Thermodynamic Aspects of. Interaction Between Premixed Hydrogen. Flame and Water Droplets. GAI Guodong Sergey KUDRIAKOV.

Transcription

1 Thermodynamic Aspects of Interaction Between Premixed Hydrogen Flame and Water Droplets GAI Guodong Sergey KUDRIAKOV DEN/DANS/DM2S/STMF/LATF Saclay, France

2 Contents Ø Introduction Ø Lumped-parameter Approach Ø CREBCOM CFD model Ø Conclusions and Perspectives CEA Saclay PAGE 2

3 Introduction ü The course of a serious accident ü Spray system of the PWR ü Interaction between flame and spray

4 THE COURSE OF A SEVERE ACCIDENT Severe accidents Zr+2H 2 O=>ZrO 2 +2H 2 H 2 +1/2O 2 rejection of H 2 Formation of a mixture potentially explosive H 2 + H 2 O + Air sources of ignition Propagation of flames: (Rapid deflagration or... detonation) Pressure loads à Threat to the containment We must estimate the consequences of an hydrogen explosion!!! CEA Saclay PAGE 4

5 SPRAY SYSTEM OF THE PWR Spray system (PWR 900 MWe) The mixture can be ignited during the spray phase!!! In order to: Ø Limit the pressure inside the containment via steam condensation The action of spray can lead to: 1. Overpressure Mitigation Ø Capture the fission products to prevent radioactive release Ø Mix gaseous species or, on the contrary, 2. Flame Acceleration Source: [Foissac 2011] Source: [Wingerden 1995], [Gupta, 2014] CEA Saclay PAGE 5

6 PHENOMENA A front of the turbulent flame A wave of pressure Spray droplets Source: [Zhao 2011] strong weak Ø Break-up of the water droplets Ø Dilution of flame with the fin droplets The velocity of the flame is reduced. Ø Increase the flame surface S Ø Interaction flameturbulence The velocity of the flame increases. CEA Saclay PAGE 6

7 OBJECTIVES v Investigation on Thermodynamic aspects of the interaction flame-spray, behind the flame front, with the Lumped-parameter and CFD model v Analysis of real Experimental works by applying the CREBCOM model with Evaporation process CEA Saclay PAGE 7

8 Lumped-parameter Approach ü Main hypothesis and definitions ü Governing equations ü Test cases

9 MAIN HYPOTHESIS AND DEFINITIONS Ø Ideal gas mixture Ø One irreversible chemical reaction H 2 + ½O 2 à H 2 O Ø T "#" and P "#" constant, m * change with X,- Ø The system is closed and adiabatic Volume fraction of liquide phase: α = V 1"2 V 343 Conservation of m5 * = m5 6 m5 6 = 7 n 6"# 1"2 =>? 8 M 8 + m,- ; e * = 7 Y " "#" h " * + G 8AB Conservation of energy: " * e * = e 6 P Q 7 Y " "#" c =," (T ) " dt N 1"2 1"2 + Y,- ; u,- ; CEA Saclay PAGE 9

10 SEVERAL CASES CONSIDERED Table: Initial conditions for different cases q Case I: Comparison with CHEMKIN code q Case II: Limiting liquid volume fraction (α 1"R ) q Case III: Influence of liquid volume fraction (α) q Case IV: «Accidental» initial conditions* *Source: [Malet 2008] CEA Saclay PAGE 10

11 CASE I: COMPARISON WITH CHEMKIN Lumpedparameter CHEMKIN 4.8% of H 2 is present in combustion products (CHEMKIN) Good coincidence for lean and rich composition of hydrogen CEA Saclay PAGE 11

12 CASE III: INFLUENCE OF LIQUID FRACTION 9 8 alpha = 2x10-4 alpha = 3x10-4 alpha = 4x10-4 alpha = 2x10-4 alpha = 3x10-4 alpha = 4x Asymptotic Pressure (bar) Asymptotic Temperature (K) X H2 (-) X H2 (-) T max (K) P max (bar) α = α = 2 10 b@ α = 3 10 b@ α = 4 10 b@ *Pressure for stoichiometric initial hydrogen-air mixture Mitigation of the flame Effective depressurization effect CEA Saclay PAGE 12

