Three-dimensional modelling of circulating fluidized bed processes by a semi-empirical approach

Transcription

1 Lappeenranta University of Technology From the SelectedWorks of Kari Myöhänen May, 2012 Three-dimensional modelling of circulating fluidized bed processes by a semi-empirical approach Kari Myöhänen, Lappeenranta University of Technology Available at:

2 ERCOFTAC Spring Festival 2012 Three-dimensional modelling of circulating fluidized bed processes by a semi-empirical approach Kari Myöhänen Lappeenranta University of Technology LUT Energy May 10, 2012 Acknowledgements: The presentation contains material, which has been partially produced in co-operation with Foster Wheeler Energia Oy and VTT Technical Research Centre of Finland. Some material has been produced by or is related to work by Markku Nikku (LUT), Jarno Parkkinen (LUT), and Matti Koski (Oilon).

3 Introduction Circulating fluidized bed (CFB) processes are used for many purposes: Energy production (combustion, gasification, chemical looping combustion). Petrochemical processes (e.g. catalytic cracking). Other chemical and metallurgical processes. Trends in energy production CFB processes: increasing unit sizes, new processes, improved efficiencies. In CFB processes, the reactions occur in a suspension formed by gas and solids just modelling the fluid dynamics is very difficult. Optimal design, development, and operation of CFB processes requires comprehensive modelling tools, which can simulate the whole process: Flow dynamics of gases and solids Reactions Attrition of solid particles Heat transfer

4 Example CFB process: CFB boiler Primary air 2. Secondary air 3. Fuel, limestone, make-up feed 4. Refractory lined lower furnace 5. Furnace walls membrane walls 6. Internal heat transfer surfaces 7. Separator (cyclone) 8. Downcomer / return leg 9. External bubbling bed heat excanger 10.Cross-over duct 11.Backpass with heat exchangers 12.Electrostatic precipitator (ESP) 13.Stack

5 Development of CFB boiler scale 600 Samcheok, KR Power capacity (MWe) Pihlava, FI Tri-State, Nucla, US Duisburg, DE Kauttua, FI Lünen, DE Provence/Gardanne, FR Turow, PL Nova Scotia, CA NPS, Tha Toom, TH Emile Huchet, FR Kajaani, FI Jacksonville, US Ebensburg, US Baima, CN Seward, US Alholmen, FI EC Tychy, PL agisza, PL Year CN China DE Germany FI Finland FR France KR South-Korea PL Poland TH Thailand US United States

6 Example of process development: oxycombustion Oxygen Carbon diox. Water vapour Nitrogen Other Oxygen Carbon diox. Water vapour Nitrogen Other Flue gas Flue gas Fuel CFB boiler Fuel CFB boiler Air Oxygen Air-fired Oxygen-fired

7 Example of new processes: Caoling Carbon dioxide capture and storage process CaO+CO 2 CaCO 3 Interconnected dual fluidised beds Carbonator as absorber Calciner as regenerator CO 2 depleted flue gas CO 2 rich gas to compression Carbonator Calciner Coal CaCO 3 Flue gas from a combustor unit Oxygen from ASU and fluidizing gas Purge of ash and sintered solids

8 Model approaches The modelling approaches of the CFB process can be categorized to: Fundamentals-oriented CFD models (commercial codes, e.g. Fluent, open source codes, e.g. MFIX). Practice-oriented, semi-empirical models ("in-house" models). The recent development of the CFD models has been fast and recently published articles present examples of modelling full scale CFB units (e.g. Zhang et al., 2010; Wang et al., 2011). Simulated process times have been in the order of < 60 seconds. In a semi-empirical approach, fundamental balance equations (e.g. continuity of mass, energy, and species) are combined with empirical data and empirical correlations (e.g. total bed mass, solid concentration profile). This enables practical calculations, which can support the design of the processes. When the model has been validated based on field test measurements, it can be applied to study similar processes and cases with good accuracy. The range of applicability is limited by the available validation data Importance of reliable and detailed measurements!

