© Strojni{ki vestnik 46(2000)7,422-428 © Journal of Mechanical Engineering 46(2000)7,422-428 ISSN 0039-2480 ISSN 0039-2480 UDK 624.19:628.83:614.84 UDC 624.19:628.83:614.84 Pregledni znanstveni ~lanek (1.02) Review scientific paper (1.02) Model dolo~itve kriti~ne hitrosti ventilacije v primeru po`ara v cestnem predoru pri izotermnih pogojih A Critical-Velocity Determination of Road-Tunnel Ventilation in Isothermal Conditions of a Fire-Incident Model Test Petar Donjerkovi} - Miodrag Drakuli} Ta prispevek predstavlja eksperimentalno določitev kritične hitrosti zraka in značilnosti za oceno učinkovitosti vzdolžne ventilacije v cestnih predorih v primeru požara. Določitev kritične hitrosti zraka pri izotermnih razmerah vzdolžne ventilacije je potekala v cestnem predoru "Sveti Rok", ki so jo izvajali v Institutu za pomorske in posebne raziskave v Zagrebu. Posebna pozornost je bila namenjena analizam porazdelitve dima v neposredni bližini simuliranega ognja za različne požarne obremenitve in različne vzdolžne hitrosti zraka. Analiza je bila narejena s pomočjo vizualizacije porazdelitve "mrzlega" dima, ki je bil simuliran z mešanico hlapov helija in parafina. Prispevek ocenjuje eksperimentalne rezultate kritične hitrosti zraka na modelu ter rezultate nedavnih raziskav. © 2000 Strojniški vestnik. Vse pravice pridržane. (Ključne besede: predori cestni, sistemi prezračevalni, kritična hitrost zraka, določevanje hitrosti) This paper shows an experimental determination of the critical air velocity, characteristic for an evaluation of the efficiency of road-tunnel longitudinal ventilation in fire-incident conditions. Determination of the critical air velocity was made in isothermal conditions of a model testing of the “Sveti Rok” road-tunnel longitudinal ventilation, performed in the hydrodynamic laboratories of the Marine Research and Special Technologies Institute in Zagreb. Special attention was paid to the analysis of smoke distribution in the close vicinity of a simulated fire source for different fire loads and different longitudinal air velocities. The analysis was made by visualisation of the distribution of “cold” smoke simulated with the mixture of helium and paraffin vapours. The paper evaluates the results of an experimental determination of the critical air velocity on the scale-model with available recent results of ventilation-system field tests. © 2000 Journal of Mechanical Engineering. All rights reserved. (Keywords: road tunnels, ventilation systems, critical velocity determination, air velocity) 0 UVOD Ventilacijski sistemi imajo pomembno vlogo pri preprečevanju požarov v predorih zaradi odvajanja toplote in pri "nadzoru" dima. Zelo pomemben je "nadzor" dima pri vzdolžnih ventilacijskih sistemih, saj je pri tem tipu ventilacije dim, z uporabo impulznih ventilatorjev, nameščenih na stropu, odstranjen skozi isti prostor, kjer poteka evakuacija potnikov. Zaradi tega mora ventilacijski sistem v primeru požara zagotoviti dotok pare, ki mora potiskati dim v nasprotni smeri evakuacije potnikov in dostopa gasilcev. Kritična hitrost zraka je prav gotovo najbolj pomemben parameter za določitev učinkovitosti vzdolžne ventilacije v primeru požara. Označena je kot "najmanjša vzdolžna hitrost pare, ki je povprečena prek prereza, pri kateri je dim "odstranjen". Glavna zahteva za uspešno odvajanje dima je doseči 0 INTRODUCTION Ventilation systems have an important role in tunnel-fire protection because of their capability to with-draw the heat generated during the fire and to “control” smoke. Smoke control is extremely important in longitudi-nal tunnel ventilation systems since in this type of ventilation the smoke is forced, with the help of ceiling-hung impulse fans, through the same space of the tunnel tube through which the passengers’ evacuation is carried out. Because of this, in the case of fire, the ventilation system must secure an efficient air stream which will push smoke in the direction opposite to the evacuation of passengers and firemen access. The critical air velocity is most cer-tainly the essential parameter for determining