13 CREBCOM CFD model ü Numerical model and main hypothesis ü Validation of the model and pressure transient ü Experimental investigation

14 NUMERICAL COMBUSTION MODEL CREBCOM model: Mass conservation: Species transport: Momentum conservation: Energy conservation: eft u e3 efo e3 efj k e3 ef e3 + h ρu = 0 + h ρuy l = ρω l + h ρu u + PI = ρg + h ρuh 3 = ρg h u ρ h 6,8 ω 8 + S z{ " Combustion rate: ω = Q h criterion function Thermal source term: S z{ = H(T T * ) Progress variable: ξ r, t = Y, - r, t Y,-,"#" Y,-,6"# Y,-,"#" = 0 fresh gas ˆξ ξ = 1 burnt gas Source: [Efimenko 2001] CEA Saclay PAGE 14

15 MAIN HYPOTHESIS OF EVAPORATION Fresh gas Preliminary assumptions: Combustion ξ > ξ thres Y N o o o o o Stationary droplets in fresh gas : v ž{4? = 0 No interaction between droplets and fresh gas Droplets evaporate totally and immediately across the flame if criteria satisfied Evaporation takes place in a closed adiabatic cell N e int > e lat? Criteria for evaporation: Y Evaporation Burnt gas 1. ξ r, t = j - {,3 bj -, > ξ j threshold -, bj -, 2. Sufficient volumetric internal energy for the complete evaporation of the liquid phase (heat-up of droplets and latent heat) CEA Saclay PAGE 15

16 PRESSURE TRANSIENT 8 7 0Dnospray 0Dspray CREBCOMnospray CREBCOMspray Volume fraction: α = 2 10 b4 Pressure evolution (bar) No spray Spray t?t>l 0.1 s P?t>l 1.1 bar Two main effects of evaporation Validation of the CREBCOM code Good Coincidence Time (s) Second criteria of evaporation No evaporation period Deceleration of flame Effective depressurization Delay of peak pressure CEA Saclay PAGE 16

17 EXPERIMENTAL FACILITY D = m Opposite spray Fig. Experimental tube Spray nozzle sketch Fig. Arrangement of nozzles Geometry used in the CREBCOM calculation Source: [Carlson 1973] CEA Saclay PAGE 17

18 EXPERIMENTAL RESULTS Fig. Important test cases Evaporation rate: α = dα dt X H2 = 16 % P AICC = 5.9 bar Fig. Pressure transient evolution Important heat loss Mitigation effect of spray Pressure(psi) X H2 = 16 % Continuous spray and evaporation P 0 = bar Time(s) *AICC = Adiabatic Isochoric Complete Combustion Source: [Carlson 1973] CEA Saclay PAGE 18

19 FLAME VELOCITY DETERMINATION Source: [Kuznetsov 1999] Run-up distance: XS : position where the V 61>Rt reaches C sp in the combustion products X D = γ C For 16%H Á, C? = 787 m/s Source: [Dorofeev 2009] 1 κ ln γd h D h = BR + K γ = c? μ Á (σ 1) Á S Ã v max C sp = 787 m/s δ D B Å X D 110 > L D = 30 Å ÆRÇÈ SLOW Deflegration Regime Maximal Flame Velocity v max 70 m/s Flame velocity (m/s) 16%H Á Relative distance x/d (-) X D vamax 1 2 L D = 15 v max 70 m/s CEA Saclay PAGE 19

20 SIMULATION RESULTS 4 Test 7 No spray H carlson 4 Test 8 With spray carlson-spray 1700-spray Pressure evolution (bar) Pressure evolution (bar) Time (s) Time (s) Test T 0 (K) P 0 (atm) T liq (K) x(m) K 0 (m/s) H (J/m 3 Ks) α (s b1 ) b@ *Experimental geometry with 13 transducers of pressure Suitable choices for H, K 0 and α Effective mitigation effect of spray system CEA Saclay PAGE 20

