1、266 2009 ASHRAEABSTRACTEmergency smoke control systems within buildings and underground transportation facilities use air movement and pressurization to contain and remove smoke in order to provide safe exiting for occupants. Wind can be a hindrance to the smoke control systems in several ways.For s
2、moke control systems that use natural ventilation for exhaust or makeup air, wind pressures may seriously disrupt the system performance by redistributing airflows and caus-ing extra unwanted mixing of smoke in exit pathways. Such disruption can occur during relatively mild wind conditions, dependin
3、g on the exposure of the building openings and the wind climate at the site. Another possible effect of wind is to carry smoke that has already been exhausted to the outdoors back into the building make up air system. Makeup air loca-tions should be placed away from smoke exhausts, ideally on differ
4、ent faces of the buildings and with large vertical sepa-ration.Although wind effects have been mentioned in several codes and guidelines, there is no detailed discussion of the extent of the potential problems, nor standard analysis tech-niques for design purposes. This paper discusses the design is
5、sues, provides recommendations, and presents some real-world examples that illustrate the potential issues and ways of reducing wind effects.INTRODUCTIONWind can have a significant effect on the movement of smoke inside and outside a building and within underground transportation facilities. Wind pr
6、essures can dramatically alter exhaust and make-up airflow rates and airflow distributions. These problems can occur either for large building openings using natural venting strategies or for cases with powered exhaust coupled with naturally vented make-up air (e.g., open doors and windows). Undergr
7、ound transportation facilities are influenced by pressures on portals and pedestrian exit ways at the surface. All facilities have risks that the smoke exhausted from the building may be re-entrained into make-up air loca-tions. Smoky air brought back into a building can adversely affect occupants t
8、rying to exit.The focus of this paper is on effects of wind on smoke control systems designed for large open spaces like atriums, malls, and airport terminal buildings. It does not directly address smoke control related to residential or commercial spaces such as in a high rise building nor does it
9、address pres-surized stairwell design issues.Useful references on the above topics are: Klote the expansion or acceleration the make-up air may have in the architectural space; and the relative locations of potential fire locations in the space. For situations where potential fire locations are clos
10、e to the openings of the building (i.e., less than 15 m (50 ft) as suggested by the authors as a simple rule of thumb), then wind speed effects should defi-nitely be included in the smoke control system design. For projects with this situation, the authors suggest that more common wind speeds could
11、create problems with naturally ventilated openings than a design based on make-up air supplied effectively with mechanical systems.Another approach to evaluating the importance of wind effects is to estimate the overall probability of building open-ing speeds greater than an acceptable threshold for
12、 the project, such as 1 m/s (200 fpm). A detailed estimate would predict pressure coefficients for a variety of wind directions and eval-uate opening air speeds for a series of wind directions and wind speeds. The estimate would also determine the wind climate in terms of probability for various win
13、d speed/direc-tion combinations. Then the overall probability of excessive air speeds is the sum of probabilities of wind speed/direction combinations that produce excessive air speeds.A simple calculation is presented below for a cube shaped building in St. Louis, with an opening facing south towar
14、ds the prevailing wind direction. Pressure coefficients for the leading face were estimated to be +0.6 for winds directly aimed at the building face (from due south), +0.5 for wind directions at 22.5 degrees from due south, and +0.3 for wind directions 45 degrees from due south.The overall estimated
15、 probability of excessive opening speeds (1 m/s or 200 fpm) is 21% for this hypothetical cube building located in St. Louis. More protected locations or loca-tions aimed away from prevailing winds will have lower prob-abilities. However, for many building opening locations and cities, the probabilit
16、y of excessive air speeds through openings is likely to be greater than 10% and perhaps 20% in some situ-ations.There is no guidance in the literature that states whether the 10% to 20% probability of wind disruption at natural makeup air vents is acceptable. In the authors opinions, consideration s
17、hould be given in the design to reducing the use of natural ventilation for make up air, or at least in taking measures to reduce external wind pressures. Such measures could be selection of building faces shielded by neighboring structures, and using areas not affected by the majority of prevailing
