CFD simulation of 1 MW th Carbonator using DDPM-DEM model

Transcription

1 Energy Systems and Technology Prof. Dr.-Ing. B. Epple Otto-Berndt-Str Darmstadt / Germany Phone: CFD simulation of 1 MW th Carbonator using DDPM-DEM model 6 th High Temperature Solid Looping Cycles Network Meeting Alexander Stroh, M. Sc., Dr. Jochen Ströhle, Prof. Bernd Epple , Milan

2 SCARLET Consortium Alexander Stroh CFD simulation of 1 MW th Carbonator using DDPM-DEM model 1

3 Outline Introduction Carbonate-Looping-Process Discrete element method (DEM) model Experimental & numerical boundary conditions Results & model validation Influence of parcels number Drag-models and bed inventory on time averaged pressure profile along reactor axis Carbonation reaction implementation Outlook Alexander Stroh CFD simulation of 1 MW th Carbonator using DDPM-DEM model 2

4 Carbonate-Looping-Process Alexander Stroh CFD simulation of 1 MW th Carbonator using DDPM-DEM model 3

5 DDPM-DEM model description Collision modelling of DDPM-DEM approach DDPM can be used for any granular flow (e.g. fluidized bed, hoppers, pneumatic solid transport applications) Explicit particle tracking using Discrete Element Method (DEM) based on the Euler-Lagrange approach Soft-sphere contact model resolve particle-particle collisions Computational expensive for more than 500,000 particle parcels and small particle step size Model simplification Particles are represented by spherical parcels Parcel collisions account for several particles with specific mass and volume Collision force laws (spring, spring-dashpot, friction) Alexander Stroh CFD simulation of 1 MW th Carbonator using DDPM-DEM model 4

6 Experimental & numerical boundary conditions Temperature and Inventory assumed to be constant during steady-state operation Isothermal Temperature Inventory 908 [K] 220, 240, 260, 280 [kg] Mass flow (kg/s) Species Boundary Boundary type Air Air + CO2 inlet Mass flow inlet Carbon dioxide Air + CO2 inlet Mass flow inlet Air Solid inlet Mass flow inlet - - Outlet Pressure outlet Alexander Stroh CFD simulation of 1 MW th Carbonator using DDPM-DEM model 5

7 Ansys DDPM-DEM modeling Assumption bed mass constant over time Particle injection implemented through User-Defined-Function Outlet Cooling lances Parcel number N parc > 200,000 Parcel diameter d parc < m Particle diameter d p = 91 µm Drag models DEM collision model Fluid flow time step size Particle time step size Syamlal O Brien, Gidaspow, Gibilaro, Wen&Yu, EMMS Spring dashpot for normal forces & friction-dshf model for tangential forces <1e-3 s <2e-4 s 1. Determine bed inventory in each time step? 2. Inject particles through surface injection using DEFINE_DPM_INJECTION_INIT macro Numerical mesh investigation structured coarse grid with 31,207 cells & fine unstructured grid with 91,031 cells Injection area Alexander Stroh CFD simulation of 1 MW th Carbonator using DDPM-DEM model 6

8 Height (m) Solid mass flow rate (kg/s) Ansys DDPM-DEM modeling influence of bed inventory Syamlal O'Brien drag model kg solid inventory 240 kg solid inventory 260 kg solid inventory Increasing solid flux Increasing inventory Pressure (Pa) Alexander Stroh CFD simulation of 1 MW th Carbonator using DDPM-DEM model 7

9 Height (m) Ansys DDPM-DEM modeling influence of drag models 6.5 Drag models comparison using 220 kg of bed inventory Drag overprediction Syamlal O'Brien WEN & YU Pressure (Pa) Gidaspow Gibilaro EMMS[1] Alexander Stroh CFD simulation of 1 MW th Carbonator using DDPM-DEM model 8

10 Height (m) Ansys DDPM-DEM modeling influence of parcels number Pressure profile along carbonator axis Experiment EMMS 300,000 parcels Gidaspow 200,000 parcels Gidaspow 300,000 parcels Gidaspow 500,000 parcels Pressure (Pa) Alexander Stroh CFD simulation of 1 MW th Carbonator using DDPM-DEM model 9

