
compounds of Cu and Fe such as CuFeO2 are not formed in this system.

Figure 1. Potential diagrams for Cu-S-O, Fe-S-O systems (ignore the compounds of Cu and Fe)

Table 1. Theoretical reaction paths by the potential diagrams

Temp.℃ Reaction paths
FeS → Fe3O4 → Fe2O3
1300 ℃
Cu2S → Cu → Cu2O
FeS → Fe3O4 → Fe2O3
1000 ℃
Cu2S → Cu → Cu2O → CuO
FeS → Fe3O4 → Fe2O3 → Fe2(SO4)3
700 ℃
Cu2S → Cu → CuO•CuSO4 → CuSO4

Theoretically estimated reaction paths shown in Table 1 are complicated because of the formation
of intermediate products. The reaction paths of the dusts should be relatively simple in order to reduce
complexity of the model calculation. Therefore, this study defined the simplified reaction paths based on
the equilibrium calculations and the industrial investigations of dust in the WHB radiation section.
According to these observations, Cu and Fe exist as sulfate (mainly CuSO4) and oxide (mainly Fe2O3 and
Fe3O4) respectively. Figure 2 shows proposed reaction paths in the present study. The sequence of
chemical reactions was assumed that equation (1), (2) and (3) occurs sequentially. And the simultaneous
reactions of oxidation and sulfation were ignored because of complexity of the model.

FeS + 7/4O2 = 1/2Fe2O3 + SO2 (1)

Cu2S + 3/2O2 = Cu2O + SO2 (2)

Cu2O + 2SO2 + 3/2O2 = 2CuSO4 (3)

Figure 2. Proposed reaction paths of flue dusts in this study

Reaction rate equations of flue dusts

Oxidation reactions

The initial phase of flue dusts from the S-furnace is molten sulfide because the temperatures in the
furnace and the WHB inlet are higher than the melting point of dusts. Sohn (1993) noted that overall rate of
oxidation reaction of molten sulfide is controlled by the gas film mass transfer because reactivity of molten
sulfide is usually extremely high. In other words, the oxidation rate of molten sulfide is determined by the
gas film mass transfer. Therefore, the gas film mass transfer controlling was applied to oxidation reaction
rate of dusts in this study. The O2 gas consumption rate JO2 of oxidation reaction is calculated by equation
(4) and (5). The shape of dusts was assumed as spherical. And their shape change was ignored.
 𝑘𝑔
𝐽𝑂2 = 4π𝑟𝑜2 (𝑝𝑂𝑏2 − 𝑝𝑂𝑠 2 ) (4)
𝑅𝑇

𝑘𝑔
𝐽𝑂2 = 4π𝑟𝑜2 𝑝𝑂𝑏2 𝑎𝑡 𝑝𝑂𝑠 2 ≅ 0 (5)
𝑅𝑇

For the oxidation reaction of the dust particle, the reaction progress and O2 gas partial pressure
around the dust particle are illustrated in Figure 3.

Figure 3. Schematic illustration of dust particle on oxidation reaction

The diffusivity of O2 gas D is estimated by equation (6). This was based on the Chapman-Enskog
equation and simplified as function of temperature in the range of 773 to 1723 K.

3/2
𝐷 = 3.67𝑥10−9 𝑇𝑔 (6)

The gas film mass transfer coefficient kg is calculated by following equations (7), (8), and (9).

𝐷 1/2 (7)
𝑘𝑔 = (2 + 0.6𝑅𝑒𝑝 𝑆𝑐1/3 )
𝑑𝑝

𝑑𝑝 (𝑢𝑝 − 𝑢𝑔 )𝜌𝑔 (8)

𝑅𝑒𝑝 =
𝜇
𝑢𝑔 (9)
𝑆𝑐 =
𝜌𝑔 𝐷

Sulfation reaction

The flue dust moving to the WHB change from liquid droplets to spherical solid particles as
temperature decreases. As mentioned above, the sulfation reaction of dusts is the reaction of Cu 2O with
gases as shown by the equation (3). The reactant Cu2O for sulfation reaction can be considered as solid
phase through the information on its melting point and the general temperature in the WHB radiation
section. Therefore, the sulfation reaction is defined as a solid-gas reaction. The unreacted shrinking core
model commonly used for solid-gas reactions was adopted for the sulfation reaction in present study.

