Numerical simulations of fluidization dynamics in a hot model of a CLC process

Chemical Looping Combustion (CLC) is one of the most promising alternatives for solid fuel combustion. CO2 concentration in the exhaust gas is high in CLC technology which enables high efficiency of CO2 capture from flue gas. The use of solid oxygen carriers is a characteristic feature of a CLC process. Oxygen carriers are mainly metal oxides which are characterized by high oxygen transfer capacity and high mechanical resistance. Since the CLC technology is not sufficiently recognized due to its complexity the development of models with real conditions of the CLC equipment is of practical significance. The paper presents numerical simulations of the dynamic fluidized bed for Chemical Looping Combustion using CeSFaMB software. The model was validated on the basis of the results obtained from experiments, which were carried out on the Fluidized-Bed Chemical-Looping-Combustion of Solid-Fuels (FB-CLC-SF) unit. The studies were conducted in air atmosphere at temperature of 850°C. The validation of the 1.5D model showed that the maximum relative error between experiment and simulations results does not exceed 12%.


Introduction
Burning of fossil fuels is one of the main sources of CO2 emissions into the atmosphere.It is therefore essential to develop new combustion technologies, which allows the use of fossil fuels while reducing CO2 emissions [1].One of the interesting alternative to this problem is Chemical Looping Combustion (CLC).This technology is characterized by using solid oxygen carriers for fuel combustion [2].The absence of air in the combustion chamber allows to obtain the exhaust gas containing mostly CO2 and H2O (Fig. 1).The almost pure CO2 is produced of the flue gas cleaning and condensation of steam.Therefore, the chemical looping combustion process allows to avoid expensive CO2 absorption installations [3,4].
This paper presents the results of numerical simulations of the fluidization dynamics in a hot model of a CLC process which are a continuation of the previous works carried out on the cold model [3,5].
The test units for cold and hot model (FB-CLC-SF) are located in Institute of Advanced Energy Technologies at the Czestochowa University of Technology and they were built in the framework of Polish -Norwegian research project NewLoop "Innovative Idea for Combustion of Solid Fuels via Chemical Looping Technology" [6].

An object of investigations
The innovative ideology of NewLoop project combines two complementary technologies: CLOU (Chemical looping with Oxygen Uncoupling) and iG-CLC (In-situ Gasification Chemical Looping), which create a so-called HCLC (Hybrid Chemical Looping Combustion).The CLOU process is characterized by the use of solid oxygen carrier, which does not require direct contact with the fuel to release oxygen in the fuel reactor.By contrast, the iG-CLC process requires that the oxygen carrier is in contact with the fuel [5,7].
The experimental studies, as well as validation of the model were carried out on a Fluidized-Bed Chemical-Looping-Combustion of Solid-Fuels (FB-CLC-SF) unit (Fig. 2a).The system consists of two main circular contours.The first circular contour is composed of an air reactor with circulating fluidized bed, riser, cyclon and return system.The second circular contour consists of a fuel reactor with bubbling fluidized bed and particulate collector.(Fig. 2b).The main dimensions of the FB-CLC-SF unit are presented in Tables 1 and 2.

Research material
The sand with the Sauter mean diameter of particles of 166 μm, true density of 3200 kg/m 3 and the sphericity of 0.7 are used during the investigations (Fig. 3).

Characteristics of CeSFaMB software
The numerical simulations of the CLC process were carried out using the CeSFaMB software.(Comprehensive Simulator of Fluidized and Moving Bed equipment), which is especially dedicated to the modeling of fuel conversion implemented in bubbling and circulating fluidized bed (Fig. 4).CeSFaMB simulator requires the formulation of the so-called initial and boundary conditions.In order to confirm the reliability of the emerging model, both introduced and designated values should be verified by experimental studies.Otherwise, the results of the model calculations are purely hypothetical [5].CeSFaMB allows to determine the following operating parameters of the fluidized systems: [8-10]: temperature profile, particle-size distribution profiles, concentration and mass flow profiles of 20 gas components (Ar, CO2, CO, O2, N2, H2O, H2, CH4, SO2, NO, N2O, NO2, HCN, C2H6, H2S, NH3, C2H4, C3H6, C3H8, C6H6) and tar throughout the bed in the emulsion and bubble phases velocity of the bubbles in the bed, diameter of bubbles depending on the height of the fluidized bed, the minimum fluidization velocity, superficial velocities in the bed and freeboard, mass flow of gases in the bed, mass flow of particles, void fractions in the bed and freeboard.

