Study on scale-up characteristics of oxy-fuel combustion in circulating fluidized bed boiler by 3D CFD simulation

The oxy-fuel combustion CFB technology as a promising carbon capture technologies needs to study the scale-up process for the commercial diffusion. Numerical simulation would be a rational tool to investigate the gas-solid flow and oxy-fuel combustion process before constructing an expensive and complicated industry-scale plant. A three-dimensional (3D) CFD simulation according to the Eulerian-Lagrangian approach was applied to simulate the hydrodynamics of gas-solid flow and oxy-fuel combustion process in lab-scale, pilot-scale and industry-scale CFB boiler (from 0.1 MW_th to 330 MW_e). Results present the differences of the boiler configuration, the gas-solid flow and the oxy-fuel combustion characteristics between lab-scale, pilot-scale and industry-scale CFB boilers. The cross-section thermal load gradually decreased, while the cross-section area increased with the thermal inputs increased. In the lab-scale and pilot-scale oxy-fuel CFB, the particle velocity field was more uniform than that in the industry-scale CFB. The carbon conversion ratio increased with an increase in the thermal input. The emission of CO, NO and SO₂ in the industry-scale oxy-fuel CFB boilers was lower than those in the lab-scale and pilot-scale. A larger oxy-fuel combustion power plant is beneficial to carbon capture and low pollutant emission. The results would be beneficial to the design and operation of industry-scale oxy-fuel CFB.     

Radiative heat transfer in splash and dilution zones of the pressurized oxy-fuel combustion CFB considering particle-dependent scattering

Radiative heat transfer is dominant in splash and dilution zones of the oxy-fuel combustion pressurized circulating fluidized bed (PCFB). However, the particle radiation fields interact with each other due to the high particle concentration. Moreover, the particle concentration distribution, the flue gas components and pressure change drastically due to pressurized oxy-fuel combustion technology. In this paper, the radiative heat transfer in splash and dilution zones of the oxy-fuel combustion PCFB is investigated when particle-dependent scattering is considered. The results show that the maximum error of the particle radiation model is less than 1% in predicting the incident heat flux at the wall when particle-dependent scattering is considered. The particle concentration distribution has a significant impact on the radiative heat transfer in splash and dilution zones of oxy-fuel combustion PCFB. The simplified particle concentration distribution model cannot capture the zero-source term phenomenon in the core region and has a large error in the annular region. In addition, the mechanism and importance of the flue gas composition and pressure on the radiative heat transfer in splash and dilution zones of oxy-fuel combustion PCFB are analyzed.     

Simulation of co-gasification of coal and wood in a dual fluidized bed system

A comprehensive three-dimensional (3D) model based on multiphase particle-in-cell (MP-PIC) approach is developed to simulate the co-gasification process of coal and wood in a dual fluidized bed (DFB) system. The pyrolysis of coal and wood, formation and conversion of tar and gaseous pollutants, as well as gasification and combustion reactions are integrated with dense gas-solid flow and heat transfer. The pilot DFB system comprises a bubbling bed gasifier and a fast fluidized bed combustor connected by loop seals. The model correctly predicts profiles of pressure and temperature, the yield and components of the product gas. The gas composition distributions, allocations of particle mass, and solid residence time inside the reactors are explored. The effects of various fuel blend ratios and particle sizes on gasification performances are also investigated. The results show that increasing coal ratio accelerates the steam gasification due to higher char content, which results in the increment of H₂ and CO concentrations. The tar content in the product gas continues to decline, while the emissions of NH₃ and H₂S increase. The size variation of feedstock is not enough to dramatically affect product gas components. The tar content and product gas yield appear a slight upward trend with the smaller size. The variations of NH₃ and H₂S concentrations are consistent with those of bed temperature.     

