CFD Simulations of Gases Flow in Calciners

As there is a broad range of constructive features and – more important – operating conditions regarding calciners that are used in clinkering plants, CFD simulations could be an effective tool in increasing their performance. A methodology that could be used to optimize their functioning is given here. In brief, CFD simulations are carried out both for computing the average gases velocity at their passage through the calciner and visualize exactly the flow patterns within the vessel. By properly adjusting dimensional and design features with the known outputs from the neighboring apparatus and with the required retention time of the particles in the calciner corresponding to a given decarbonation degree one could reach high running efficiency with low energy loses, without the need for demanding and expensive live trials.


CFD Simulations of Gases Flow in Calciners
As there is a broad range of constructive features and -more important -operating conditions regarding calciners that are used in clinkering plants, CFD simulations could be an effective tool in increasing their performance.A methodology that could be used to optimize their functioning is given here.In brief, CFD simulations are carried out both for computing the average gases velocity at their passage through the calciner and visualize exactly the flow patterns within the vessel.By properly adjusting dimensional and design features with the known outputs from the neighboring apparatus and with the required retention time of the particles in the calciner corresponding to a given decarbonation degree one could reach high running efficiency with low energy loses, without the need for demanding and expensive live trials.

Keywords: computational fluid dynamics, clinker calciners
Knowledge of the way gas and particles flow into an industrial reactor is important not only from a scientific viewpoint but, also, within efforts to optimize reactor's functionality.Calciners, as components of the clinkering installations, claim -usually -50 to 60% of the fuel consumption while providing up to 95% decarbonation degree for the hot meal entering into the rotary kiln.Within calciners, several processes take place instantaneously, such as combustion, complex heat transfer, decarbonation, gas and material transport, particles clustering (and the opposite phenomenon) etc. difficult to be modeled analytically altogether at that stage.Moreover, inputs and outputs of the calciner in terms of material, gases and fuel flows, heat flows, materials characteristics (size distribution, decarbonation degree) should be integrated in the global model of the clinkering installation, comprising, also, the rotary kiln [1], cyclones and grate cooler.Taking into account all these remarks, the paper should be regarded as constituting a step in a broader optimal design of the clinkering installation, when the outputs of a component becomes inputs for another.For each of the components the term "optimal design" [2] comprises all the constructive and functional characteristics that could optimize the local performances while, relating at the same time, to the other apparatus in the given system.The methodology used here was to maximize the retention time of the particles in the calciner so the calcining process should be carried at its maximum extent (without affecting the reactivity of the material further on, into the rotary kiln).This can be done, in the authors' opinion, by fine-tuning dimensional features with gases velocities at the inlets of the calciner and by computing the average velocity of the gases in their passage through the calciner.As it is known, particles retention time could be correlated, at a given temperature, with the decarbonation degree [3].By setting the last one and the inlet velocities (which results from processes taking place in the adjacent apparatus), it could be determined the corresponding, optimal shape of the calciner.The goal of the paper is not to particularize a certain calciner, part of a given clinkering installation, but only to give the methodology and to present and analyze some applications.

Method
As analytical methods are excluded due to their inability to deal with such complex phenomena that take placeconsidering, as it is the case here, only gases flow -CFD (Computational Fluid Dynamics) methods were used.While CFD methods are based on solving flow equations under different circumstances [4-7], their applications, typically, include other models such as combustion, heat transfer, pollutants generation etc.Such an application will comprise three major steps: a) preprocessing, in which the reactor is drawn and meshed according to the required accuracy and the problem is defined (boundary conditions are given); b) processing (a solving method is selected); c) postprocessing, required to show results in such a way that they could be easily analyzed.As it could be easily observed, the first stage is the most important not only in terms of working time but also for the success of the simulation.It is a common habit to repeat the simulation thus the whole work will be subjected to a heuristic approach (this is the point in which experience will be a key factor for the selection of the proper solving method, mesh refinement, values of different coefficients involved in the model).Being an iterative method while considering its inner complexity, one should pay a particular attention to the convergence criterion, as long it is not confirmed as an universal solution.Among different stopping criteria, here there were used -along with mass balancing -a maximum value (10 -3 ) to be reached by the residuals summed over the whole volume (fig. 1) or/and a very slow residual decrease (or even a flattening of the evolution) (fig.2).

Results and discussions
The first simulation targets an ILC type calciner (fig.3).Boundary conditions are set to: tertiary air velocity -25m/ s, gases from kiln -20m/s and coal particles suspension -10m/s.Gases resulting from coal combustion were included into inlet velocities and afterwards the result was rounded to actual values.
The useful volume was meshed to 172167 tetrahedral cells, found to be both effective in terms of accuracy and computation time expenditure.The averaged gases velocity over the entire useful volume was found to be 7.702 m/s.The results of the simulation consist not only in valuable, numerical values; images of the way gases flow inside the vessel provide additional information about the zones having different velocities and flow orientation, for the given calciner configuration and inputs.For example, in that case, figures 4 and 5 show the existence of dead zones while the mainstream concentrates on the rest.It is easily to understand that an optimized configuration (comprising both geometrical model and boundary conditions) will ask for the most possible part of the vessel to be in a swirling type flow (in order to increase path in a lesser amount of volume) and, then, to acquire an uniform flow in every cross section to be set over the rest of the calciner.This is not the case here, that could be regarded as a good example of a bad combination of constructive and functional characteristics of the calciner.
The second case study concerns an ILC-Low NO x- calciner, incorporating the cold end of the rotary kiln.The reason for this was: that the gases exiting the kiln in their path to calciner will experience a steep turn and this will entail changes in their orientation the velocity.Therefore, the corresponding direction and velocity profile will be, in  the transition area, not uniform as assumed in the first case study.It is to be expected, by that way, that simulation to become more scrupulous.However, in the simulation was used a simplification: cross-sectional area filled with material was neglected.That area occupies about 10% of the total cross-sectional area of the rotary kiln; the reason it was neglected is related to the difficulties encountered during the design stage.Although it is certain that it affects the local flow distribution, it is not exactly known the influence exerted on the overall flow (balancing with the major factor.i.e. the gases change in direction).Gases velocity when they leave the kiln was set 10m/s while tertiary air was assigned with different velocity values.Working area was covered with 172167 tetrahedral cells.
To get a complete picture, the results could be given both numerically (table 1) and visually (figs.6,7) and consequently differences could be observed and analyzed easier.

Conclusions
Data gathered through CFD methods offer a wide range of information, both quantitative and qualitative.Altogether, these information could lead -by attempting successive changes in constructive and functional features -to optimize the predicted performance of the calciner in agreement with given technological requirements.Modeling the calciner to acquire its optimal features has to be a part of the whole effort to design the entire clinkering a c b c b a plant, as connections to all the other components involve awareness of the mass and heat transfer between them.This observation imply for the clinkering plant's optimal modelling that components design will continuously exchange information between them, and the complete process could iterate until the technological requirements will be fulfilled.