9 Semi-empirical models The semi-empirical models (or engineering type models) are targeted in modelling the complete process, including the reactions and heat transfer. The different semi-empirical models can be categorized based on dimensionality: 0D-models: correlation based models for the total process. 1D-models: reactor divided to vertical sections. 1.5D-models (i.e. core-annulus approach). 3D-models: reactor divided to control volumes in three dimensions. The number of feeding points and the mixing of different reactants is limited, which results to a non-homogeneous process, which can be best modelled by 3D-models. Gas flow Solid flow Upper zone Annulus Core Bottom zone Core-annulus approach

10 Large furnaces - limited mixing of reactants agisza CFB 460 MWe

11 Comprehensive 3D-models for CFB Despite the clear demand for comprehensive 3D-models, the number of published and actively developed models is very small: 1) Model by Technical University Hamburg-Harburg First published in 1999 (Knoebig et al., 1999). Recent development: Wischnewski et al. (2010). 2) Model by Chalmers University of Technology First published in 2008 (Pallarès et al., 2008). Recent development: Palonen et al. (2011). 3) Model presented in this presentation. First version in 1989 (Hyppänen et al., 1991). Recent development: Myöhänen (2011).

12 Model frame of CFB3D Flue gas, fly ash u + v + Control volume method w v u Heat transfer to walls and internal surfaces Combustion, gasification & other reactions Gas, solids Exchange of gas / solids Solids Gas Separator(s) - separation eff. - heat transfer - reactions Solids External heat exchangers - heat transfer - reactions w + Inlet sources - sec. gas - fuel - limestone - sand Solids to furnace Fluidization gas Recirculation of flue gas / fly ash Fluidization gas Bottom ash

13 Concentration field of total solids The concentration field of total solids is defined by empirical equations. The parameters for vertical distribution are fitted based on measured pressure profiles. In horizontal direction, the solid concentration is assumed flat, except for a denser wall layer, which is solved as superimposed over the main furnace model. The volume fraction of solids at wall layer is determined as a function of the local average volume fraction of solids across the crosssection of the furnace.

14 Solution of velocity fields and mixing The net solid velocity field is solved by potential flow approach taking account for the different sources (e.g. circulating mass flow from return legs) and sinks (e.g. sinks due to reactions). The net gas flow field is solved by combining a continuity equation and a simplified momentum equation. The different mixing processes are modelled by dispersion. E.g. continuity equation for gaseous species: Convection Dispersion Sources Reactions Potential field and velocity field of solids

15 Particle size fractions and comminution of solids Solid materials (combustible fuel, ash, sand, limestone) are divided into six particle size fractions (population balance approach). Comminution of solids is simulated by a rate model, in which the mass change is proportional to the mass. Agglomeration of particles can be simulated as well, but normally the particle size is decreasing (mechanical wear, temperature shocks and effects of chemical reactions). Fraction 6 Fraction 5 Fraction 4 Fraction 3 Fraction 2 Fraction 1 Fragmentation Chipping Attrition Abrasion q m,c61 q m,c63 q m,c52 q m,c41 q m,c65 q m,c54 q m,c43 q m,c32 q m,c21 Agglomeration q m,c64 q m,c53 q m,c42 q m,c31 q m,c62 q m,c51

16 Combustion model Inert Evaporation H 2 O Example modelling results from bottom of a furnace Ash Moisture Devolatilization Volatiles Char Char combustion +O 2 Devolatilization Char combustion HCN, NH 3 H 2 S Char gasification +H 2 O, +CO 2 CO H 2 NO, N 2 O H 2 O CO, CO 2 Max CO, CO 2 CH 4, C 2 H 4 H 2 S SO 2 N 2 H 2 H 2 CO Gas combustion reactions CO + 0.5O 2 CO 2 Shift conversion CO + H 2 O CO 2 + H 2 H O 2 H 2 O CH 4 + 2O 2 CO 2 + 2H 2 O Min C 2 H 4 + 3O 2 2CO 2 + 2H 2 O H 2 S + 1.5O 2 H 2 O + SO 2

17 Mixing of fuel The mixing of reacting fuel (char, volatiles, moisture) can be simulated by two alternative methods: Target-dispersion model (old method, requires empirical knowledge of the target profile to which the fuel is mixing by dispersion). Convection model (new method, momentum balance solved for fuel). The new method is better suited for reactive and elutriative fuels, such as many biofuels. Illustration of differences in small scale: Model domain Target-dispersion, coarse d p Fuel convection, coarse d p Fuel convection, very fine d p Outlet Fuel feed (1 m/s) Grid air

18 Sorbent model for calcitic limestone CaCO 3 CO 2 Modelled sulphur dioxide profile in oxygen-fired combustion CO 2 CaO CaSO 4 Carbonation SO 2 +CO 2 Desulphation CaCO 3 CaO