the efficiency of longitudinal ventilation in fire conditions. It is defined as the “minimum longitudinal air stream velocity, averaged along the cross-section, at which the smoke backlayering is made impossible”. Achieving the critical air velocity is a 00-7 VBgfFMK stran 422 P. Donjerkovi} - M. Drakuli}: Model dolo~itve kriti~ne hitrosti - A Critical-Velocity Determination Model kritično hitrost zraka, ki omogoča intervencijo gasilcev ter njihovo varno delo v bitki z ognjem. Nezanesljivi ventilacijski sistemi, ki ne zagotovijo kritične hitrosti zraka v primeru požara, neposredno ogrožajo gasilce, ki so vstopili v predor in pričeli gasiti, pa tudi potnike, ki se umikajo prek nezadimljenega dela predora proti izhodu, drugemu predoru ali v zaklonišče. 1 TEORETIČNA DOLOČITEV KRITIČNE HITROSTI ZRAKA Izrazi za oceno kritične hitrosti zraka v predoru (w) so razviti iz najbolj preprostih oblik, ki povezujejo c odvisnost: f kjer pomenita Q toplotno obremenitev in B širino predora. Thomasova empirična enačba: basic prerequisite for successful tunnel smoke purging which enables the intervention on the part of fire brigades and their safe work on fire fighting. An unstable ventilation system, which cannot reach the critical air velocity in fire conditions, directly endangers firemen who have entered the tunnel tube and begun fire fighting, as well as the passengers who are moving, in the stage of evacuation, along the unsmoked part of the tunnel towards the exit, the other tunnel tube or the shelter. 1 THEORETICAL DETERMINATION OF CRITICAL AIR VELOCITY Expressions for estimating the critical air velocity in the tunnel (wC) have evolved from the most simple form, which shows the functional connection: (1), where Q is the fire thermal load and Bt the tunnel tube width, through the Thomas’ empirical expression: w C gQH r0T0cpA (2), kjer so g gravitacijski pospešek, h višina predora, r gostota zraka, T temperatura vstopajočega zraka, c specifična toplota zraka, A površina prečnega prereza predora, kar vse vodi k spremembi Thomasovega izraza [1]: where g is gravity, H is the tunnel tube height, r0 is the air density, T0 is the temperature of the incoming air, cp the specific air heat and A the surface of the tunnel tube cross-section, all up to the modification of Thomas’ expression which has been confirmed by field tests [1]: K1K2 gQH 1/ (3), kjer so: - količnik K1 = Fr -1/3 - količnik K2 = 1+ 0,037 (strmina)0,8 - povprečna temperatura dima T = Q +T Q +T r0cp Enačba (3) za oceno kritične hitrosti (w) zahteva iterativni postopek, ki ga uporabljamo tudi pri računanju povprečne temperature dima. Froudovo število v količniku K še dodatno pripomore k zapletu ocene, saj ne poznamo vseh elementov za njegovo določitev. 2 DOLOČITEV KRITIČNE HITROSTI ZRAKA S PRESKUSI 2.1 Splošno Določitev kritične hitrosti zraka s preskusi: a) testiranje ventilacije in možnosti požara, b) na testnem modelu ventilacije in požara pri izotermnih in neizotermnih pogojih. Najbolj obsežne terenske raziskave obnašanja različnih ventilacijskih sistemov v primeru požara so bile opravljene v zapuščenem r0TDcpA where: - coefficient K1 = FrC -1/3 - coefficient K2 = 1+ 0,037 (gradient)0,8 - average temperature of smoke TD = r0cpAwC The above-mentioned formula (3) for the critical velocity (wC) estimate requires an iterative procedure since the same is also contained in the expression for the average smoke temperature. Intro-duction of the Froude number through the coeffi-cient K1 additionally complicates the estimate since not all elements for its determination are known. 