11 Ansys DDPM-DEM reaction modeling Applied reaction rate expression according Romano [2] dx = k dt ss n 1 X 2 3 (CCO2 C CO2,eq) with S n = V MCaCO3 X Maxρ CaO M CaO h intrinsic rate constant specific available surface area Applied constants from Abanades et al. [2,4] k s (m 4 /smol) 6.05*10-10 According Charitos et al. [3] dx dt = k ss 0 X max X 2 3 (CCO2 C CO2,eq) V MCaCO3 (m³/mol) 36.9*10-6 ρ CaO (kg/m³) 3320 h (m) 50* Alexander Stroh CFD simulation of 1 MW th Carbonator using DDPM-DEM model 10

12 X (conversion degree) Ansys DDPM-DEM reaction modeling Comparison of reaction rates Romano & Charitos C CO2 - C CO2eq ~ mol/m³ X MAX = 0.10 Charitos Romano Time (s) Alexander Stroh CFD simulation of 1 MW th Carbonator using DDPM-DEM model 11

13 Ansys DDPM-DEM reaction modeling Single particle experiment - Injected parcel with kg - Particle velocity set to zero through UDF - Gas inlet velocity 10 m/s - Different CO 2 /N 2 concentrations Recording carbonation degree over time in order to validate reaction model Alexander Stroh CFD simulation of 1 MW th Carbonator using DDPM-DEM model 12

14 Ansys DDPM-DEM reaction modeling Alexander Stroh CFD simulation of 1 MW th Carbonator using DDPM-DEM model 13

15 CO 2 (area weighted average kg/kg) Ansys DDPM-DEM reaction modeling 0.06 CO 2 concentration at Carbonator outlet Relative deviation of mean values ~ 10% Simulation Mean value from Experiment Time (s) Alexander Stroh CFD simulation of 1 MW th Carbonator using DDPM-DEM model 14

16 Ansys DDPM-DEM reaction modeling - Local particle velocities > 4 m/s - dense region underpredicted with conventional drag models - CO 2 gas concentration locally different - at the outlet ~3 % [kg CO 2 /kg Gas] (2.2 vol. %) - CO 2 mainly captured in the bottom zone of Carbonator Alexander Stroh CFD simulation of 1 MW th Carbonator using DDPM-DEM model 15

17 Outlook Further 3-D model validation with new designed cold flow experiments Improve implemented reaction model Extend reaction model for particle size classes Validation with pilot plant data for different experimental conditions Drag model development, implementation of filtered drag models in DDPM-DEM Sensitivity analysis (e.g. Bed inventory influence on capture efficiency, Make-Up Flow, PSD) Full-Loop simulation and experimental validation with 1MW th plant Alexander Stroh CFD simulation of 1 MW th Carbonator using DDPM-DEM model 16

18 The end Thank you for your attention. Contact M.Sc. Alexander Stroh L01/01 room 342 Tel. (06151) Alexander Stroh CFD simulation of 1 MW th Carbonator using DDPM-DEM model 17

19 References [1] Nikolopoulos, A., et al. (2010). "An advanced EMMS scheme for the prediction of drag coefficient under a 1.2 MW th CFBC isothermal flow Part I: Numerical formulation." Chemical Engineering Science 65(13): [2] Romano MC. Modeling the carbonator of a Ca-looping process for CO 2 capture from power plant flue gas. Chemical Engineering Science. 2012;69: [3] Charitos A, Hawthorne C, Bidwe A, Sivalingam S, Schuster A, Spliethoff H, et al. Parametric investigation of the calcium looping process for CO 2 capture in a 10kW th dual fluidized bed. International Journal of Greenhouse Gas Control. 2010;4: [4] Abanades JC, Anthony EJ, Lu DY, Salvador C, Alvarez D. Capture of CO2 from combustion gases in a fluidized bed of CaO. AIChE Journal. 2004;50: Alexander Stroh CFD simulation of 1 MW th Carbonator using DDPM-DEM model 18