19 Validation case: air-fired 235 MWe CFB burning lignite Modelled vs. measured profiles Oxygen Concentration (%-dry) Furnace Width (m) Meas. 2.5 m Meas. 2.0 m Meas. 1.1 m Meas. 0.7 m Model 2.5 m Model 2.1 m Model 1.0 m Model 0.6 m Heat flux Oxygen CO Uneven fuel feed distribution

20 Calculation example: Compostilla 300 MWe A planned circulating fluidized bed boiler for flexible operation with airfired and oxygen-fired combustion. The limestone reactions are different in oxygen-fired conditions due to high partial pressures of CO 2 and SO 2. Temperature Sulphur capture Air-fired Oxygen-fired Air-fired Oxygen-fired

21 Gasifier study Thermal capacity: 50 MWth Air-blown, atmospheric gasification Fuel: wood based biomass 49.8% volatiles, 9.0% char 40.0% moisture, 1.2% ash Geometry Outlet to cyclone Surface mesh coloured by velocity magnitude Shift conversion CO + H 2 O CO 2 + H 2 Sorbent: calcitic limestone Geometry approximated by hexahedral control volumes, mesh size cells The 3D modelling can be applied to study the progress of reactions along the furnace height and to support optimal design. Fuel, limestone secondary air Fuel, sand sec. air Return leg Grid air

22 3D-modelling of a pilot calcinator for CaOling Calcination process: CaCO 3 CaO+CO 2 Gas and solids to separator CO 2 depleted flue gas CO 2 rich gas to compression Carbonator Calciner Coal CaCO 3 Flue gas from a combustor unit Oxygen from ASU and fluidizing gas Purge of ash and sintered solids Fuel and make-up sorbent feed Return leg Feed from carbonator Fluidization gas Bottom ash

23 Discussion and summary The development of the CFB processes requires valid modelling tools, which can be used to optimize the design and operations. The application of the currently available CFD models is still limited. Only a few comprehensive process models exist, which can model large circulating fluidized bed reactors three-dimensionally. The presented model has been successfully applied for studying different CFB processes. The development of the model is a continuous process: the different sub-models can be always improved as more knowledge is achieved.

24 References Hyppänen, T., Lee, Y.Y., Rainio, A. (1991). A three-dimensional model for circulating fluidized bed boilers. Proc. 11th Int. Conf. on Fluidized Bed Combustion. Knoebig, T., Luecke, K., Werther, J. (1999). Mixing and reaction in the circulating fluidized bed - A three-dimensional combustor model. Chem. Eng. Sci., 54, pp Myöhänen, K. (2011). Modelling of combustion and sorbent reactions in three-dimensional flow environment of a circulating fluidized bed furnace Ph.D. thesis. Lappeenranta University of Technology. Nicolai, R., Goedicke, F., Tanner, H., Reh, L. (1993). Particle induced heat transfer between walls and gas-solid fluidized beds. Proc. 4th Int. Conf. on Circulating Fluidized Beds. Pallarès, D., Johnsson, F., Palonen, M. (2008). A comprehensive model of CFB combustion. Proc. 9th Int. Conf. on Circulating Fluidized Beds. Palonen, M., Pallarès, D., Ylä-Outinen, V., Larsson, A., Laine, J., Johnson, F. (2011). Circulating fluidized bed combustion build-up and validation of a three-dimensional model. Proc. 10th Int. Conf. on Circulating Fluidized Beds. Weng, M., Nies, M., Plackmeyer, J. (2011). Computer-aided optimisation of gas-particle flow and combustion at the Duisburg circulating fluidized bed furnace. VGB PowerTech 8, pp Wischnewski, R., Ratschow, L., Hartge, E.-U., Werther, J. (2010). Reactive gas-solids flows in large volumes - 3D modeling of industrial circulating fluidized bed combustors. Particuology, 8, pp Zhang, W., Johnsson, F., Leckner, B. (1993). Characteristics of the lateral particle distribution in circulating fluidized bed boilers. Proc. 4th Int. Conf. on Circulating Fluidized Beds. Zhang, N., Lu, B., Wang, W., and Li, J. (2010). 3D CFD simulation of hydrodynamics of a 150 MWe circulating fluidized bed boiler. Chemical Engineering Journal, 162, pp