18、 wind directions.The difficulty is there is not currently a standard approach by designers to reach a common layout and sizing of openings in buildings. As well, there is pressure in the design industry to minimize costs of designing engineered smoke control systems. Both factors make it difficult f
19、or buyers of these services, usually architects, facility owners and developers, to approve more comprehensive analyses that would cost more in engineering and also potentially lead to more costly solutions. The authors opinion is that codes and standards should change to clarify these concerns.SMOK
20、E EXHAUST PLUME DISPERSIONThe effect of wind on smoke exhaust emitted outdoors is a second important effect (besides wind pressure disruption on interior wind flows) discussed in this paper. This section discusses the behaviour of smoke exhaust after being emitted to the outdoors. The goal is to ach
21、ieve maximum exhaust dilu-tion at building air intakes and makeup air locations and mini-mize the re-entrainment of smoke back into the building. If the exhaust is not diluted enough before being re-entrained, then some areas of the building may experience hazardous smoke levels.As shown in Figures
22、2 through 4, winds can create recir-culation regions and wake flows around a building. These regions can have wind shear (i.e., large gradient in speed) and high turbulence that will degrade the initial momentum of the exhausts and transport the exhaust back to the building.Table 2. Calculated Wind
23、Pressures and Air Speeds through Openings for a Typical City for 1%, 2.5%, and 5% Wind SpeedsWind Speed Level1% 2.5% 5%Umet10.8 m/s(24.1mph)9.1 m/s(20.4 mph)8.3 m/s(18.6 mph)Building Height = 10 m (33 ft)UH(10 m)7.8 m/s(17.4 mph)6.6 m/s(14.8 mph)6.0 m/s(13.4 mph)p25 Pa(0.1 in.H2O)18 Pa(0.07 in.H2O)1
24、5 Pa(0.06 in.H2O)Uopen2.3 (460 fpm) 2.0 (388 fpm) 1.8 (354 fpm)Building Height = 30 m (100 ft)UH(30 m)9.9 m/s(22.1 mph)8.4 m/s(18.8 mph)7.6 m/s(17.0 mph)p41 Pa(0.165 in.H2O)29 Pa(0.116 in.H2O)24 Pa(0.096 in.H2O)Uopen3.0 (588 fpm) 2.5 (495 fpm) 2.3 (451 fpm)272 ASHRAE TransactionsRowan Williams Davie
25、s carbon monoxide (CO) maximum of 2000 ppm for a few seconds, averaging 1500 ppm or less for the first 6 minutes of exposure, averaging 800 ppm or less for the first 15 minutes of the exposure;Figure 9 Favorable configuration of high exhaust and lower-level intakes for make-up air.274 ASHRAE Transac
26、tionssmoke obscuration levels that are continuously main-tained below the point at whichdoors and walls are discernible at 33 ft (10 m).Milke (2000) provides an assessment of smoke hazards from fires in large spaces. A significant conclusion is that visi-bility can be reduced to less than 1 m even t
27、hough the temper-ature rise and generation of toxic gases are relatively modest (i.e., temperature rise of 10C (18F) and CO concentration of 100 ppm). The conclusion from this paper and opinions2of many designers of smoke control systems is that often if tena-ble visibility targets can be achieved s
28、hort-term exposures to other smoke hazards by normally healthy adults exiting a building will be acceptable. Dilution targets are, therefore, proposed below in order to achieve tenable levels of visibility.It is because of the above concerns about smoke hazards that designers of smoke control system
29、s often conduct assess-ments of smoke transport using hypothetical fire scenarios which rapidly grow to relatively high heat releases rates that causes the worst-case growth of a large volume-flow rate plume of dense hot smoke. For large spaces, these assessments involve calculations with algebraic
30、equations or CFD computer modeling. Prediction of visibility distances in exit pathways using these methods and comparison to tenability criteria is often used to judge success or failure of the perfor-mance of a proposed smoke control system.The authors opinion is that in most cases the assessments
31、 should assume a conservative worst-case fire event based on fuel that has a high soot yield rate and relatively high mass optical density. Relatively high values of these factors will result in predictions of shorter visibility distances in locations affected by smoke. This is often a suitable assu
32、mption when no information of fuels that may be involved in a fire can be assured for the large space being designed. Further, it recog-nizes that there are many uncertainties in the range of possible fire scenarios and so it seems reasonable to attempt to account, in part, for this uncertainty with
33、 a worst-case fuel assumption. This is reinforced by Milke (2000) who raises a point of caution, via a reference to Tewarson (1995), noting that both soot yield rate and mass optical density values can vary by orders of magnitude for different ventilation conditions.For the projects that the authors