According to the Barin (1995), CuSO4 is not stable and thermally decomposes above 800℃. The
sulfation reaction of Cu oxide occurs only when the particle temperature T p is lower than the normal
decomposition temperature of CuSO4 TNDT which is applied on the reaction paths as shown in Figure 2.
This study assumed that the size of dust does not change during the reaction. O2 and SO2 gases are required
for the sulfation reaction. Since the O2 gas partial pressure pO2 is lower than SO2 gas partial pressure in off-
gas stream of the uptake shaft and the WHB, it was assumed that the overall reaction rate is determined by
the partial pressure of O2. The O2 gas consumption rate JO2 of sulfation reaction is determined by equation
(10).
𝑏
𝑝𝑂 /𝑅𝑇 (10)
𝐽𝑂2 = 4π𝑟𝑜2 (𝑟
2
2
1 𝑟𝑜 𝑜 −𝑟𝑖 ) 𝑟𝑜 1
+( ) +( )
𝑘𝑔 𝑟𝑖 𝐷𝑒 𝑟𝑖 𝑘𝑟

On the sulfation reaction of the dust particle, the reaction progress, O2 gas partial pressure around
the dust particle and changes of product layer and unreacted core are represented in Figure 4.

Figure 4. Schematic illustration of dust particle on sulfation reaction

In the unreacted shrinking core model for the sulfation reaction, the effective diffusivity of O2 gas
through the produced Cu2O-Fe2O3 oxide layer De is determined by equation (11). The labyrinth factor ε
was assumed as 0.2 which is from Sohn’s (1993) work for oxidation kinetics of copper concentrates.

𝐷𝑒 = 𝜀 2 𝐷 (11)

The rate constant of interfacial reaction between unreacted core and product layer kr was
calculated by equation (12). The activation energy for this interfacial reaction was determined from
experimental data of Ranki-Kilpinen (2004), and the value is 153 kJ/mol. Frequency factor ko was defined
as log 7.5 through the following calculation processes as shown in Figure 4 with the observed data and the
trial and error method.

𝐸𝑟 (12)
𝑘𝑟 = 𝑘𝑜 exp (− )
𝑅𝑇

Physical properties and thermochemical data

The density of gas ρg is calculated by the ideal gas law. The gas viscosity μg is determined by the
Sutherland’s formula. The thermal conductivity of gas kc,g is estimated by the following equation (13)
which refer to Bird et al. (2007).

5𝑅 (13)
𝑘𝑐,𝑔 = (𝐶𝑝 + )𝜇
4𝑀 𝑔
 The density of molten copper matte whose grade is 67% was applied to the dust density in this
study. The density value was calculated using the empirical equation by Goto (1988). The value was 4900
kg/m3. Present study assumed that the density of dusts was constant in order to reduce complexity of the
model.

The melting phenomena of dust particles are considered in the model of this study. As dust
reaction progresses, the melting temperatures vary with the changes of phases and compositions in dusts.
Therefore, the melting points of dusts should be simplified in order to achieve the convenience of
calculation. In this study, the melting temperature of the dusts was defined as 1150℃ from available phase
diagrams. (Cu2S-FeS phase diagram by FactSageTM 7.1 software and Cu2O-Fe2O3 phase diagram by
Nikolic et al (2009))

Thermochemical data such as enthalpy increment, heat capacity and formation heat of various
substances used the data of Pankratz (1982, 1987, 1994) and Barin (1995).

Flow sheet of the dust reaction model

The flow sheet of this modeling study is shown in Figure 5. The calculation procedure is as
follows. Firstly, the conditions such as density, diameter of dust particles, initial gas temperate and O2
partial pressure, dust amounts and various variables are given. The model computes sequentially gas
properties, particle reactions, heat transfer from particle to gas and wall, mass balance and reaction ratio
with time.

Figure 5. Flow sheet of the dust reaction model

RESULTS AND DISCUSSION

Simulation results and effects of temperature

The dust reaction model by this study shows various simulation results as presented in Figure 6.
The calculation conditions are as follows. (dust amounts W=0.76 kg/s, dust diameter dp=0.3 mm, labyrinth
factor ε=0.2 and frequency factor ko=log7.5)
 (a) Changes of dust reaction ratio and temperature (b) Changes of phases and amounts of dust

(c) Changes of pO2 and unreacted core radius (d) Changes of resistance fraction

Figure 6. Simulation results by the dust reaction model

The changes in dust temperature and dust reaction ratios with time were plotted in Figure 6. (a). It
is confirmed that the oxidation reactions of dusts (X1 = Fe as oxide / Total Fe, X2 = Cu as oxide / Total Cu)
occur rapidly. It is considered that these are due to the influence of increasing diffusivity D by relatively
high gas temperature. As mentioned above, this model considers the reaction heats of dust and gases and
heat transfer of dust particle between gases and wall. The rapid temperature rise at the beginning is because
the exothermic heat by dust oxidation is more dominant than heat transfer to gases and wall. The gradual
temperature drops of dust particle with time show that the influence of heat transfer is larger than heat
generation by oxidation and sulfation reactions.

