H. PRETREL and L. AUDOUIN Laboratoire ETIC (IUSTI-IRSN)

Transcription

1 GDR Incendie, 3 juin-1er juillet 11 - Marseille Ecoulement à un passage de porte sous les effets combinés de la ventilation naturelle et forcée. PRETREL and L. ADOI Laboratoire ETIC (ISTI-IRS)

2 Content Context of te study Teoretical background about doorways flow Fire experiments Facility description Test grid Experimental determination of doorway flow Typical test beaviour Data analysis Pressure cascade Velocity profiles Correlative approac Metodology Results Conclusions GDR Incendie, 3 juin-1er juillet 11 Marseille

3 Context Smoke movement, an important issue for safety in nuclear installation Propagation means for eat, soot and toxic species Filter Clogging Failure of E&E devices Doorway flow Large mass flow rate of smoke Complex flow: - Bidirectional flow (fres air and smoke) - Interface and mixing layer State of te art atural ventilation - Steckler-akaya (1985) - Bryant (5) - Quintiere (9) Mixed (forced and natural) ventilation - Positive pressure ventilation (PPV) - Mecanically ventilated compartment fire VETILATIO ETWORK P m u smoke Fres air FIRE Electrical failure GDR Incendie, 3 juin-1er juillet 11 Marseille 3

4 Teoretical background (1) Direction of ventilation P Z Z outflow o inflow Adjacent room fire room P s P P( z) ( ) o z + P { S g ( z) ydrostatic P. atural ventilation Static P. Forced ventilation GDR Incendie, 3 juin-1er juillet 11 Marseille 4

5 Teoretical background () Direction of ventilation P Local velocity outflow Z Z ( z) P( z) ( z) inflow o pper and lower velocities Adjacent room fire room P s P 1 Mass flow rate o, ( 1 ) + P o, S PS 1 1 i Cd dz o Cd dz Objective: Investigate te combined effect of te fire (natural convection) and te ventilation (forced convection) on te doorway flow GDR Incendie, 3 juin-1er juillet 11 Marseille 5

6 Test configuration P 1 P 3 L L L1 1 3 L3 Tree compartments assembly Two doorways Initial one-way air flow for ventilation (upstream and downstream doorway) Fire wit gas propane burner DIVA facility of IRS D 1 Q D 3 GDR Incendie, 3 juin-1er juillet 11 Marseille 6

7 Tests grid Two parameters Fire eat release rate RR Q Ventilation flow rate T C Q P g A (m 3 /) Test matrix g A Variables investigated: Velocity profiles eutral plan position In & out flow rates q v Seven tests RR[1-4]kW [5-31]m 3 / [-8] -1 per room GDR Incendie, 3 juin-1er juillet 11 Marseille 7 q v Q (kw)

8 Doorway flow measurement Centreline temperature and velocity profiles (7 points) Bi-directional velocity probes (Mc Caffrey) ( z) k P ( z) eutral plane.14m Z ( z ) 1.79m 1.45m In & out flow rates 1.9m in C w ( z) ( z) dz m & C w ( z) ( z) dz out.75m.39m.5m w GDR Incendie, 3 juin-1er juillet 11 Marseille 8

9 Typical test beaviour P 1 P 3 L L L1 1 3 L3 D 1 Q D 3 q ext P 1 q adm P 3 RR P L 143-4cm 14cm 179cm 179cm 143-4cm 14cm GDR Incendie, 3 juin-1er juillet 11 Marseille 9

10 Pressure cascade P 1 P 3 L L -1,5-1, P (Pa) Before ignition -1,5-1, During combustion P (Pa) L1 1 D 1 Q D 3 3 L3 -,5, nodes 1 3 -,5, nodes 1 3,5 1, 1,5, P5-6 P6-6 P7- P1-3 P-3 P3-3 P4-3,5 1, 1,5, P5-1 P6-3 P7- P1- P-3 P3-4 P4-1 Fire modifies te static pressure cascade Small difference of pressure between rooms (~1Pa) GDR Incendie, 3 juin-1er juillet 11 Marseille 1

11 Velocity profiles Te fire RR and ventilation flow rate modifies te velocity profile Te increase of fire RR favours te spread of smoke Te effect of te ventilation depends on te relative position of te doorway and te fire source (upstream or downstream) Doorway D 1-31m 3 / 1 z RR-386kW RR-9kW RR-194kW RR- 97kW -,6 -,4 -,,,,4,6 Doorway D 1 - RR1kW 1 z qv- 7m3/ qv- 31m3/ L1 1 P 1 P 3 L L D 1 Q D 3 3 L3 Doorway D 3-31m 3 / 1 z RR-386kW RR-9kW RR-194kW RR- 97kW -,6 -,4 -,,,,4,6 Doorway D 3 - RR1kW 1 z qv- 7m3/ qv- 31m3/ Te ventilation may stop te propagation of smoke or te entrance of fres air -,6 -,4 -,,,,4,6 -,6 -,4 -,,,,4,6 GDR Incendie, 3 juin-1er juillet 11 Marseille 11

