ASHRAE LV-11-012-2011 Computer Modeling of Stairwell Pressurization to Control Smoke Movement During a High-Rise Fire.pdf

1、786 ASHRAE TransactionsABSTRACTComputer programs are valuable tools that can guide thedesign of any fire safety plan, particularly in high-rise build-ings that present many demanding design and smoke manage-ment challenges. A software package is used to examine thepossibility of employing air-handli

2、ng units to pressurize fireescape stairwells in order to maintain a smoke-free environ-ment throughout the fire. The computer results suggest thatequipment with modest capacity can achieve that goal, and theopportunity for success is increased when the stair shaft isconstructed with airtight materia

3、ls and all openings in theshaft are tightly sealed. The program output is used as a guide for the selection ofair-handling equipment that has sufficient capacity to keepsmoke from invading the stairwells. As the door at the fire flooris opened, the capacity of the positive pressure fans must beincre

4、ased to prevent smoke from entering the stair shaft. Pres-surizing stairwells must always balance the benefits of provid-ing fresh air in the fire escapes with the disadvantages ofcreating a situation that makes it difficult to open a fire escapedoor against the increased pressure inside the stairwe

5、ll. INTRODUCTIONHigh-rise fires present a challenging task to both firefight-ers and fire protection engineers. Extended escape routes thatexist in high-rise structures not only make fighting a fire onupper floors very difficult, but also create problems for occu-pants who decide to evacuate the bui

6、lding via fire escapes. Firefighters carrying bulky equipment to upper floors, and build-ing occupants trying to descend to the exterior must competefor available space in the fire escape stairwells. Complicatingthis counterflow of humanity is the possibility that the stair-wells will become contami

7、nated with smoke. Clearly, a lifesafety plan to keep smoke from entering the stairwells wouldbe beneficial to both parties. This paper describes a smokecontrol program that can help design air-handling equipmentused to pressurize and to maintain a clean-air environment inhigh-rise fire escape stairw

8、ells during the fire. Smoke is able to move as a result of very small pressuredifferences, and those pressure differences are magnified intall structures. In high-rise buildings, large pressure differ-ences exist between the bottom and top of the building, andthey tend to pull any smoke generated by

9、 the fire toward theupper floors. Adding to the upward pull, the buoyancy of thehot gases generated by the fire drive the generated smokeupward through any openings in the building and particularlyinto the elevator shafts and fire escape stairwells where it expe-riences a low resistant path to the t

10、op of the building. Theupward force experienced by the lighter, hot gases is oftenreferred to as stack effect or chimney effect, and it can be amajor factor in determining the route that smoke will takeduring a high-rise fire. To face these challenges, fire protection engineers mustbe able to assess

11、 the potential paths that smoke may take andthen design countermeasures to keep the smoke away fromescape routes and occupied spaces within the building. Safetyoptions include pressurizing floor spaces so that people whochoose to remain in the building or are unable to exit the build-ing safely woul

12、d be able to remain in a smoke-free area. Othermeasures would include forcing fresh air into the stairwells sothat occupants can descend from the upper floors through asmoke-free environment. To select a successful smoke management plan for aspecific building design and a reasonable set of fire cond

13、itionsComputer Modeling ofStairwell Pressurization to ControlSmoke Movement During a High-Rise FireW.Z. Black, PhD, PEMember ASHRAEW.Z. Black is a Regents Professor Emeritus at the George W. Woodruff School of Mechanical Engineering, Georgia Institute of Technology,Atlanta, GA.LV-11-0122011. America

14、n Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in ASHRAE Transactions, Volume 117, Part 1. For personal use only. Additional reproduction, distribution, or transmission in either print or digital form is not permitted without ASHRAES prior writte

15、n permission.2011 ASHRAE 787requires a design tool that can quantitatively assess how smokewill tend to move throughout the structure. Considering themultitude of factors that influence the movement of smoke, itis impractical and far too expensive to carry out comprehen-sive experimental tests to el

16、iminate designs that could poten-tially fail and narrow the selection to those that would createsafe conditions inside a high-rise. A software package that candetermine the pressure distribution inside the structure andcalculate the flow of smoke in all building compartments, onthe other hand, could

