CFD simulation of the preheater cyclone of a cement plant and the optimization of its performance using a combination of the design of experiment and multi-gene genetic programming
Graphical abstract
Introduction
The application of cyclone separators in separating particles from gas flow began hundreds of years ago. Owing to the wide range and simplicity of the cyclone structure that leads to a low cost of investment and maintenance, they are still widely used in the chemical and petrochemical industries. The main parameters of cyclone performance are separation efficiency and pressure drop [1]. Many efforts have been made to derive an appropriate mathematical model for the study and prediction of flow behaviour as well as to find suitable cyclone geometries. However, due to the hydrodynamic complexity of cyclones, no exact mathematical model for cyclones has been offered yet. For numerical solutions, CFD is a proper tool [2], [3]. The key to the successful use of CFD for finding turbulent flow pattern depends on the selection of the right turbulence model. Selecting a suitable turbulence model has been studied by many researchers [2], [4]. The result of the Reynolds stress model (RSM) among the turbulence models has a very good agreement with the experimental data [5], [6], [7], [8] since the isotropy of the flow assumption is ignored. In addition, large-eddy simulation methodology can predict turbulent flow inside the cyclone [4], [9], [10].
The pressure drop and the cut-diameter in cyclones are important objective functions, to be optimized simultaneously.Cyclone geometry has a significant effect on its performance; various researches have been conducted to investigate the effect of geometrical parameters on cyclone performance [11], [12], [13]. Safikhani et al. [14] obtained the pressure drop and cut-diameter by CFD and then optimized the cyclone performance with the assistance of multi-objective optimization; the artificial neural network was then used to calculate the objective function. Elsayed and Lacor performed many studies in order to optimize the cyclone [15], [16], [17]. They used multi-objective optimization for the performance of the cyclone [18]. Two radial-basis function neural networks were used to model the pressure drop and the cut-diameter. Their results showed a noticeable influence of the vortex finder diameter and its length, the inlet width, and the total height on the efficiency of the cyclone. Fathizadeh et al. [19] used Leith and Licht theory to prepare some algorithms to obtain the dimensions of the Stairmand cyclone. Vortex finder length was optimized using genetic algorithm to prevent the direct escape of particles from vortex finder.
Owing to environmental concerns-and as the cement industry is one of the major producers of greenhouse gases-strict regulations have been enforced in this industry. In recent years, different researches on the various techniques that improve the cement production process and reduce resultant pollution [20] have been carried out. Mikulcic et al. [21] showed the application of CFD in the implementation of the calcination reaction, which is carried out in calciner in order to reduce carbon dioxide emissions in the cement industry. Mikulcic et al. [22] numerically studied gas-solid flows in a cement cyclone using the Eulerian-Lagrangian approach. The heat exchange phenomenon in the cyclone, arising from the heat exchange between the gas-solid phase and the endothermic calcination reaction, was investigated. Cristea and Conti [23] studied the heat exchange between the gas and the solid phase in a cement plant preheater. They concluded that the most of the heat distribution occurs within the duct, and the temperatures of the gas and particles in the cyclones remain approximately constant and equal. Ferrite and Ari [9] optimized the preheater cyclone by means of the Taguchi method. They optimized the cyclone performance by changing the four parameters of diameter, vortex-finder length, velocity inlet, and the concentration of particles. By modernizing the design of cyclone preheaters, significant energy saving can be achieved. The role of the first-stage cyclones is significantly important as specifying the separation efficiency. Wasilewski and Duda [24] for optimizing the first-stage cyclones proposed guidelines depending on the objective function. In another study [25] they optimized the geometry of cyclone. Mariani et al. [26] optimized the length and angle of vortex finder to increase the separation efficiency and decrease the exit gas temperature.