34、 have been involved with, they often select polystyrene or polyurethane foam as the fuel for the design fire. These fuels have between 19% to 24% soot per unit mass (SFPE Handbook (1995). Selecting polystyrene as the design fuel is also useful since it has a property of a conservatively high value o
35、f mass optical density i.e., 1.4 m2/g, as quoted by Klote dm= mass optical density (m2/g). Table 3.4 of Klote or, Table 3. Three Story Atrium-Predicted Dilution Requirements of Smoke to Achieve10 m (33 ft) and 30 m (100 ft) Visibility TargetsTotal Heat Release RateVolume Flow at Flame Limiting Heigh
36、tEquation (4)Soot Emission Rate Polystyrene1Exhaust Flow Rate from Atrium2Internal DilutionRequired Total Dilution to Achieve 10 m (33 ft) Visibility3Required Total Dilution to Achieve 30 m (100 ft) Visibility3Btu/s MW cfm m3/s g/s lb/s cfm m3/s 1000 1.055 11,620 5.5 9 0.02 54,131 25.5 4.7 30 902000
37、 2.110 23,240 11.0 17 0.04 73,263 34.6 3.2 29 855000 5.275 58,099 27.4 43 0.094 115,067 54.3 2.0 27 8210,000 10.55 116,198 54.8 87 0.191 170,286 80.4 1.5 27 80Notes:1Polystyrene: Heat of Combustion 25.6 kJ/g; Soot Yield Rate 21%; Mass Optical Density 1.4 m2/g (SFPE Handbook, 1995)2Hypothetical examp
38、le of an axisymmetric plume (NFPA 92B, 2008) rising to a smoke layer maintained at 9.14 m (30 ft) above the base of the design fire.3Assume 100% of the make-up air is contaminated with smoke at a visibility of 10 m (33 ft) using K = 2 in Jins equation.Table 4. Six Story Atrium-Predicted Dilution Req
39、uirements of Smoke to Achieve10 m (33 ft) and 30 m (100 ft) Visibility TargetsTotal Heat Release RateVolume Flow at Flame Limiting HeightEquation (4)Soot Emission Rate PolystyreneExhaust Flow Rate from Atrium1Internal DilutionRequired TotalDilution to Achieve 10 m (33 ft) visibility2Required TotalDi
40、lution to Achieve 30 m (100 ft) visibility2Btu/s MW cfm m3/s g/s lb/s cfm m3/s 1000 1.055 11,620 5.48 9 0.02 182,829 86.3 1.6 41 1242000 2.110 23,240 11.0 17 0.04 235,412 111.1 2.5 36 1075000 5.275 58,099 27.4 43 0.094 335,136 158.1 4.4 31 9410,000 10.55 116,198 54.8 87 0.191 447,556 211.2 6.6 29 88
41、Notes:1Hypothetical example of an axisymmetric plume (NFPA 92B, 2008) rising to a smoke layer maintained at 20.12 m (66 ft) above the base of the design fire.2Assume 100% of the make-up air is contaminated with smoke at a visibility of 10 m (33 ft) using K = 2 in Jins equation.276 ASHRAE Transaction
42、sDtot= 105:1 to achieve 30 m (100 ft) visibility of the make-up air (in the worst case).The required external dilution Dextcan be estimated in Equation (12) by first calculating the internal dilution contri-bution Dintin Equation (11). Dintis defined here as the ratio of the volume flow rates at the
43、 flame limiting height, Vlim (cfm), of the fire of total heat release rate Q (Btu/s) to that of the proposed exhaust from the building, Vexh (cfm). Equation (10) is merely a simple combination of Equations (4) to (7).Vlim= 11.625 Q (10)Dint= Vexh/Vlim(11)Dext= Dtot /Dint(12)An example of the applica
44、tion of the above would be for the design of a smoke control system for a short atrium. For a design fire size of Q = 5,000 Btu/s (5.275 MW) and the proposed exhaust rate of Vexh= 115,000 cfm, (54 m3/s) then the required minimum target external dilution would be Dext= 18:1 to achieve the 10 m (33 ft
45、) visibility at the make-up air intake locations, or at least Dext= 53:1 if a more conservative target of 30 m (100 ft) visibility is agreed upon. Considering the broad range of uncertainties in the potential fire events and the impacts it may have on occupants exiting a building it would, of course
46、, be more prudent to design the exhaust and make-up air locations based on the 30 m (100 ft) target for most projects.SUMMARY AND DISCUSSIONWind effects are important to consider in designing emer-gency smoke removal systems. Wind pressures exerted on naturally ventilated makeup air openings can cre
47、ate disruptive air movement within a building for moderate wind speeds that can occur frequently. Also, smoke exhaust can be carried by the wind back towards building openings, which creates potential safety problems for people exiting the building. Building codes and guidelines mention wind effects
48、, but no standard criteria or analysis techniques are available.This paper demonstrates how wind pressures can be esti-mated using simple equations and pressure coefficients for simply-shaped buildings. Wind tunnel modeling is recom-mended to determine pressure coefficients for more complex building
49、 geometries including surrounding buildings. Proba-bilities for these wind pressures can be estimated from wind climate information for the building site. For an example case presented, disruptive wind speeds could occur with a proba-bility of 20%. To avoid such a high probability would require design decisions which limit the percentage of make-up air provided through natural vent openings on buildings to a rela-tively small percentage or that other provisions are taken in the smoke control design to overcome potential adverse wind effects.Several suggestions ar