2 EXPERIMENTAL DETERMINATION OF CRITICAL AIR VELOCITY 2.1 General Experimental determination of critical air velocity can be made by: a) field tests of ventilation and fire incident; b) scale-model tests of ventilation and fire incident in isothermal and non-isothermal conditions. So far the most complex and comprehensive field test of the behaviour of various ventilation systems in fire-incident conditions was performed in the abandoned Me- stran 423 glTMDDC P. Donjerkovi} - M. Drakuli}: Model dolo~itve kriti~ne hitrosti - A Critical-Velocity Determination Model predoru Memorial [1] (ZDA, 1993 do 1995) pod pokroviteljstvom Federalne cestne administracije, skladno s programom Tehničnega komiteja 5.9 ASHREA. V sklopu omenjenih raziskav so bile poleg kritične hitrosti vzdolžnega sistema ventilacije raziskane tudi toplotne obremenitve do 100 MW. Rezultate ocen kritične hitrosti dobimo s spremenjeno Thomasovo enačbo. Raziskave so pokazale, da kritična hitrost zraka ne presega vrednosti 3 m/s v razponu toplotnih obremenitev od 5 do 100 MW. Testni model pri neizotermnih pogojih temelji na toplotnih virih z bistveno zmanjšano toplotno obremenitvijo. Oka in Atkinson sta leta 1996 izvedla preskus neizotermnega testnega modela, ki je ugotavljal porazdelitev dima v povezavi z vzdolžno ventilacijo [2]. Glavni namen je bil določitev kritične hitrosti zraka v odvisnosti od različnih toplotnih obremenitev požara, ki je bil simuliran z zgorevanjem propana. Testni modeli požara in ventilacije pri izotermičnih pogojih navadno temeljijo na ti. "mrzlem" dimu, ki je pravzaprav mešanica enega od lahkih plinov (npr. helij) z zrakom, dušikom ali ogljikovim dioksidom. Za namen boljše vizualizacije so dodali barvne komponente, npr. paro glicerina. Toplotne obremenitve v testnem modelu ne ustrezajo tistim v primeru požara, z njimi niti ne moremo predstaviti vseh potrebnih fizikalnih lastnosti požara kot celotnega pojava. Kakorkoli že, navkljub očitnim omejitvam, ki spremljajo raziskave ventilacije v primeru požara pri izotermnih pogojih, dobljene vrednosti resničnih meritev v primerjavi z računalniškimi simuliranji ne odstopajo veliko. Omenjeni rezultati se nanašajo na določitev kritične hitrosti zraka in vizualizacije porazdelitve dima v bližini simuliranega vira ognja. 2.2 Model določitve kritične hitrosti pri izotermnih pogojih 2.2.1 Model, namen in pogoji testiranja Določitev kritične hitrosti zraka pri izotermnih pogojih so izvajali skupaj z modelom testiranja ventilacije v predoru Sveti Rok in v Inštitutu za pomorske in posebne raziskave v Zagrebu [3]. Model je vključeval uporabo vzdolžne ventilacije v razmeroma dolgem predoru z dvosmernim prometom. Namen testiranja je bil, poleg določitve vpliva prometa na učinkovitost delovanja ventilacijskega sistema, tudi napoved obnašanja sistema v primeru požara. Kritična hitrost zraka je bila določena za tri toplotne obremenitve 5, 10 in 20 MW. Podatki in rezultati so prikazani v preglednici 1. _____00 7 SnnsjtalcJUMllBilrSfl | ^BSfiTTMlliC | stran 424 morial tunnel [1] (USA, 1993 to 1995.), under the auspices of the Federal Highway Administration, according to the Technical committee 5.9 ASHREA program. The critical velocity of the longitudinal ventilation system, including the fire loads up to 100 MW was also determined within the scope of these tests. The tests resulted in a modifica-tion of the standard Thomas’ formula for the critical velocity estimate which thus acquires the form of the above-mentioned expression (3). Subject tests have established that the critical air velocity does not exceed the value of 3 m/s for a wide range of fire loads from 5 to 100 MW. Model tests in non-isothermal conditions are based on the real heat source with an essentially re-duced heat load that is appropriate to the model. An example of a non-isothermal model test is testing of the smoke distribution in conjunction with the longitudinal ventilation made by Oka and Atkinson in 1996 [2]. The basic purpose of this testing was a determination of the critical air velocity depending on different heat loads of fire stimulated by the propane burner operation. Model tests of the fire incident and tunnel ventilation in isothermal conditions are usually based on so-called “cold” smoke which is actually a mixture of one of the “light” gases (e.g. helium) with air, nitrogen or carbon dioxide. For the purpose of improved visualisation, coloured components are added, for example, glycerine vapours. This type of testing cannot model the heat loads adequately in case of fire, nor is it possible to present all the necessary physical properties of fire as a complex phenomena. However, in spite of obvious limitations which accompany the isothermal tests of ventilation in fire conditions, the obtained results of some physical values which characterise the behaviour of the ventilation system in fire conditions, have shown good correla-tion with field tests and computer simulations. First of all, the mentioned results refer to a determination of the criti-cal air velocity and a visualisation of the smoke distribution close to the simulated fire source. 