Section snippets
Background
In traditional cement manufacturing technology, the preheater system is positioned before the rotary kiln in order to increase the heat transfer between the raw materials and hot gases. Raw materials in the preheater are gradually dried, heated, and calcined by moving against the flow of the hot gases coming from the rotary kiln. In Fig. 1, a rotary kiln with a suspension preheater similar to that one from Kerman cement plant is shown. Preheaters are built in several stages in order to increase
The governing equations
For an incompressible and steady-state fluid flow, the continuity equation for the mean flow is as follows [29]:
The time-averaged Navier–Stokes equation is as follows [29]:
The last term is defined as the Reynolds stress tensor, which reflects the effects of turbulent fluctuations in the fluid flow. The RSM used in this study is required for solving equations for each of the Reynolds stresses. In the Reynolds stress model, the
Cyclone geometry
The cyclone was a scroll type in which the radius of the cylindrical part (R) was 1.26 m and the radius of the scroll part (Rs) was 1.722 m, which was located at a distance of δr = 0.64 m from the cylinder centre. Cyclone dimensions are provided in Table 1; it consists of cyclone diameter (D), inlet height (a) and width (b), vortex-finder diameter (De) and length (S), the height of outlet pipe (l) (it was selected greater than usual, in order to be sure of accessing fully developed flow in the
Multi-gene genetic programming
Genetic algorithm is a search technique in computer science used to find approximate solutions for optimization. Genetic algorithms apply the principle of Darwin's theory. (According to Darwin's theory, the generations that have better characteristics than other generations will have a better chance of survival and reproduction; their superior characteristics are also transmitted to the next generation). Genetic algorithm with random initial population generation checks the problem space
The mesh independency and the validation of the model
The computational results of the three different meshes for the preheater cyclone pressure drop and efficiency are provided in Table 2, Table 3 respectively. Since the pressure drop and cyclone efficiency difference between both the sequential meshes were minimum (< 5%), the CFD results are independent of the mesh size. To reduce computational times, the mesh number of 2,208,864 was used in this study. The last row in Table 2, Table 3 represent experimental pressure drop and efficiency of the
Conclusions
The first-stage cyclones' efficiency in the preheater system of a cement plant is very important. In this study, one of the twin cyclones in the first-stage of the preheater in the Kerman cement plant was investigated using a combination of CFD, the design of the experiment, and multi-gene genetic programming. Since its structure had been changed, a decrease in its efficiency had taken place. In the same pressure drop and volumetric flow rate, the Stairmand cyclone efficiency was 2.5% greater
Nomenclature
- a
inlet height (m)
- b
inlet width (m)
- B
the cone-tip diameter (m)
- D
cyclone diameter (m)
- De
vortex-finder diameter (m)
- Dij
stress diffusion
- Ef
efficiency
- Eu
Euler number
- h
cylindrical part height[m]
- hs
part of the scroll height (m)
- H
the cyclone total height (m)
- Hc
conical height (m)
- L
the height of outlet pipe (m)
- P
pressure (Pa)
- Pij
shear production
- r
radius of cyclone (m)
- S
vortex-finder length (m)
- U
inlet velocity (m/s)
- ui
time averaged fluid velocity i (m/s)
- u′i
fluctuating velocity to direction i (m/s)
sub grid scale velocity
Acknowledgements
The authors acknowledge the Kerman cement plant, Iran in allowing them to use their experimental data.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
References (45)
- et al.
Computational fluid dynamics (CFD) and empirical modelling of the performance of a number of cyclone samplers
Aerosol Science
(1996) - et al.
Solids concentration simulation of different size particles in a cyclone separator
Powder Technol.
(2008) - et al.
The effect of modeling of velocity fluctuations on prediction of collection efficiency of cyclone separators
Appl. Math. Model.
(2013) - et al.
A CFD study of the effect of cone dimensions on sampling aerocyclones performance and hydrodynamics
Powder Technol.
(2006) - et al.
Numerical investigation of the effects of the central channel on the flow field in an oil–gas cyclone separator
Comput. Fluids
(2014) - et al.
Numerical simulation of effect of inlet configuration on square cyclone separator performance
Power Technology
(2011) The efficiency improvement of a large-diameter cyclone – the CFD calculations
Sep. Purif. Technol.
(2013)- et al.
The effect of cyclone vortex finder dimensions on the flow pattern and performance using LES
Comput. Fluids
(2013) - et al.
The effect of cyclone inlet dimensions on the flow pattern and performance
Appl. Math. Model.
(2011) - et al.
Characteristics of the collection efficiency for a cyclone with different vortex finder shapes
Aerosol Science
(2004)