Numerical modelling of the large-scale Virtuhcon Benchmark for non-catalytic natural gas reforming Authors: Yury Voloshchuk, Andreas Richter, Bernd Meyer Institute of Energy Process Engineering and Chemical Engineering TU Bergakademie Freiberg, Germany 1
Motivation For HP POX processes, CFD allows for an accelerated development of new technologies HT/HP processes require intensive validation of the numerical model In literature, experimental data for HP POX systems are typically based on lab-scale (simplifications and limitations necessary) Results away from real process To overcome this limitation, large-scale experiments for HP POX were performed at TU Bergakademie Freiberg (Virtuhcon Benchmark) Comprehensive data set was published in Fuel 2015 (all data for CFD modelling were provided) Operating conditions were selected in such a way that it stresses the chemical model (outlet conditions away from equilibrium) CFD modelling of the Virtuhcon Benchmark to test the accuracy of state-of-the-art CFD models 2
CONTENT Large-Scale Experiments (Virtuhcon Benchmark) CFD Model Numerical Results and Validation Conclusion 3
Pilot plant in IEC, Freiberg* System specifications max. pressure: 100 bar temperature: 1000-1500 ºC thermal power: 5 MW adjustable reactor volume Feedstock natural gas light or heavy fuel oils residues of oil processing HP-POX pilot plant Gasification modes Autothermal reforming GasPOX MPG * 6 th International Freiberg Conference on IGCC and XtL Technologies, 19-22 May 2014 4
Plant scheme 5
Reactor design and boundary conditions for Virtuhcon Benchmark* Operating pressure: 50, 60 and 70 bar Reactor temperature: 1200, 1400 ºC Three streams: natural gas, steam, oxygen-steam mixture *A. Richter, P. Seifert, F. Compart, P. Tischer, B. Meyer, A large-scale benchmark for the CFD modelling of non-catalytic reforming of natural gas based on the Freiberg test plant HP-POX, Fuel, volume 152, 15 July 2015, p. 110-121 6
Table 1. Input characteristics for different test runs Case 1 Case 2 Case 3 Case 4 Pressure, bar(g) 50 60 70 50 Reference values Target temperature, ºC 1200 1200 1200 1400 Gas residence time, s 14.8-16.2 Primary steam Mass flow rate, kg/hr 78.02 90.01 107.23 64.39 Temperature, ºC 289.60 299.20 309.70 277.50 Secondary steam Mass flow rate, kg/hr 23.15 28.63 31.50 22.77 Temperature, ºC 240.20 239.60 240.60 239.90 Oxydiser Mass flow rate, kg/hr 265.37 312.19 355.88 280.25 Temperature, ºC 240.20 239.60 240.60 239.90 Natural gas Mass flow rate, kg/hr 224.07 267.37 314.48 195.90 Temperature, ºC 359.10 362.10 365.30 355.20 7
Inflow CFD Model 2d axisymmetric Eddy-Dissipation Concept Adiabatic wall P-1 radiation model ATR mech (reduced GriMech 3.0, C3 species neglected) 28 species, 114 reactions Outflow 8
Numerical grid A comprehensive mesh study was carried out High quality is significant for the correct prediction of flame position and shape Natural gas Primary steam O2 + Secondary steam Upwind 9
Case 1 (50 bar (g), 1200 ºC) Temperature distribution (K) Temperature Maximum temperature: 3123.65 K Outlet temperature: 1620.17 K 10
Case 1. Comparison with optical measurement Temperature Measured flame length from Optisos: 265 mm Calculated flame length: 267 mm 11
Case 1. Flow field X Velocity 0 10 20 30 40 50 60 70 80 Maximum velocity: 83 m/s. 12
Case 1. CH4 and CO distribution (vol.-frac) Methane CO Methane CO 13
Case 1 Case 2 Case 3 Case 4 14
Results comparison between experiments and numerics Case1 Exp. Case 1 Num. Case 2 Exp. Case 2 Num. Case 3 Exp. Case 3 Num. Case 4 Exp. Case 4 Num. Pressure, bar(g) 50 60 70 50 Temperature, K 1475 1475 1475 1675 H2, %vol 48.27 45.80 48.32 45.99 48.40 45.22 48.06 47.74 N2, %vol 0.65 0.25 0.65 0.54 0.63 0.57 0.67 0.53 CO, %vol 23.79 23.96 23.73 23.72 23.64 23.45 25.61 26.27 CO2, %vol 4.19 3.62 4.10 3.57 4.13 3.64 3.89 3.19 H2O, %vol 19.33 21.36 19.04 20.93 18.64 21.26 21.71 22.25 CH4, %vol 3.76 4.94 4.17 5.20 4.55 5.83 0.06 0.011 Calculated temperature at the outlet, K 1620 1625 1608 1828 Measured flame length, mm 265 271 274 Calculated flame length, mm 267 278 263 311 Measured flame width, mm 41 43 45 Calculated flame width, mm 38 40 39 30 15
Conclusion Position of the flame and its shape are sensitive to mesh quality The flame characteristics were reproduced with a moderate error The outlet composition of the gas has a good agreement between numerics and experiments Deviation in calculating the conversion progress: gas at the outlet more away from equilibrium than experiment 16
Thank you for your attention TU Bergakademie Freiberg Institute of Energy Process Engineering and Chemical Engineering 09596 Freiberg Germany Phone: +49 3731 39-4217 Fax: +49 3731 39-4555 E-Mail: Yury.Voloshchuk@tu-freiberg.de Web: www.iec.tu-freiberg.de 17