Recent advances in high-fidelity simulations of pulverized coal combustion

Although renewable energy has recently attracted a lot of attention, coal-fired power plants will still play an important role in electricity production in the future due to coal’s abundance in the world. However, combustion of pulverized coal is extremely complex which makes experimental measurements very challenging even impossible. As an alternative, computational fluid dynamics (CFD) is a powerful tool to investigate pulverized coal combustion (PCC), and can help to design and develop advanced coal-fired facilities. With the development of computational capacity, high-fidelity simulations of PCC have been developed and significant achievements have been obtained in the recent years. In this review, the recent advances in CFD of PCC, especially in fully-resolved direct numerical simulation (DNS), point-particle DNS and large eddy simulation (LES) of PCC are summarized. The progresses on both numerical methods and coal combustion models are highlighted. The state-of-the-art applications of these high-fidelity simulations are also reviewed, including the investigation of combustion characteristics of pulverized coals, the development of advanced coal combustion models, and the development of clean coal technologies. It is concluded that fully-resolved DNS, point-particle DNS, and LES have their merits and demerits, so the optimal use of each approach depends on specific research purpose. For coal devolatilization and char-oxidation models, the combined models in which advanced models are used to calibrate the kinetic parameters of simple models are quite attractive in terms of computational accuracy and computational burden. For chemical reactions, the flamelet model is popular, but the predictive capability and robustness need to be further improved. Besides, LES is rapidly approaching to simulate industrial PCC. However, the computational cost is still quite expensive, and more efficient algorithms and sub-grid models are required. To this end, DNS database, benchmark measurements, and international collaborations are important. Machine learning may also play an important role in promoting the development of high-fidelity simulations of PCC in the future.     

Experimental study of fluidization characteristics of Geldart-D particles in pressurized bubbling fluidized bed

The pressurized bubbling fluidized bed shows great advantage in retreating municipal solid waste because it could effectively capture CO₂ and enhance the reaction rate of the process of combustion and gasification. In the present work, fluidization characteristics of Geldart-D particles at elevated pressure were experimentally investigated, such as flow pattern, pressure drop, minimum fluidization gas velocity. At the same fluidization gas velocity, as elevating operating pressure, the fluidization of Geldart-D particles became more intense, the bubbles got larger, the standard deviation and the power density of dominant frequency of the pressure drop signal increased. While, under the same fluidization number, as increasing operating pressure, the fluidization of Geldart-D particles became smoother, the bubble size decreased, both the standard deviation and the power density of dominant frequency of the pressure drop signal decreased. It seems that, under elevated pressure, the fluidization behavior of Geldart-D particles would transition to that of Geldart-B particles. Finally, the minimum fluidization velocity of the Geldart-D particles was found decreased with the increase of the operating pressure. A new correlation for the prediction of the minimum fluidization velocity of Geldart-D particles at elevated pressure was also formulated based on the present experimental results.     

Simulation on coal-fired supercritical CO2 circulating fluidized bed boiler: Coupled combustion with heat transfer

Using supercritical carbon dioxide (S-CO₂) as the working fluid integrated in a circulating fluidized bed (CFB) boiler is a rising technology used to improve the power generation efficiency and reduce gas pollutant emissions in coal-fired power generation systems. This study established a comprehensive 3-D model based on an Eulerian-Lagrangian frame to simulate the combustion process. A new method was presented using constant heat flux as the boundary obtained from the coupled simulation of heat transfer and combustion. The gas phase was described with large eddy simulation (LES). The solid phase used the multi-phase particle-in-cell (MP-PIC) approach. Simulations were carried out in a 10 MW S-CO₂ CFB boiler (with cross section area of 3.557 × 3 m² and height of 21.01 m). Combustion characteristics obtained in boundary heat flux and excess air ratio were numerically investigated. Results showed that the temperature profile was relatively uniform in the whole boiler and the furnace temperature increased with the increase of boundary heat flux. Emissions of CO₂ and SO₂ declined with the increase of boundary heat flux while CO emission increased. An increased excess air ratio caused a decrease in furnace temperature and the rise of CO and SO₂. The characteristics of combustion and pollutant emissions were optimal with the heat flux at around 25–37 kW/m² and an excess air ratio at 1.18–1.25.     

Simulation of coal pressurized pyrolysis process in an industrial-scale spout-fluid bed reactor

A three-dimensional Eulerian-Lagrangian model, facilitated with multiphase particle-in-cell (MP-PIC) method, was developed to simulate gas-solid flow and pyrolysis characteristics of coal (with the capacity of 500 thousand tons per year) in an industrial-scale spout-fluid bed reactor (H = 16.6 m and D = 3.1 m), aiming at providing guidance for industrial application of pressurized grade conversion of coal. The performance of the reactor and the effects of operating parameters such as coal feeding rate, semi-coke to coal ratio, and particle sizes were numerically investigated. It was found that the flow pattern in this case is a “jet in the fluidized bed with bubbling”. The raise of pressure has a positive impact on the spouting structure and the flow uniformly. The increase of the semi-coke to coal ratio is beneficial to the coal pyrolysis, but the improvement of the pyrolysis is limited and the number of particles in the reactor will be sharply increased. With the increase of particle sizes, the flow pattern in the pyrolysis reactor tends to be stable while the mixing effectiveness is getting worse. It is suggested that the particle size of the material should range within 0–6 mm.     

