1、10 2016 ASHRAEThis paper is based on findings resulting from ASHRAE Research Project RP-1600.ABSTRACTThe primary purpose of this study is to develop engineer-ing methods to assess the impact of increased makeup airvelocity in atria. The current restriction defined by NFPA 92(NFPA2015)states:“Themake
2、upairvelocityshallnotexceed200 ft/min (1.02 m/sec) where the makeup air could come intocontact with the plume unless a higher makeup air velocity issupported by engineering analysis.” This limitation not onlylimits creative and aesthetic atria designs but may also repre-sentasignificantcost.Thisstud
3、yanalyzestheeffectofmakeupair injected by a variety of vent sizes at elevations at or belowthe limiting elevation of the flame through numerical simula-tions. This study focuses on identifying worst-case scenariosfor the interaction of makeup air with an axisymmetric plumeby applying computer modeli
4、ng to simulate multiple configu-rations, observe the results, and adapt further simulations toelicit the most extreme cases. A mass flow rate diagnostic isused to assess the increase in entrainment (i.e., smoke produc-tion.) This mass flow diagnostic is developed to provide acomparative analysis, as
5、sessing the increase in the rate ofsmoke production with a specified makeup air velocity withthat produced with no mechanical makeup air. The propor-tionalincreaseinentrainmentisdefinedasanalphafactor.Themost significant smoke production increase and smoke layerstabilizationdescentisassociatedwitha1
6、MW(950Btu/s)fire,with lesser increases observed for 2.5 and 5 MW (2370 and4740 Btu/s) fires. As the makeup air is introduced further fromtheedgeoftheflame,theapparenteffectoftheairflowvelocityis reduced.INTRODUCTIONThe primary purpose of this study is to develop engineer-ing methods to assess the im
7、pact of increased makeup airvelocity in atria. The current restriction defined by NFPA 92states that makeup air must not exceed 1.02 m/s (200 fpm)during the operation of a mechanical smoke exhaust system(NFPA 2015). The makeup air required for atria during theoperation of a mechanical smoke exhaust
8、system may involvelargeareasofpassiveopeningsandmechanicalventilation,alllocated below the smoke layer. This limitation not only limitscreative and aesthetic atria designs but also involves signifi-cant costs. Many engineering designers use alternative meth-ods to exceed the limit of the code, claim
9、ing that the velocitylimit is too restrictive.PROJECT SCOPEAtria have become increasingly popular design featureswithin large commercial spaces, contemporary hotels, andmultilevel shopping centers. Modern atria serve as prominentaesthetic features, and are often several stories high. Manyatria have
10、glazed roofs and large windows for grandeur and afeeling of space and light.The openings in the floors created to form an atrium posefire and smoke challenges. The large space allows for easiersmoke spread between floors and adjacent openings. A prin-cipal design objective for fire protection system
11、s in an atriumis to protect occupants from the adverse effects of smoke andcontain the fire and smoke to its room of origin. As atria fash-ion larger openings, the ability to compartmentalize smokeA CFD Study to Identify Methods to IncreaseMaximum Velocity of Makeup Air forAtrium Smoke ControlChrist
12、ine Pongratz James A. Milke, PhD, PE Arnaud Trouve, PhDChristine Pongratz is a graduate research assistant at the University of Maryland, College Park, MD, and a graduate fire engineer at Arup,London, UK. James A. Milke is a professor and chair and Arnaud Trouve is a professor in the Department of F
13、ire Protection Engineering,University of Maryland, College Park, MD.ST-16-002 (RP-1600)Published in ASHRAE Transactions, Volume 122, Part 2 ASHRAE Transactions 11and fire decreases. Consequently, alternative designapproaches must be used to maintain life safety objectives.The hazard of smoke spread
14、in atria must be addressed tosatisfy identified objectives. Possible objectives detailed inNFPA 92 (NFPA 2015) include the following:Maintain a tenable environment in the means of egressfrom large-volume building space during the timerequired for evacuationControl and reduce smoke spread between the
15、 fire areaand adjacent spacesProvide conditions within and outside the fire zone toassist emergency response personnel in conductingsearch and rescue operations and in locating and con-trolling the fireContribute to the protection of life and reduction ofproperty lossAid in post fire smoke removalIn
16、 order to accomplish the design objective(s), an engi-neered smoke management system is considered for all atria.NFPA 92 includes requirements for the design of smokemanagement systems (NFPA 2015). By evaluating the designcharacteristics of the atrium, numerous smoke managementapproaches may be cons
17、idered.There are various design approaches for atria that areintended to maintain tenable conditions for occupants. Amechanical smoke exhaust system is a commonly usedapproach in North America (Klote et al. 2012). The system isdesigned to use mechanical exhaust to stabilize the bottom ofthe smoke la
18、yer at a predetermined height. The exhaustremoves smoke from the upper levels of the atrium to preventaccumulationofheatandsmokeandpreventthedescentofthesmoke layer interface below the predetermined height.The mechanical ventilation system assumes the forma-tion of a smoke layer at the ceiling. Fact
19、ors that may affect thesmoke layer include sprinkler activation, HVAC systems, aircurrents striking the plume, upward thrusting airflows, and airforced into the upper layer by means other than the plume.In order to provide effective mechanical ventilation,makeup air supply must be provided. The supp
20、ly of makeupair may enter into the compartment from passive openings,such as doors or windows, or additional mechanical ventila-tion. The amount of air that must be supplied is not dictated inthe model building codes, but several restrictions areaddressed in the design of this component of the syste