17、 be used to identify those factors thatwould lead to an improved life safety plan. Parametric runswith the programs would permit a fire safety engineer to focuson a design plan what would result in the greatest possibilityof safe egress from the building.This paper illustrates how a smoke control pr

18、ogramcalled COSMO (COntrol of SMOke) can be used to design areasonable smoke control plan for a high-rise building in theevent of a first-floor fire. The code is used to explore reason-able smoke control options, such as pressurizing various areaswithin the building, including floor spaces, elevator

19、 shafts,and stairwells. The COSMO program has previously investi-gated the concept of pressurizing floor areas so that peoplewho are unable to evacuate the building will have a safe areato survive the fire (Black 2009a, 2009b, 2009c). However, inthis paper the computer results focus on the concept o

20、f pres-surizing stairwells by using positive pressure ventilating fanslocated throughout the fire escapes.For such a complex problem as smoke movement in arealistic building, a word of caution is necessary. Output froma computer program is obviously quantitative. However, eventhough the program cons

21、iders all of the factors that influencethe movement of smoke, the output should be best viewed asproviding qualitative trends that will give reasonable guidanceto the design of a smoke control system for a high-rise build-ing. The computer output presented here illustrates conditionsinside a particu

22、lar building for a fixed set of fire conditions.The qualitative results of the program give clear indication ofhow a smoke control plan should evolve, but the quantitativeresults should only be used for the specific case being consid-ered. Results for other sets of fire conditions and other build-in

23、g types can be easily simulated with the program. Repetitive,parametric runs with the program will clearly identify thosefactors that play a dominant role in controlling smoke move-ment within a building. Equally important is the ability of thesmoke movement program to isolate those elements in buil

24、d-ing design and construction that have little or no influence onthe smoke flow during a fire.SMOKE MOVEMENT SOFTWAREThere are several smoke movement and smoke controlprograms available to fire protection engineers and code offi-cials. Early papers on the subject provided simplifiedapproaches to smo

25、ke management schemes, equations todetermine the location of the neutral pressure plane (NPP),and basic models for the vertical spread of gases in high-risestructures (Tamura and Wilson 1966; McGuire 1967a, 1967b;Barrett and Locklin 1968 1969; Hutcheon and Shorter 1968;Tamura 1969). Papers that foll

26、owed (Tamura and Wilson1970; Tamura 1972; Tamura and Shaw 1973; McGuire andTamura 1975; Lie and McGuire 1975; Jones and Quintiere1984) refined previous smoke movement work by analyzingthe flow of smoke into all areas of a building. Perhaps the most widely accepted and most frequentlyused types of co

27、mputer programs are classified as networkmodels, which are based on a series of interconnected regionswithin the building that have a uniform set of properties. Anetwork program called CONTAM that was initially devel-oped to study air quality in buildings has been frequently usedto evaluate smoke mo

28、vement in buildings (NBS 1981, 1982;Klote and Milke 1992, 2002; Klote and Fothergill 1983).Other network programs designed to track air quality in build-ings (SMOKESIM Yuill and Haddad 1994, and ASCOSKlote and Tamura 1991) and multi-zone models (COMISFeustel 1999) have also been applied to the proce

29、ss quanti-fying smoke movement in buildings. Network models subdi-vide the building into a number of nodes and apply anincompressible static fluid equation to each node, whichrelates the nodal pressure to elevation. The result of this exer-cise is an array of algebraic equations equal to the number

30、ofnodes throughout the building. The set of equations is thensimultaneously solved for the pressure at each node. Once thepressure distribution is known, the flow rate between eachnode can be calculated. Many network models do not includeconsideration of heat transfer so that the temperature distrib

31、u-tion of the gases must either be assumed or provided by sepa-rate programs. If a temperature distribution is simply assumed,it may violate the requirements of the conservation of energy,and it would result in erroneous information. Another distinctly different type of smoke movementprogram is base

32、d on a finite element analysis (FEA) or compu-tational fluid dynamics (CFD) models that have recentlygained popularity. Several texts describe these methods, whichcan be applied to arrive at detailed flow patterns near a devel-oping or steady fire (Patankar 1980; Spalding 1981). Theseprograms have b

33、een used to examine the flow of gases aroundthe source of a fire and into a small number of enclosures adja-cent to the fire. When using an FEA program, a mesh ofelements is superimposed over the domain of interest, and theconservation equations are applied to each element. A finegrid is used where