Partitioning behavior of trace elements during pilot-scale fluidized bed combustion of high ash content lignite

This study describes the partitioning of 20 trace elements (As, B, Ba, Cd, Co, Cr, Cu, Hg, Li, Mn, Mo, Ni, P, Pb, Sb, Se, Sn, Tl, V, Zn) and eight major and minor elements (Al, Ca, Fe, K, Mg, Na, Si, Ti) during the combustion of high ash content lignite. The experiments were carried out in the 0.3 MW(t) Middle East Technical University (METU) atmospheric bubbling fluidized bed combustor (ABFBC) test rig with and without limestone addition. Inert bed material utilized in the experiments was bed ash obtained previously from the combustion of the same lignite without limestone addition in the same test rig. Concentrations of trace elements in coal, limestone, bottom ash, cyclone ash and filter ash were determined by inductively coupled plasma optical emission spectroscopy (ICP-OES). Partitioning of major and minor elements are influenced by the ash split between the bottom ash and fly ash and that the major proportion of most of the trace elements (As, Ba, Cr, Hg, Li, Mo, Ni, Sn, V, Zn) are recovered in fly ash. Limestone addition shifts the partitioning of Ba, Cr, Mo, Ni, Sn, V, Zn from bottom ash to fly ash.     

Numerical simulation of a 3-D gas-solid fluidized bed: Comparison of TFM and CPFD numerical approaches and experimental validation

This paper presents the results of a 3-D numerical simulation of a freely bubbling fluidized bed, based on the Eulerian–Lagrangian approach, using the software Barracuda (CPFD-Barracuda). The main results obtained were assessed in terms of frequency analysis, bubble pierced length, bubble size, bubble passage frequency and bubble velocity. The results obtained were also compared with experimental data obtained in a 3-D fluidized bed using pressure and optical probes, and with the numerical results using the more common Eulerian-Eulerian approach, implemented in the commercial software Fluent (TFM-Fluent).The results show that CPFD-Barracuda satisfactorily predicts the global behaviour of bubbling beds with a low computational cost, although it computes smaller bubble sizes and lower bubble velocities than TFM-Fluent and experiments. Additionally, the spectra of pressure and particle volume fraction obtained with CPFD-Barracuda resemble those from the experiments and the TFM-Fluent simulations, but with a larger contribution of lower frequencies. The peaks of the pressure spectra from CPFD-Barracuda are close to those from the experiments and the TFM-Fluent simulations, whereas those in the solid volume spectra seem to be underestimated by CPFD-Barracuda. The results also indicate that the particle fraction threshold value chosen to distinguish bubbles contours notably influences the results of the bubble characteristics, especially for TFM-Fluent, whereas CPFD-Barracuda is less sensitive to this threshold value.     

Investigation of non-uniform characteristics in a 300 MWth circulating fluidized bed with different coal feeding modes

Circulating fluidized bed (CFB) has been practised in many engineering fields, but the non-uniform characteristics deteriorate combustion performance and system control. In this study, the improvement of external-loop and in-furnace non-uniformity of a 300 MW_th industrial-scale CFB with multiple cyclones by a dual-side coal feeding mode was numerically quantified. The results show that the pressure is non-uniformly distributed among three external loops, where the pressure in the middle loop seal is lower than that in the corner loop seals by 4.5 %. The pressure gradient positively correlates with the solid holdup. As compared with the traditional single-side coal feeding mode, the dual-side coal feeding mode: (i) promotes the final mixing degree by 11.15 %; (ii) extends the residence time of coal particles by 10.3 %; (iii) reduces the magnitude and fluctuation of the horizontal solid flux by 28.5 % and 77%, respectively; (iv) narrows the temperature range and reduces the mean temperature by 7 °C; (v) enhances the combustion of coal particles by consuming more O₂; (vi) decreases the concentrations of SO₂ and NO by 3 % and 5 %, respectively. The present study shows the superiority of the dual-side coal feeding mode to the traditional single-side coal feeding mode in the optimization of CFBs in practical industrial processes.     

Effect of circulating fluidized bed combustion ash on the properties of roller compacted concrete

Circulating fluidized bed combustion (CFBC) ash, which has such a high content of f-CaO and SO_3, is a waste or by-product of petroleum coke combustion power stations. The purpose of this study is to investigate the effects of CFBC ash on the properties of roller compacted concrete (RCC). CFBC ash was used to replace fine aggregate with various dosages (5%, 10% and 15%) by weight. All mixtures were designed according to ACI 211.3R and prepared for testing. During casting, cylinders were vibrated and compacted with different pressures of 25 g/cm², 50 g/cm² and 75 g/cm², respectively. Test results show that CFBC ash can increase the water absorption and effectively reduce the initial surface absorption. Meanwhile, CFBC ash has a positive effect on compressive strength, splitting tensile strength, and sulphate attack resistance of hardened RCC. SEM revealed that the main hydration products of specimens containing CFBC ash are AFt (ettringite), C–S–H (hydrated calcium silicate) and portlandite. Based on the presented observations and results, RCC with the dosage of 5% CFBC ash as fine aggregate replacement and the roller compaction pressure of 75 g/cm² is recommended.     