2.2 Determination of critical velocity in iso-thermal conditions of model tests 2.2.1 Model, purpose and conditions of testing Determination of the critical air velocity in isothermal conditions was carried out during model testing of the ventilation of the Sveti Rok tunnel in the hydrodynamic laboratories of the Marine Research and Special Technologies Institute in Zagreb [3]. The proposed design solution included the use of longitu-dinal ventilation in a relatively long single-tube tunnel with bi-directional traffic. Determination of the critical velocity was made for three standard fire loads of 5, 10 and 20 MW. Because of the complexity of the men-tioned tests the interval of fire loads was not extended to values above 20 MW. Input data and the results of the subject tests are shown in Table 1. P. Donjerkovi} - M. Drakuli}: Model dolo~itve kriti~ne hitrosti - A Critical-Velocity Determination Model Zgoraj omenjene raziskave so se odvijale na tako imenovanih majhnih fizikalnih modelih (razmerje 1:25). Model predora je imel obračališče, dve zavetišči (oziroma vhoda v zavetišči), tri obcestna počivališča in pet baterijskih ventilatorjev z resnično dolžino 675 m (dolžina na modelu 27 m). Pleksi steklo (debeline 3 mm) je bilo nameščeno po predoru in omogočalo nadzor širjenja dima med simuliranjem požara in tudi nadzor pretoka prometa. The above-mentioned model tests were car-ried out on a so-called “small” physical model (Scale 1:25) and for tunnel modelling the central section which includes one turning gallery, two shelters (i.e. entrances into shelters), three lay-bys for stopping vehicles and five fan batteries with the actual length of 675 m (model length 27 m). Perspex (3 mm in thick-ness) was used for the tunnel lining to enable monitoring of the spreading of smoke during the fire incident simulation as well as the control of traffic flow. Preglednica 1. Podatki in rezultati za določitev kritične hitrosti na modelu in v predoru Table 1. Input data and results of critical velocity determination on model and on site Toplotna obremenitev Thermal load ekvivalentna površina gorenja equivalent burning area nastala količina dima generated volume of smoke ekvivalentna površina gorenja equivalent burning area nastala količina "mrzlega dima" generated volume of "cold smoke" pretok zraka skozi predor air flow through tunnel hitrost toka zraka skozi model predora air stream velocity through the model tunnel tube razdalja med "čelom " dima in mestom gorenja distance of the smoke layer face from the place of fire razdalja 1 med "čelom" dima in mestom gorenja (na kraju samem) distance 1 of the smoke layer face from the place of fire (on site)________________________________________ hitrost toka zraka skozi predor (na kraju samem) -kritična hitrost wc air stream velocity through the tunnel (on site) -critical velocity wc MW 2 m 3 m /s 20 2 cm l/s 6,4 10.6 19.2 3 m /s 0,0393 0.0458 0.0467 m/s m m m/s 5 2 4 8 32 64 128 0,423 -2,0 do/to -1,8 - 50 do/to - 45 2,115 10 33 0.493 -2.0 do/to -1.8 20 60 0.503 -2.0 do/to -1.8 - 50 do/to - 45 - 50 do/to - 45 2,465 2,515 (1 ) Negativen predznak pomeni, da je razdalja merjena v nasprotni smeri od smeri toka zraka med odstranjevanjem dima (1) Negative sign denotes that the distance is measured in the direction contrary to the direction of air flow during the “smoke purging” regime 2.2.2 Rezultati V skladu z rezultati testnega modela je odvisnost kritične hitrosti od toplotne obremenitve prikazana na sliki 1. Razvidno je, da krivulja kritične hitrosti enakomerno narašča glede na toplotno obremenitev in pri veliki