Research methodology
The experimental studies were carried out in three tests in which the CLC unit is operated stably.In each test, the velocity of the gas was changed in the air reactor.The study aimed to determine the effect of change of gas velocity in the air reactor on the basic parameters of the CLC system.Modeling the dynamics of fluidized bed for CLC process performed using a CeSFaMB software.Since the geometry is rather complex: Air Reactor (with the Circulation Fluidizing Bed approach) and Fuel Reactor (with the Bubbling Fluidizing Bed approach) were considered separately and boundary conditions between these two main parts of the system were used to link them.
The method of modeling CLC process using CeSFaMB simulator was presented in the papers [3,5].These articles show the results for the CLC cold model where the process is carried out at a temperature of 25 ºC.In contrast, this work presents the results for CLC hot model -at a temperature of 850 ºC.
The main operating parameters for these tests, including the information on the gas flux, pressure and  3.

Model validation
Numerical simulations were carried out taking into account the process conditions during experimental tests.The comparison and verification of the calculation and experiment results involves the following parameters: -main pressure losses, -solids recirculation flux, -void fractions, -superficial gas velocity, -gas mass flow rate, -bubble diameter, -rising velocity of gas bubbles.

Main pressure losses and solids recirculation flux
The average pressure losses across the beds for tests No. 1 -3, predicted by the model as well as the errors of the predictions are given in Table 4.The relative errors between experimental and calculated results are lower than 10 %.Total mass of solids recycled to the bed from the cyclone system in the main cycle equal 0.0026 kg s -1 , 0.0033 kg s -1 and 0.0034 kg s -1 , for tests No. 1, 2 and 3, respectively.The relative errors between experimental and calculated results are lower than 1 %.

Void fractions
The dense fluidized bed is formed in the lower part of the reaction chamber.By contrast, the lean region is located above the dense region and covers a predominant part of the chamber [11][12][13][14].
The comparison between experimentally determined and predicted by the model void fractions in dense region in the air reactor and fuel reactor is given in Figures 5 and 6.A similar comparison in lean region in the air reactor and fuel reactor is shown in Figures 7  and 8.
The relative errors between experimental and calculated results are lower than 12% for dense zone and 1% for lean region, respectively.

Superficial gas velocity and gas mass flow rate
The superficial gas velocity depends on the geometry of the reaction chamber whereas the gas mass flow rate is highly dependent on the process temperature [5].
The calculation results are shown in Figures 9  and 10.The maximum relative errors between experimental and calculated superficial gas velocities for the dense bed region of AR and FR do not exceed 2% and 5%, respectively.
The maximum relative errors of gas mass flow rate estimation are located within the range of ± 5% compared to the experimental data.This comparison is shown in the Figures 11 and 12

Bubble diameter and rising velocity of gas bubbles
In the bubbling and circulating fluidized bed the increase of bubbles diameter occurs with the height of the reaction chamber.Similarly, the rising velocity of gas bubbles increases with the height of the reaction chamber.Furthermore, the rising velocity of gas bubbles is dependent on the bubble diameter.These correlations are described by the following formulas [3,15]: ݀ = 0.430൫ܷ − ܷ ൯0.4(ℎ + 0.1272)0.8݃ି.ଶ (1) The comparison between experimentally determined and predicted by the model for bubble diameter and rising velocity of gas bubbles are shown in Figures 13-16.The calculated results do not exceed 10 % compared to the experimental data, for bubble diameter and velocity of a bubble rise.

Empirical models for basic dynamic parameters in CLC process
The successfully validated mathematical model allows to determine the influence of the superficial gas velocity on the following basic dynamic parameters of a CLC process [15]: solid circulating rate (GS) in the air reactor, number of a particle cycles in air and fuel reactor, particles residence time in air and fuel reactor, mass of solids in air and fuel reactor.
The influence of the superficial gas velocity on solid circulating rate is given in Figure 17.An increase in superficial gas velocity in air reactor causes an increase in the solids circulating rate.This dependence can be described by the following formula (3): The relations of the superficial gas velocity and number of cycles in air and fuel reactors are given in Figures 18 and 19.The analyzed relations is described in equations (4 and 5): The results of the reduction in residence time of particles both in air and fuel reactors are shown in Figures 20 and 21.This can be described by the following correlations: The dependences between superficial gas velocity and mass of solids contained in reactors are given in Figures 22 and 23.
The increase of the superficial gas velocity in air reactor causes the decrease in the total mass of solids contained in air reactor (Fig. 22) and an increase in the total mass in fuel reactor (Fig. 23).These behaviour are given in the forrmulas 8 and 9: ݉ ிோ = 0.51 ‫ݒ‬ ோ + 1.85