Development and combustion characteristics of microwave plasma-assisted fluidized bed combustor

A novel waste treatment method that can efficiently decompose waste, suppress by-product generation, and operate at a low cost is urgently required. Herein, a microwave plasma-assisted combustor was developed and its combustion characteristics were investigated for application in solid waste treatment. The experimental conditions for obtaining good states of fluidization, mixing, and plasma formation were examined prior to the combustion experiments. Subsequently, the optimal experimental conditions, such as the filling amount of bed particles, bed particle diameter, microwave irradiation position, and microwave output, required for good combustion were achieved. Combustion experiments based on these conditions revealed that a good fluidization state is required to obtain a good combustion state in this device, although the combustion condition does not necessarily depend on the system pressure at each O₂ flow rate. Comparison of the conditions with similar fluidization states at O₂ flow rates of 1–4 L/min revealed a maximum fuel conversion ratio at 4 L/min owing to the combustion promotion caused by increased O₂ partial pressure. The fuel did not remain in the fluidized bed combustor after the combustion experiments.     

Characterization of low chromium ferritic steels for use in fluidized bed combustion systems

There is evidence from certain materials evaluation studies that suggests that low chromium ferritic steels are susceptible to oxide scaling at a gr:ater rate than would have been expected if the steels had been expectsed m a. conventional coal-fired combustor. Such a result is of economic sigmfwance since it tmplles that the changeover point in a superheater from low alloy ferritic to austenitic tubing would have to be at a lower metal temperature than would otherwise be necessary, with consequent Cast penalties. There were, however, some anomalies in the data because of uncertamty over the precise metal temperatures of the steels included in the studtes. A programme of work was therefore established to resolve this problem.

Ferritic steel samples were exposed at carefully controlled metal tempe.ratures under steady state conditions in an AFBC test rig. The scaling characteristics of these samples were compared with data produced under controlled. condttwns m atr. In contrast to earlier, less closely defined tests, the results of thts present study suggest that the oxidation rates of 2.25 wt% Cr-lwt% Mo steel are similar to those expenenced in air and in conventional combustion systems. On this basis the upper temperature limit for use of this steel for in-bed superheater tubes would be set at 560°C. In contrast 9 wt% Cr-l wt% Mo steel showed enhanced oxidation rates, greater than those found in either air or conventional combustion systems, and would appear to offer no advantage over 2.25 wt% Cr-lwt% Mo steel. The possible impact on these recommendations for systems where high chlorine coals will be burned, or where erosion-corrosion effects are likely, is discussed.     

Multi-fluid modelling of hydrodynamics in a dual circulating fluidized bed

In this work, a multi-fluid model based on the Eulerian-Eulerian framework is used to study the gas-solid hydrodynamics, such as solid distribution, particle motion and solid velocity, in a three-dimensional (3D) dual circulating fluidized bed (DCFB). The influence of four different drag force models, including two classic models, i.e. Gidaspow, EMMS drag model and two recent drag models, i.e. Rong and Tang drag model, on hydrodynamics in DCFB are assessed. Numerical results show that the characteristics of solid distribution and velocity in different sections are distinct. For qualitative analysis, all the drag models can predict a reasonable radial solid distribution and pressure distribution, but only the EMMS, Rong and Tang drag model can capture the phenomenon of dense solid concentration in the low part. For quantitative analysis, the solid circulating rate predicted by the EMMS drag model is the closest to the experimental value while the Gidaspow drag model shows the most significant deviation. The overall assessments confirm that the drag model selection has a significant influence on the simulations of gas-solid flow in DCFBs. This study sheds lights on the design and optimization of fluidized bed apparatuses.     