toplotni obremenitvi se asimptotično približuje vrednosti 3 m/s, z drugimi besedami, občutljivost kritične hitrosti se zmanjšuje v območju večjih toplotnih obremenitev, kar v praksi omogoča uspešno odstranjevanje dima iz predora v širokem razponu toplotnih obremenitev do 100 MW s hitrostjo okrog 3 m/s. Slika 2 prikazuje medsebojno odvisnost kritične hitrosti zraka in toplotne obremenitve s Froudovim številom kot parametrom. Prej omenjeni spremenjeni Thomasov izraz (3), ki je bil preverjen v predoru Memorial, je bil uporabljen tudi za določitev družine krivulj w = f(Fr,Q). Razvidno je, da je krivulja w = f(Q), ki smo jo dobili s ekstrapolacijo merjenih vrednosti testnega modela, ki so bili izvajani v 2.2.2 Results In compliance with the results of model tests, the critical velocity dependence on the fire load is shown in Figure 1. It is clear that the critical velocity is a uniformly growing fire load curve which for high fire loads asymptotically aims at the value of 3 m/s. In other words, the “sensitivity” of the critical velocity drops in the area of high fire loads and in practice it enables the successful smoke purging of the tunnel for a wide range of fire loads up to about 100 MW with the speed of about 3 m/s. Figure 2 shows the interdependence of the critical air velocity and the fire thermal load with the Froude number as a parameter. For this, previously mentioned modified Thomas’ expression (3), checked during testing in the Memorial tunnel was used for determining the curve family wC = f (Fr, Q). It is vis-ible that the curve wC = f (Q), obtained by extrapolation of measured values of model tests carried out in the Marine Research and Special Technologies Insti- P. Donjerkovi} - M. Drakuli}: Model dolo~itve kriti~ne hitrosti - A Critical-Velocity Determination Model Iinštitutu za pomorske in posebne raziskave (BI), se dobro ujemajo s teoretično krivuljo za vrednost parametra Froudovega števila 4,5, še posebno v območju višjih toplotnih obremenitev. Enako vrednost Froudovega števila so omenili tudi avtorji testiranj v predoru Memorial [1] kot vrednost, ki najbolj približa teoretične in praktične preskuse. Določene odmike od kritičnih hitrosti, merjenih v BI, v primerjavi s teoretičnimi vrednostmi v območju toplotnih obremenitev, lahko pojasnimo z dejstvom, da je bila gostota mrzlega dima nespremenljiva za celotno področje med 5 in 20 MW, kar pa ni stvarno. Omenjena gostota ustreza 2,6 m/s tute (BI), matches well the theoretical curve for parameter value of Froude number 4.5, particularly in the area of higher fire loads. The same value of Froude number is mentioned by the authors of tests per-formed in the Memorial tunnel [1], as the value which best “adapts” the theoretical expression to the field tests. Certain deviations from critical velocities mea-sured in the BI, in comparison with theoretically com-puted value in the range of lower fire loads, are ex-plained by the fact that the density of “cold” smoke was kept constant for the entire testing interval from 5 to 20 MW, which is not a physical reality. The men-tioned density corresponded to an assumed smoke 2,5 2,4 2,3 2,2 2,1 1,9 wC = 0,2885Ln(Q) + 1,7006 o 10 15 20 toplotna obremenitev ognja / fire thermal load Q MW 25 Sl. 1. Odvisnost kritične hitrosti zraka od toplotne obremenitve ognja Fig. 1. Dependence of critical air velocity upon simulated fire load 6,0 m/s 5,0 4,0 ekstrapolacija glede na BI meritve . extrapolation acc. to BI measurements 3,0 0,0 2,0 1,0 0 10 20 30 40 50 60 70 80 90 100 MW toplotna obremenitev ognja / fire thermal load Q Sl. 2. Odvisnost kritične hitrosti zraka od toplotne obremenitve s Froudovim številom kot parametrom Fig. 2. Dependence of critical air velocity upon simulated fire load, with Froude number as parameter 00-7 VH^tTPsDDIK stran 426 2 0 5 P. Donjerkovi} - M. Drakuli}: Model dolo~itve kriti~ne hitrosti - A Critical-Velocity Determination Model načrtovani temperaturi dima 1350 °C, kar je še vedno prevelika vrednost za požare take toplotne obremenitve. Upoštevajoč dejstvo, da je dejanska temperatura dima za omenjene temperaturne obremenitve manj od 1350 °C in bi zato morala biti gostota mrzlega dima večja od tiste, ki so jo uporabili pri testiranjih (0,182 kg/m3). Iz tega lahko sklepamo, da bi krivulja BI in teoretična krivulja za w dale boljše rezultate, kakor če bi računali z manjšimi toplotnimi obremenitvami. 