When dust temperature Tp is lower than the normal decomposition temperature of CuSO4, the
sulfation reaction of Cu2O proceeds as presented in Figure 6. (a) and (b). These show that the rates of
sulfation reaction (X3 = Cu as sulfate / Total Cu) decrease gradually. This model considers the gas
temperature drop with time. It is considered that the gradual reduction of sulfation reaction rate was mainly
influenced by the temperature decrease.

Figure 6 (d) shows the changes of the resistance fractions for sulfation reaction. The resistances of
each step of sulfation reaction are gas film mass transfer controlling resistance λ1, ash layer diffusion
controlling resistance λ2 and interfacial reaction controlling resistance λ3. And they are determined by
equation (14), (15) and (16) respectively.
 1 (14)
𝜆1 =
𝑟𝑜2 𝑘𝑔

(𝑟𝑜 − 𝑟𝑖 ) (15)
𝜆2 =
𝑟𝑜 𝑟𝑖 𝐷𝑒

1 (16)
𝜆3 =
𝑟𝑖2 𝑘𝑟

And the resistance fractions of each step η1, η2 and η3 are calculated by equation (17).

𝜆𝑖 (17)
𝜂𝑖 =
(𝜆1 + 𝜆2 + 𝜆3 )

The mentioned decrease of sulfation reaction rate are considered to the results of complex effects
of the temperature drop with time and changes of gas film, ash layer and unreacted core radius as reaction
progresses. In addition, the change of unreacted core radius is shown in Figure 6 (c).

It is suggested that the relatively higher temperature (but lower than decomposition temperature of
CuSO4) is required in order to maximize the sulfation reaction.

Effects of dust particle sizes

The computation results of effects of dust sizes are shown in Figure 7. In given conditions, Fe is
completely oxidized, but amounts and phase fractions of Cu bearing substances such as Cu2S, Cu2O and
CuSO4 change by the dust particle sizes. Fractions of CuSO4 decrease with increasing the dust sizes, which
means larger dust particles need more time to complete the reactions. As presented in equation (5), (7), (8)
and (10), the dust particle size (diameter dp and initial radius ro) affects the particle Reynolds number Rep,
the gas film mass transfer coefficient kg and the O2 gas consumption rate JO2. It is confirmed that the
decrease of dust particle size increases the reaction rate by the equation (7). To maximize the dust reactions
ideally smaller dust sizes are needed.

Figure 7. Mass fractions of each phase by various dust particle diameters

Effects of gas partial pressure

As shown in the Figure 8 (a), the reaction ratio increases with increasing the initial O2 partial
pressures. This is the results of increase of O2 gas consumption rate JO2 by increasing O2 partial pressure as
presented in equation (5) and (10).

Both simulation results and industrially investigated data are presented in the Figure 8 (a). It
shows that the calculation results are positioned in the range of measured values. And it is shown that the
simulation and the actual could be relatively similar.

(a) Dust reaction ratio by O2 partial pressure (b) O2 in remained gases

Figure 8. Dust reaction ratio and remained O2 in gases by various O2 partial pressures

The increase of O2 gas amounts is practically an effective way in order to maximize the oxidation
and sulfation reactions of dusts. By the dust reaction model, the required O2 amounts could be estimated
for achieving the target reaction ratio. However, the large input amounts of O2 make the increase amounts
of remained O2 as shown in Figure 8 (b). The residual O2 gases will react with SO2 and H2O gases, then,
produce the SO3 and H2SO4, respectively. As is well known, they affect the corrosion of facilities and the
weak acid formation of waste acid treatment process. When considering the side effects of increase of O2
gas amounts, it is necessary to control O2 gas amounts in proper ranges. The dust reaction model could be
used to evaluate the appropriate levels of the dust reaction ratios and the remained O 2 gas.

CONCLUSIONS

The present study developed the dust reaction model in the S furnace uptake shaft and the S-WHB
of the Onsan copper smelting process. In the modeling, the reaction paths of dusts are reviewed based on
thermodynamic equilibria with investigated data. The reaction rate equations of dusts adopted the gas film
mass transfer controlling and the unreacted shrinking core model for oxidation and sulfation reaction,
respectively. The developed model can describe effects of various variables on dust reactions. And varied
changes such as gas partial pressures, particle temperatures and resistance fractions by the dust reaction
could be also simulated through this model.

Using the dust reaction model, this study reviewed the effects of temperatures, dust particle sizes
and O2 gas partial pressure, especially.

On the effects of temperatures, higher temperatures increase the rates of dust reactions because of
the relationship between temperature and the diffusivity, the effective diffusivity and the rate constant of
interfacial reaction. The relatively higher temperature (but, lower than decomposition temperature of
CuSO4) is required in order to increase the sulfation reaction.