21 ENERGY BALANCE Energy Conservation: Energy contributions (J/m3/s) 6x10 6 5x10 6 4x10 6 3x10 6 2x10 6 1x10 6 Combustion Energy Loss Spray Effect Spray Combustion Heat Loss Time (s) Pressure evolution is associated with energetic evolution Evaporation Rate (kg/s) Evaporation Rate Total Evaporation Rate Time (s) Total evaporated mass: m,- ; = N t=>? α ΔtV zt11 ρ,- ; kg Mean evaporation rate: Q t=>? = R -Û 3 uüu = kg/s < Q?{>á = 4.6 kg/s CEA Saclay PAGE 21

22 FLAME VELOCITY EVOLUTION Flame Propagation Direction Heat loss : v 5 m/s Spray effect : v 10 m/s Effective deceleration of flame Flame Front Sketch CEA Saclay PAGE 22

23 MESH EFFECT Fig. Original Mesh Fig. Finer Mesh x 10 cm x 5 cm Combustion rate: x can affect chemical reaction rate K 0 needs to be increased to get the good peak time Pressure evolution (bar) x = 10 cm x = 5 cm Time (s) Influence on K 0 carlson dx = 5 cm dx = 10 cm CEA Saclay PAGE 23

24 Conclusions and Perspectives

25 CONSLUSIONS AND PERSPECTIVES HIGHTLIGHTS: v The lumped-parameter and CFD code can give reliable results for isochore and adiabatic combustion, with or without water spray; v The depressurization and mitigation effect of spray droplets is demonstrated, with increase of the liquid volume fraction; v The transient evolution of pressure and flame velocity can be simulated by choosing suitable parameters for the combustion, heat loss and evaporation process. PERSPECTIVES: v A more sophisticated model for the parameter K0 can be proposed, taking into account flame-spray interaction; v A more sophisticated model for evaporation rate can be proposed, taking into account the diameter of droplets. CEA Saclay PAGE 25

26 OPERATORS IN CAST3M DETO PRIM KONV PENT VARI CALLM CALMU MODELISER DOMA DIFF PROG CHPO EVOL CEA Saclay PAGE 26

27 More Miracles with Cast3M CEA Saclay PAGE 27

28 REFERENCES [Carlson 1973] L.W.CARLSON, R.M.KNIGHT and J.O.HENRIE Flame and detonation initiation and propagation in various hydrogen-air mixtures, with and without spray, Atomics International DIvision Rockwell International. May 11, [Foissac 2011] A. Foissac, J. Malet, M.R. Vetrano, J.M. Buchlin, S. Mimouni, F. Feuillebois, O. Simonin, Droplet size and velocity measurements at the outlet of a hollow cone spray nozzle, Trans Atomization and Spraysto deal with hydrogen hazards in large scale facilities. 21(11), pp , [Zhao 2011] H. Zhao, H.F. Liu, J.L. Xu and W.F. LI, Experiment study of drop size distribution in the bag breakup regime, Industrial and Engineering Chemistry Research, 50, pp , [Efimenko 2001] Efimenko A.A., Dorofeev S.B., CREBCOM Code System For Description of Gaseous Combustion, J.Numer.Methods Fluids 76, pp , [Gupta 2014] S. Gupta et al, Hydrogen Combustion During Spray Operations Tests HD30 to HD35, OECD-NEA THAI-2 Project, Report No TR-HD-30-35, [Kuznetsov 1999] M. Kuznetsov, V. Alekseev, A. Bezmelnitsyn, W. Breigung, S. Dorofeev, I. Matsukov, A. Veser, Yu. Yankin Effect of Obstacle Geometry on Behavior of Turbulent Flames Institut fur Kern-und Engergietechnik July, [Dorofeev 2009] Dorofeev, S.B., Hydrogen flames in tubes: Critical run-up distances International Journal of Hydrogen Energy, 34, pp , [Malet 2008] J.Malet et al. Sprays in Containment: Final Results of the SARNET Spray benchmark Nuclear Engineering & design 25 Sept, CEA Saclay PAGE 28

