A Dynamic Model of a Submerged Plasma Slag Fuming Process
241
of heat and mass; the heat and mass transfer to the subsequent zones are assumed
only after attaining the thermodynamic equilibrium. The heat supplied and lost in
each sub-unit operation were also taken into account. Microsoft Excel was used as
a user interface to set up the initial conditions for the simulations and to write the
results.
The submerged plasma zinc fuming process is a complex process with several
reactions between petcoke-slag, gas-slag, and petcoke-gas. Modelling the process
and fitting with the reality would be challenging without making assumptions and
simplifications. Therefore, the following assumptions and simplifications are made
in the model while maintaining the crux of the process. It is assumed that the slag
bath is homogenous (uniform in temperature and composition), the slag bath does
not splash, the tuyere gas partially bypasses the slag bath, there is no interaction
between the slag bath surface and the off-gas, the unreacted slag does not exist in
the slag bath, and the freeze lining thickness changes with the process conditions.
Model Predictions
The dynamic model of the submerged plasma slag fuming process includes several aspects of Metallo’s slag fuming process, such as change in bath height during
fuming, changes in freeze lining thickness, two reaction zones in the reactor (in
scenario-2), and heat losses in different reaction zones (in scenario-2). These features helped to predict the fuming rate, slag bath and off-gas temperature, slag bath
and off-gas composition, and several other process parameters.
The data for the simulations were taken from Metallo’s process database. The
batches with a stable fuming cycle were selected for the simulation. For the fuming
cycle considered to be stable, the following criteria must be satisfied: the plasma
generator power and the under-pressure created at the post-combustion duct must be
constant during the fuming, there must be no fluctuations in the reactor heat losses,
there must be no outliers in the measured ZnO content of the slag (the coefficient
of determination (R
2 ) of the ZnO content of the slag with time must be greater than
99%), and there must be no solid slag filling during the fuming process.
Two scenarios were simulated using the model to understand the rate-limiting
factors of the fuming process. In the first scenario, it is assumed that the process
happens at thermodynamic equilibrium, and there are no kinetic factors affecting the
process. In the second scenario, it is assumed that the process is affected by kinetic
factors. From the literature, it is known that the tuyere gas can partially bypass the
slag bath, and it is not in thermal equilibrium. Therefore, the parameter tuyere gas
efficiency, which is the amount of tuyere gas in equilibrium with the slag bath, is
used to fit the calculated ZnO content with the measured ZnO content in the slag.
The initial temperature of the slag bath (T SB,i ) is not measured and is an unknown.
Therefore, it is used as a variable to understand the importance of this parameter on
the fuming process.
241
of heat and mass; the heat and mass transfer to the subsequent zones are assumed
only after attaining the thermodynamic equilibrium. The heat supplied and lost in
each sub-unit operation were also taken into account. Microsoft Excel was used as
a user interface to set up the initial conditions for the simulations and to write the
results.
The submerged plasma zinc fuming process is a complex process with several
reactions between petcoke-slag, gas-slag, and petcoke-gas. Modelling the process
and fitting with the reality would be challenging without making assumptions and
simplifications. Therefore, the following assumptions and simplifications are made
in the model while maintaining the crux of the process. It is assumed that the slag
bath is homogenous (uniform in temperature and composition), the slag bath does
not splash, the tuyere gas partially bypasses the slag bath, there is no interaction
between the slag bath surface and the off-gas, the unreacted slag does not exist in
the slag bath, and the freeze lining thickness changes with the process conditions.
Model Predictions
The dynamic model of the submerged plasma slag fuming process includes several aspects of Metallo’s slag fuming process, such as change in bath height during
fuming, changes in freeze lining thickness, two reaction zones in the reactor (in
scenario-2), and heat losses in different reaction zones (in scenario-2). These features helped to predict the fuming rate, slag bath and off-gas temperature, slag bath
and off-gas composition, and several other process parameters.
The data for the simulations were taken from Metallo’s process database. The
batches with a stable fuming cycle were selected for the simulation. For the fuming
cycle considered to be stable, the following criteria must be satisfied: the plasma
generator power and the under-pressure created at the post-combustion duct must be
constant during the fuming, there must be no fluctuations in the reactor heat losses,
there must be no outliers in the measured ZnO content of the slag (the coefficient
of determination (R
2 ) of the ZnO content of the slag with time must be greater than
99%), and there must be no solid slag filling during the fuming process.
Two scenarios were simulated using the model to understand the rate-limiting
factors of the fuming process. In the first scenario, it is assumed that the process
happens at thermodynamic equilibrium, and there are no kinetic factors affecting the
process. In the second scenario, it is assumed that the process is affected by kinetic
factors. From the literature, it is known that the tuyere gas can partially bypass the
slag bath, and it is not in thermal equilibrium. Therefore, the parameter tuyere gas
efficiency, which is the amount of tuyere gas in equilibrium with the slag bath, is
used to fit the calculated ZnO content with the measured ZnO content in the slag.
The initial temperature of the slag bath (T SB,i ) is not measured and is an unknown.
Therefore, it is used as a variable to understand the importance of this parameter on
the fuming process.