Decreasing the dust particle size increases the reaction rate of dusts based on the relationship
between the particle sizes, gas film mass transfer coefficients and O2 consumption rates.

Increasing O2 gas amounts is practically an effective way in order to maximize the oxidation and
sulfation reactions of dusts. When considering the side effects by residual O2 gas such as corrosion, it is
necessary to control in proper ranges. The developed model could be used to evaluate the appropriate
levels of the dust reaction ratios and the remained O2.

NOMENCLATURE
Cp Heat capacity J/mol-K
D Diffusivity of O2 = 3.67x10-9Tg3/2 (unit of T is Kelvin) m2/s
De Effective diffusivity in product layer = ε2D m2/s
dp Particle diameter m
Er Activation energy J/mol
Ho-Ho298 Enthalpy increment J/mol
Convective heat transfer coefficient
hp = kc,g(2+0.6Rep1/2Pr1/3)/dp J/m2-s-K
between gas and particle
JO2 O2 gas consumption rate molO2/s
kc,g Thermal conductivity of gas = (Cp+5R/4M)μg J/m-K-s
kg Gas film mass transfer coefficient = D(2+0.6Rep1/2 Sc1/3)/dp m/s
ko Frequency factor of kr m/s
kr Rate constant of interfacial reaction = ko exp(-Er/RT) m/s
M Molar weight g/mol
pO2 O2 partial pressure [Suffix] b: bulk, s: surface atm
Pr Prandtl number = Cp,gμg/kc,g -
R Universal gas constant 8.314 J/mol-K
Rep Particle Reynolds number = dp(up-ug)ρg/μ -
ri Radius of unreacted core m
ro Initial radius of particle m
Sc Schmidt number = ug/ρgD -
Sh Sherwood number = 2+0.6Rep1/2 Sc1/3 = kgdp/D -
T Temperature [Suffix] w: wall, g: gas, p: particle ℃ or K
t Time s
u Velocity m/s
W Dust(particle) amounts kg/s
[Suffix]
1: oxidation of FeS,
X Reaction ratio, Conversion ratio -
2: oxidation of Cu2S,
3. Sulfation of Cu2O
ε Labyrinth factor, porosity factor -
εp Emissivity of particle -
[Suffix]
Resistance fraction of the respective 1: gas film mass transfer controlling,
η -
steps 2: ash layer diffusion controlling,
3: interfacial reaction controlling
λ Resistances of the respective steps [Suffix] same as above s/m3
μ Viscosity kg/m-s
ρ Density kg/m3
σ Stefan-Boltzmann constant 5.67x10-8 J/m2-s-K4
 REFERENCES

Barin, I (1995). Thermochernical Data of Pure Substances (3rd ed.). VCH Verlagsgesellschaft mbH,
Weinheim.

Bird, R. B., Stewart, W. E. & Lightfoot, E. N. (2007). Tranport phenomena (Revised 2 nd ed.). John Wiley
& Sons, Inc. (pp. 525-528 and 863-866).

FactSage, Montreal, Canada, www.factsage.com

Goto, M., (1988), The Mitsubishi continuous process - operation and design, pp. 120, Mitsubishi Materials
Corporation.

https://www.lmnoeng.com/Flow/GasViscosity.php (for gas viscosity calculation)

Nikolic, S., Hayes, P. C, & Jak, E. (2009), Experimental Techniques for Investigating Calcium Ferrite
Slags at Metallic Copper Saturation and Application to the Systems ‘‘Cu2O’’-‘‘Fe2O3’’ and
‘‘Cu2O’’-CaO at Metallic Copper Saturation, METALLURGICAL AND MATERIALS
TRANSACTIONS B, VOL 40B, DEC 2009, pp. 892-899.

Pankratz, L. B. (1982). Thermodynamic Properties of Elements and Oxides, Bulletin 672, US department
of the Interior, Bureau of Mines.

Pankratz, L. B. et al. (1987). Thermodynamic Properties of Sulfides, Bulletin 689, US department of the
Interior, Bureau of Mines.

Pankratz, L. B. (1994), Thermodynamic Properties of Carbides, Nitrides, and Other Selected Substances,
Bulletin 696, US department of the Interior, Bureau of Mines.

Ranki-Kilpinen, T. (2004). Sulphation of cuprous and cupric oxide dusts and heterogeneous copper matte
particles in simulated flash smelting heat recovery boiler conditions, Helsinki University of
Technology Doctoral Theses in Materials and Earth Sciences.

Sohn, H. S. (1993). Kinetic Studies on Smelting Reactions and Elimination of Minor Elements in Copper
Smelting Process, Doctoral thesis, Faculty of Engineering, Kyoto University.

View publication stats