29 MESH EFFECT Pressure evolution (bar) No spray Spray CREBCOMnospray CREBCOMspray carlson carlson-spray Different choices for K 0 Effective mitigation effect of spray system Due to acoustic wave, oscillation magnitude increases with K Time (s) Test T 0 (K) P 0 (atm) T liq (K) x(m) K 0 (m/s) H (J/m 3 Ks) α (s b1 ) b@ *Experimental geometry with 13 transducers of pressure CEA Saclay PAGE 29

30 CFD MODEL VALIDATION 0 relative error Relative Error of Pressure (%) x H2 (-) Mean values of pressure and temperature are calculated in CREBCOM Initial conditions Pressure Gas temperature Droplets temperature Steam concentration H 2 combustion limit Volume fraction of droplets 1.0 atm K K b@ Geometry : Good coincidence with the lumpedparameter code for the asymptotic P and T Relative error = CREBCOM 0D 0D 100% CEA Saclay PAGE 30

31 CHOICE OF K 0 (COMBUSTION RATE) Pressure evolution (bar) K 0 K0=4.5 K0=5.5 K0=10 K0=20 carlson 0Dnospray The range of K m/s < K 0 < 10 m/s Flame velocity v flame 70 m/s Source: [Efimenko 2001] Time (s) ê = σ 1 Å Ã ê ë ì for à ê ì K * = S P(σ + 1) 5. 73m/s 4 < 500 The parameter K 0 = m/s works well for the available data K 0 70 m/s CEA Saclay PAGE 31

32 OSCILLATIONS IN PRESSURE EVOLUTION Pressure evolution (bar) Dnospray 0Dspray CREBCOMnospray CREBCOMspray 6.75 T T 4 z"11>3"4# = 0.2 s T 4 z"11>3"4# = bá s T >z4o 3"z õ>=t = 2 L C? bá s T 4 z"11>3"4# = T >z4o 3"z õ>=t Acoustic wave Time (s) 0.2s The oscillations are due to the propagation of the Acoustic Wave CEA Saclay PAGE 32

33 RUN-UP DISTANCE The run-up distance Decreases with the Blockage Ratio Validation of correlation X D = γ C γ = 1 κ ln γd h c? μ Á (σ 1) Á S Ã δ D + K B Å Å ÆRÇÈ D h = BR Far from the Run-up Distance Smooth tube in Carlson s experiment Source: [Dorofeev 2009] CEA Saclay PAGE 33

34 CASE 2: LIMITING VOLUME FRACTION 1.4 limit volume fraction Limit Volume Fraction (x10-3 ) X H2 (-) Total evaporation region Q z tr = L t=>? α R> 1"R"3 = bå H Á O Á(21% in air) H Á O H Á O 1"2 H Á O =>? CEA Saclay PAGE 34

35 CASE 4 : «ACCIDENTAL» CONDITIONS Accidental scenario: Source: [Carlson 1973] Case P ini (bar) T ini gas (K) liq T ini(k) X ini H2 ( ) X ini H2 O( ) α( ) IV [0.09, 0.3] 0.45 (2.0, 3.0, 4.0) 10 b@ Asymptotic Pressure (bar) alpha = 2x10-4 alpha = 3x10-4 alpha = 4x X H2 (-) P ùúûû R> = 9.95 bar üa@ B* ýþ = 8.24 bar P R> Asymptotic Temperature (K) alpha = 2x10-4 alpha = 3x10-4 alpha = 4x10-4 T ùúûû R> = 1774 K üa@ B* ýþ = 1104 K T R> Effective depressurization in accidental scenarios X H2 (-) CEA Saclay PAGE 35

36 VOLUMETRIC ENERGY EVOLUTION 2x10 6 Volumic energy Volumic Energy (J/m3) 1.5x10 6 1x Time (s) Close to volumetric energy of combustion products in 0D AICC calculation Homogenization Effect due to the pressure wave propagation Thermal Expansion Mass convection Leads to the local decrease of volumetric energy CEA Saclay PAGE 36