Simulations on pressurized oxy-coal combustion and gasification by molecular dynamics method with ReaxFF

Pressurized oxy-fuel combustion (POFC) is recognized to have the potential to effectively capture CO₂ with low cost and high efficiency. To investigate the chemical mechanism of oxy-fuel combustion and the effects of different operating parameters on combustion characteristics, a gasification reaction model and the pressurized oxy-fuel combustion model were constructed. A series of reactive molecular dynamics (MD) simulations were conducted on the POFC model using ReaxFF force field under the pressure of 0.2, 0.3 and 0.5 MPa, and with the temperature at 1600, 1800 and 2000 K, respectively. The activation energy for atmospheric oxy-fuel combustion was firstly calculated, in agreement with reported experimental results, which verified the accuracy of ReaxFF MD method. The results of gasification and combustion showed that both temperature and pressure positively affect coal decomposition and the combustion reaction rate. The conversion mechanism of C in coal to CO₂ is a process with dehydrogenation, coal decomposition and oxidation reaction. Compared with atmospheric oxy-fuel combustion, the increase of pressure/density would accelerate the dehydrogenation reaction and the decomposition of coal structure, and improve the performance of coal combustion and promote the decomposition of coal molecule into smaller fragments, and further promote the releasing of CO₂ and small fragments.     

Prediction of segregation behavior of binary mixture in a pulsed fluidized bed

A three-dimensional simulation is carried out to investigate the impact of pulsation flow on segregation behavior of binary mixture in a fluidized bed via the multi-fluid model. The simulated results are compared against the experimental data. The impacts of pulsation frequency and operating condition on flow and segregation behavior of binary mixture are discussed. The results reveal that an excessive increase of pulsation frequency and operating temperature is not beneficial to the enhancement of segregation efficiency. The pulsation frequency plays a more important role in segregation efficiency under a small size discrepancy of binary mixture. Meanwhile, the effect of pulsed air flow waveform on segregation efficiency of particles is also analyzed.     

The distribution of heavy metals during fluidized bed combustion of sludge (FBSC)

During combustion of wastewater treatment sludge, the inorganic constituents are converted into ash which contains the major fraction of the heavy metals present. The behaviour of heavy metals in combustion processes has been studied extensively for mostly coal combustion and waste incineration. For biomass and sludge, literature data are scarce and mostly limited to laboratory experiments. The present paper assesses the partitioning of eight heavy metals (Hg, As, Cd, Cu, Pb, Cr, Ni and Zn) in the different residues from a large-scale fluidized bed sludge combustor of 4.4 m i.d. The origin of the sludge is mostly from treating urban wastewaters (>90%), although some mixed sludge (urban+industrial, <10%) is also burnt. The different residues (bottom ash, fly ash, filter cake, scrubber effluent and stack emissions) were sampled and analysed during 33 weeks, spread over a period of 1 year. The mass balance of relevant heavy metals closes for 96.5%, inaccuracies being related to the unsteadiness of the process, the accuracy of the mass flow data monitored at the plant, and on collecting representative samples. It is also shown that all heavy metals under scrutiny, except Hg, are concentrated in the fly ash as collected in the electrostatic precipitator.     

Experimental and numerical studies on a bubble-induced inverse gas-liquid-solids fluidized bed

Hydrodynamics in a newly invented bubble-induced inverse gas–liquid-solids three-phase fluidized bed has been studied via both experimental and numerical methods. With experiments in a 3.0 m column of 0.153 m in diameter, four fluidization regimes including a fixed bed regime, a bed expansion regime, a complete fluidization regime, and a freeboard regime have been identified with the increase in the superficial gas velocity. A three-phase Eulerian-Eulerian CFD model was developed to simulate the hydrodynamics in the inverse three-phase fluidized bed and the simulation results have a good agreement with the experimental data. The effects of the particle property and solids loading on the transitions across the flow regimes were numerically studied. A higher solids loading and/or a larger particle density are reported to contribute to an easier fluidization and a faster flow development to the complete fluidization regime. The radial flow structure becomes less uniform with increased inner circulation of the liquid after introducing more bubbles into the column.     

Granular temperature with discrete element method simulation in a bubbling fluidized bed

The Discrete Element Method (DEM) plays an important role in understanding and modeling the kinetic characteristics in granular systems. A soft-sphere method with a linear spring–dashpot model was used in the simulation of a bubbling fluidized bed. The time-averaged granular temperature and vertical particle velocity at different heights were numerically studied and compared to experimental measurements of Müller. The influence of a velocity-dependent coefficient of restitution and three drag models were also investigated in this work. Good agreement was found between the DEM simulation and Müller’s experiment, especially using the DiFelice drag model. The variable coefficient of restitution, with a sufficiently high yielding relative velocity, gives a granular temperature that is a little lower compared to that of a constant coefficient of restitution, while it predicts a more intense velocity fluctuation, with a lower yielding relative velocity. By comparing the granular temperature in the vertical direction and in the transverse direction, a strong anisotropy is found in the bed.