12 Correlative approac (1) Correlate flow features wit fire RR and ventilation flow rate,,,, f ( q, RR) in out v RR Input variables Experimental data fitting, o, P S, intermediate variables o, + δ P o, ( 1 ) o, S δ P + δ P S Bernoulli S,, k o o 1 m k 3 m Cd Cd ( 1 min( 1; ) ( 1 max( ; ) output variables in out, P Q GDR Incendie, 3 juin-1er juillet 11 Marseille 1

13 Correlative approac () Metodology RR Input variables Experimental data fitting, o, P S, intermediate variables o, Bernoulli + δ PS o, ( 1 ) δ P o, + δ P S S,, k o o 1 m k 3 m Cd Cd ( 1 min( 1; ) ( 1 max( ; ) output variables in out 1/ Correlate o, f(rr): D + o, ( ) D 1 D Q 1 / Intermediate variables calculation /, / inf, P from te measured doorway caracteristics:, o, 3/ Experimental data fitting A1 A3 A5 P A q 1+ A Q 1+ A q S v ( 1 ) 1 [ ( )] 4 v B1 B3 1+ B Q ( 1+ Bqv ) GDR Incendie, 3 juin-1er juillet 11 Marseille 13 ( ) C C C Q + C q O 1 PS, 1 1 v 3

14 Correlative approac (3) 1, 1,,8,6,4,, o / D + o, Q31m3/ Qm3/ Q7m3/ qv ( ) D 1 D Q Q,,5,1,15,,5 1,4,35,3,5,,15,1,5, / Doorway D 1 Doorway D 3 qv qv-7 qv- qv-31 Q31m3/ Qm3/ Q7m3/ Q31m3/ Qm3/ Q7m3/ Q,,5,1,15,,5 C + ( ) C C q C1 Q 1 v 3,15,1,5, -,5 -,1 -,15 P s,,5,1,15,,5 S Doorway D 1 Doorway D 3 A3 A5 [ 1+ AQ ( 1 A4q v )] A1 v P A q + qv-7 qv- qv-31 Q31m3/ Qm3/ Q7m3/ Q31m3/ Qm3/ Q7m3/ Q,5,4,3,,1, / inf Doorway D 1 Doorway D 3 qv qv-7 qv- qv-31 Q31m3/ Qm3/ Q7m3/ Q31m3/ Qm3/ Q7m3/ Q,,5,1,15,,5 B3 ( Bqv ) B1 Q 1 1+ B + Good accordance of te data fitting correlations GDR Incendie, 3 juin-1er juillet 11 Marseille 14

15 Top and bottom velocity,6,4,, -, -,4 -,6.3 Doorway D 3 Doorway D 1 Q7m3/ Q7m3/ Qm3/ Qm3/ Q31m3/ Q31m3/ qv--d1 qv-7-d1 qv--d1 qv-31-d1,,5,1,15,,5 Q,6,4,, -, -,4 -,6 Doorway D 1 Doorway D 3 Q7m3/ Q7m3/ Qm3/ Qm3/ Q31m3/ Q31m3/ B qv--d1 qv-7-d1 qv--l1 qv-31-d1 Q,,5,1,15,,5 D1 Q D3 Te fire RR is te dominant parameter for upper velocity Larger effect on te out flow tan te in flow Te ventilation is te dominant parameter for te lower velocity Te effect of is predominant for te outflow at te D 3 doorway Critical conditions for no smoke propagation or no fres air entrance GDR Incendie, 3 juin-1er juillet 11 Marseille 15

16 eutral plan 1,,9,8,7,6,5,4,3,,1, Doorway D 1 Doorway D 3 Q7m3/ Q7m3/ Qm3/ Qm3/ Q31m3/ Q31m3/ qv-31-d1 qv--d1 qv-7-d1 qv--d1 B Q,,5,1,15,,5 D1 Q D3 Significant difference between te two doorways Te effect of te ventilation flow rate is more significant for te downstream doorway GDR Incendie, 3 juin-1er juillet 11 Marseille 16

17 Mass flow rates D1 Q D3,15,1,5, -,5 -,1 -,15 m o Doorway D 3 Q7m3/ Q7m3/,1 Qm3/ Qm3/ Q31m3/,5 Q31m3/ A/^,5, qv-7 Doorway D 1 qv- qv-31 -,5 Doorway D 3 qv- Q,,5,1,15,,5,15 -,1 -,15 m i Doorway D 1 Q7m3/ Q7m3/ Qm3/ Qm3/ Q31m3/ Q31m3/ A/^,5 qv-7 qv- qv-31 qv- Q,,5,1,15,,5 Critical conditions for no smoke propagation or no fres air entrance: o propagation of smoke in upstream room : 1kW & 31m3/ o entrance of fres air from te downstream room: kw & 31m3/ GDR Incendie, 3 juin-1er juillet 11 Marseille 17

18 Conclusions & future works Conclusions Mecanical ventilation as a significant effect on te doorway flow Pressure differences are low (1-5Pa) for te two contributions: natural and forced ventilation Main parameters: - ventilation flow rate - relative position of te doorway (upstream or downstream te fire) Correlative approac - Relative contributions of fire RR (natural ventilation) and mecanical ventilation (forced ventilation) - Critical conditions to avoid propagation of smoke or entrance of fres air Perspectives Improve correlative model by introducing temperature measurements (termal stratification, ) Quantify te effect of fire on te static pressure( expansion effect) GDR Incendie, 3 juin-1er juillet 11 Marseille 18