3 SKLEP Vrednosti kritične hitrosti zraka, ki smo jih dobili s testiranjem ventilacije v predoru Sveti Rok pri izotermnih pogojih, se ujemajo s teoretičnimi izračuni. Majhne odmike v območjih manjših toplotnih obremenitev lahko pojasnimo s povečanjem gostote mrzlega dima, ki ustreza dejansko nižjim temperaturam dima v predorih. Gostota uporabljenega mrzlega dima, ki ustreza resnični temperaturi dima 1350 °C je nižja od tiste, ki se pojavi v predorih v primeru manjših toplotnih obremenitev. Izotermni pogoji testiranja niso vplivali na verodostojnost pridobljenih podatkov o kritični hitrosti zraka, čeprav je problem fizikalnih vrednosti povezan izključno z delovanjem ventilacijskega sistema pri neizotermnih pogojih. temperature of 1350°C, which is still too high a value for fires of this low thermal load. Considering the fact that the actual temperature of smoke for the mentioned fire loads is less than 1350°C and that therefore the density of “cold” smoke should have been higher than the one used during model testing (0.182 kg/m3), it can be concluded that the BI curve and the theoretical curve for wC would have given much better agreement than the achieved one at mentioned lower fire loads. 3 CONCLUSION Measured values of critical air velocities obtained by model tests of the Sveti Rok tunnel ventilation in isothermal conditions have a high degree of agreement with theoretically computed and on-site confirmed values. The slight deviations noted in the area of lower fire loads can be annulled by an increase of the “cold” smoke density that is appropriate to actually lower temperatures of smoke on site. In that sense the density of used “cold” smoke, which corresponds to the real temperature of smoke of 1350°C, has shown to be lower than the one which occurs on site for the case of observed smaller values of fire loads. Isothermal conditions of testing, however, have not affected the credibility of the obtained data on the necessary critical air velocities, although it is the question of the physical value which is connected exclusively with operation of the ventilation system in fire, i.e. in distinctly non-isothermal conditions. kritična hitrost toplotna obremenitev ognja širina predora višina tunela gostota zraka temperatura zraka temperatura dima spec. toplota zraka površina prereza predora Froudovo število 4 OZNAČBE 4 SYMBOLS wC Q Bt H r0 T0 TD cp A Fr m/s critical velocity W fire load m width of tunnel m height of tunnel tube kg/m3 air density K air temperature K smoke temperature J/kgK spec. air heat m2 tunnel cross-section area - Froude number 5 LITERATURA 5 REFERENCES [1] Memorial tunnel fire ventilation test program (1995), Massachusetts Highway Department and Federal Highway Administration, Boston (USA). [2] Oka, Y, G.T. Atkinson (1995) Control of smoke flow in tunnels fire. Fire Safety Journal 25, 305^323. [3] Sambolek, I; M. Čizmek (1998) Modelsko ispitivanje ventilacije tunela Sv. Rok. Brodarski institut - Marine Research and Special technologies, Zagreb. gfin^OtJJIMISCSD 00-7 stran 427 |^BSSITIMIGC P. Donjerkovi} - M. Drakuli}: Model dolo~itve kriti~ne hitrosti - A Critical-Velocity Determination Model Naslova avtorjev: prof.dr. Petar Donjerkovič Fakulteta za strojništvo in ladjedelništvo Univerze v Zagrebu Ivana Lučiča 5 10000 Zagreb, Hrvaška mag. Miodrag Drakulič Brodarski institut Pomorske raziskave in posebne tehnologije Av.V. Holjevca 20 10020 Zagreb, Hrvaška Authors’ Addresses: Prof.Dr. Petar Donjerkovič Faculty of Mechanical Eng. and Naval Architecture Zagreb University Ivana Lučiča 5 10000 Zagreb, Croatia Mag. Miodrag Drakulič Naval Institute Marine Research and Special Tehnologies Av.V. Holjevca 20 10020 Zagreb, Croatia Prejeto: Received: 15.8.2000 Sprejeto: Accepted: 10.11.2000 00-7 VH^tTPsDDIK gfin^(5)eJ][M]g&3] I stran 428