Summary and Conclusions
The first results of numerical simulations of the dynamic fluidized bed for the hot model of FB-CLC-SF unit are discussed in the paper.The simulations are carried out by the use of CeSFaMB Simulator.The standard operating conditions for the bubbling and circulating fluidized bed are used during the study.
A successfull validated model allows to carry out the simulation of system operation and determine the influence of the superficial gas velocity on the solids circulating rate, number of a particle cycles, particles residence time, as well as mass of solids in air and fuel reactor.
The increase of the superficial gas velocity in air reactor causes the increase in the solid circulating rate, number of cycles in both reactors, total mass in fuel reactor and decrease in the particles' residence time in both reactors, as well as total mass in air reactor.
The comparison of the experiment and calculations results involves the following parameters: main pressure losses, solids recirculation flux, void fractions, superficial gas velocity, gas mass flow rate, bubble diameter and rising velocity of gas bubbles.The validation of the model showed that the maximum relative error between the simulations and experiment results does not exceed 12%, which confirms the correctness of the CLC model.Therefore, the CeSFaMB simulator is a useful tool for the analysis of fluidization dynamics in CLC units.

Table 1 .Table 2 .Fig. 2 .
Fig. 2. The hot model ofFB-CLC-SF unit: (a) Photography of the FB-CLC-SF unit in the Institute of Advanced Energy Technology at the Czestochowa University of Technology; (b)The scheme of FB-CLC-SF unit[5]

13 Fig. 3 .
Fig. 3. Microscopic photograph of sand -research material used in the study

Fig. 4 .
Fig.4.The fluidized bed system considered by the CeSFaMB simulator[8] as the total mass of circulating solids are given in Table

Fig. 5 . 13 Fig. 8 .
Fig. 5. Comparison between desired and predicted void fractions in dense region of the air reactor .

Fig. 9 .
Fig. 9. Comparison between desired and predicted superficial gas velocity in dense bed region of air reactor

Fig. 10 .
Fig. 10.Comparison between desired and predicted superficial gas velocity in dense bed region of fuel reactor

Fig. 13 .
Fig. 13.Comparison between desired and predicted bubble diameter in dense bed of the air reactor

Fig. 15 .
Fig. 15.Comparison between desired and predicted rising velocity of bubble in the air reactor.

Fig. 16 .
Fig. 16.Comparison between desired and predicted rising velocity of bubble in the fuel reactor

Fig. 17 .
Fig. 17.Solid circulating rate versus superficial gas velocity in the air reactor

Fig. 18 .
Fig. 18.The number of cycles in air reactor versus superficial gas velocity in the air reactor

Fig. 20 . 13 Fig. 21 .
Fig. 20.The particles' residence time in air reactor versus superficial gas velocity in the air reactor

Fig. 22 .
Fig. 22.The total mass in air reactor versus the superficial gas velocity in the air reactor

1 Subscripts
d -diameter, m g -acceleration of gravity, m s -2 h -vertical coordinate, m Gs -solid circulating rate, kg m -2 s -1 m -mass of solids, kg tR -particles' residence time, s U -velocity, m s -, -nC -number of cycles in the reactors, cycles h -Reactor FB-CLC-SF -Fluidized-Bed Chemical-Looping -Combustion of Solid-Fuels, CeSFaMB -Comprehensive Simulator of Fluidized and Moving Bed equipment, CLC -Chemical Looping Combustion, CLOU -Chemical looping with Oxygen Uncoupling, FR -Fuel Reactor, iG-CLC -In-situ Gasification Chemical Looping.The Project "Innovative Idea for Combustion of Solid Fuels via Chemical Looping Technology" (Agreement No. POL-NOR/235083/104/2014) is funded from Norway Grants in the Polish-Norwegian Research Programme operated by the National Centre for Research and Development.CCS 2013 Call "New innovative solutions for CO2 capture".The support is gratefully acknowledged.

Table 4 .
Pressure loss across the bed,[Pa]