34、property gradients are large, and a coarsegrid is used where property variations become more uniform.The final desired mesh is one that balances mathematicalaccuracy and computational speed. FEA programs becomeimpractical and are typically not used to simulate smokemovement inside high-rise building

35、s with a large number ofindividual compartments due to the excessive amount ofcomputational time necessary to arrive at a set of properties foreach grid.The COSMO code is a differential model that is funda-mentally different from network and FEA smoke control788 ASHRAE Transactionsmodels. It is stru

36、ctured to be a design tool to assist fire protec-tion engineers and code officials when they want to quantifysmoke movement throughout the building and they want todesign measures to manage smoke movement during a fire.The mathematical basis for the COSMO model is formulatedon differential equations

37、 rather than the algebraic equationsthat are employed in frequently used programs such asCONTAM. The basic equations must be solved simultane-ously because all properties in the equations are related. Othermodels that do not consider a complete set of conservationequations produce a single algebraic

38、 equation that is a functionof only the pressure at each node. These simplified modelsneed only to solve a set of algebraic equations rather than anarray of differential equations that are necessary whenemploying a differential model. When using a differential model, all vertical shafts in thebuildi

39、ng, such as elevator shafts and stairwells, are subdividedinto a large number of volumes. The conservation of energy,mass, and momentum equations are written for each of the gasvolumes. The volumes are assumed to contain compressiblegases as they travel through the shafts, and the fundamentalconserv

40、ation equations are solved simultaneously for theproperties of the smoke. Other smoke movement models oftenassume the smoke is incompressible and stationary, which canlead to computational errors. By employing a complete set ofconservation equations, the differential analysis is morecomplex, but als

41、o more accurate because it more closely simu-lates the flow conditions that actually exist within a buildingunder fire conditions. For example, COSMO is able to takeinto account the frictional effects experienced by the smoke aswell as the buoyancy forces that draw the smoke upward in theshafts. The

42、se forces are neglected in other models, which canoften lead to large errors in the computed smoke properties.The COSMO code also includes a detailed analysis of theconvective and radiative heat transfer that takes place betweenthe hot gases in the shaft and the cool shaft walls. Therefore,the code

43、is able to determine the temperature distribution ofthe smoke in the shafts rather than estimate the temperaturedistribution, as is often the case in other more simplifiedmodels. COSMO provides flow rates, velocities, pressures,temperatures, and densities of the smoke distributed through-out the bui

44、lding. Provisions within the code allow for pressur-ization of the stairwells, elevator shafts, and floor spaces sothat the viability of using pressurized regions to create smoke-free environments for occupants of the building can be deter-mined. The code is not limited to a high-rise building, but

45、thedifficulty of maintaining a safe environment within a tallbuilding and providing smoke-free escape routes is moredemanding and more complex in a high-rise. Therefore, theoutput from a smoke control program can be extremely bene-ficial when selecting a smoke control plan that is appropriatefor a t

46、all building. The engineering basis for the program anda complete set of equations used for the foundation of the codeare given in (Black 2009b, 2010), and they are not repeatedhere. This paper focuses on one aspect of smoke control issue.It isolates the problem of pressurizing a stair shaft with ai

47、r-handling units in order to maintain non-contaminated escaperoutes during the fire. This type of smoke control plan shouldnot be considered in isolation, but only after considering acomprehensive plan that involves smoke control in other areaswithin the building to ensure that pressurization of one

48、 areacomplements and does not conflict with efforts in other areas.For example, a complete smoke control plan should not onlyconsider pressurizing stairwells, but also should involve pres-surization of floor spaces and evacuation of smoke from eleva-tor shafts. The validity and accuracy of computer

49、models for smokemovement should be supported by practical experience andavailable measurements of conditions during realistic fires.Computer output from smoke control programs has beenbolstered by measurements from numerous experimentalprojects that involved fires in actual buildings (Maddox 2004;Tamura 1985; Achakji and Tamura 1988) and in experimentalfire towers (Klote and Tamura 1991; Klote and Bodart 1985).Measured values for pressure distributions and smoke flowrates between separate building compartments help determinethe accuracy of computer programs that endeavor to simulateac