21、m.High makeup air velocity may increase air entrainmentintoaflametosignificantlyaffectfiredevelopmentandsmokemovement within a compartment. Increased air entrainmentwill increase mixing between ambient air and smoke toupsurge the volume of smoke produced. Also, the additionalair velocity may tilt th
22、e flame and disturb the upward trajec-tory of the fire plume, which may expose occupants to addi-tionalradiantheatfluxandsmoke.Aschematicdiagramoftheeffect of increased makeup air velocity on the fire plume isshowninFigure1.Theimageillustratesadisturbedfireplumewith smoke entering the balcony levels
23、 of the atrium. Theadverseeffectofthemakeupaircompromisestheabilityofthemechanical exhaust system to achieve the design goals ofmaintaining the smoke layer above a particular elevation.Currently NFPA 92 restricts the makeup air velocity notto exceed 1.02 m/s (200 fpm) to prevent significant plumedef
24、lection and disruption of the smoke layer interface. Thislimitation is further discussed by Klote et al. (2012). NFPA 92permits greater velocities of makeup air if the design issupported by engineering analysis (NFPA 2015).Although there is no restriction on the overall volumetricflow of makeup air,
25、 NFPA 92 suggests makeup air to bedesigned at85% to95% ofthe exhaust,not includingthe leak-age through the building enclosure (NFPA 2015). Coupledwith the limitation on the maximum makeup air velocity, thissuggestion can lead to large areas for makeup air vents. Over-all, the requirement often prese
26、nts design challenges andincreased costs.SEPARATION DISTANCE OF FUEL PACKAGESInadditiontotheincreaseinsmokeproductionbecauseofincreased makeup air velocity, the thermal radiation from thedesign fire to exposed items close to the fire may also increasedue to the flame tilt. When selecting a design fi
27、re, NFPA 92requires that the design consider the type of fuel, fuel spacing,and configuration (NFPA 2015). The base fuel package isconsidered the maximum probable size of fuel that is likely tobe involved in a fire situation.Figure 1 Makeup air velocity effects on fire plume inatrium (not to scale).
28、Published in ASHRAE Transactions, Volume 122, Part 2 12 ASHRAE TransactionsThe base fuel package in an array of potential fuels isselectedastheonewiththegreatestheatreleaserate(oftenthelargest fuel package in the array). In addition, the configura-tion of the design fire must consider the impact of
29、the firesradiant heat flux to its surroundings. This study uses computersimulationscoupledwithapointsourcemodeltoreportontheappropriateseparationdistancegivenidentifiedairflowcondi-tions.RESEARCH OBJECTIVESThis study investigates the adverse effects of makeup airvelocity specifically on the smoke la
30、yer interface position andthe separation of fuel packages. The objective of the study isto evaluate the impact of makeup airflow velocity introducedat elevations at or below the flame height on the smoke layerinterface and fuel package separation distance. Further, anengineering tool is proposed to
31、provide guidance on possibleadjustments in the mechanical exhaust when exceeding thecurrent 1.02 m/s (200 fpm) limitation as a function of fire size,supply vent area, and vent elevation.Past research by Heskestad (1984), Beyler (2008),Hadjisophocleous and Zhou (2007), and Kerber and Milke(2007) sugg
32、ests that velocities above 1.02 m/s (200 fpm) canalter an axisymmetric smoke plume resulting in an increase inthe amount of air entrained into the plume. While the study byKerber and Milke showed that high makeup airflow velocitiesmay result in deficient operation of the smoke control system,it did
33、not provide a detailed understanding of the exact condi-tions that led to substandard behavior and whether there maybe a range of makeup airflow velocities beyond the limit inNFPA 92 that still provide acceptable design solutions.Thisstudywascompletedusingastate-of-the-artcompu-tational fluid dynami
34、cs (CFD) model, Fire Dynamic Simula-tor (FDS) Version 6 (McGrattan et al. 2013). An FDS modelgrid analysis was completed to evaluate the proper grid reso-lution for the study. As further explained in the next section,two previous studies were completed to evaluate the effects ofmakeupairusingFDSVers
35、ion4.0.Consequently,thefirststepin the present study is to revisit the simulations created in theprevious study to confirm similar results using FDSVersion 6.0.This study is designed to isolate the effects of makeup airby introducing the air through a vent located on one side, at aclose distance to
36、the base of the fire. The fire is represented asa gas burner, centrally located within a 10 m (32.8 ft) tallatrium. The intent is to study the effects of makeup air at thebase of the fire plume to investigate the increased mass flowrate of the smoke plume to develop and quantify the results.The vent
37、 is a dedicated, duct-mounted makeup air vent andsupplies makeup air velocities of 1, 1.25, 1.5, and 1.75 m/s(197, 246, 295, and 345 fpm) directed at the fire. The fire sizesincluded in the study are 1, 2.5, and 5 MW (950, 2370, and4740 Btu/s). The duct-mounted makeup air vent is varied insize and e
38、levation for each simulation.AnanalysisoftheFDSsimulationsisconductedtoisolatetheeffectsoftheduct-mountedmakeupairventairflowonthefire. The analysis concentrates on the effects of the increasedmakeup air on the mass flow rate of the plume and, in turn, onthe smoke layer interface height within the a
39、tria. Also, addi-tional simulations are run to investigate the outcome ofincreased makeup air on radiant heat flux and in turn the sepa-ration distance of fuel packages.A parameter to measure the strength of the forced horizon-talairflowwithrespecttothebuoyantverticalflowgeneratedbythe combustion pr
40、ocess is created and compared to the FDSresults of smoke production and heat flux. Engineering designtool correlations for both increases in smoke production andheat flux are formulated as corrections to the expressions forsmoke flow rates and separation distances used in NFPA 92.PAST STUDIESThe ori
41、ginal study initiated by ASHRAE, undertaken byHadjisophocleous and Zhou (2007) at Carleton University,and the University of Maryland study by Kerber and Milke(2007) conclude that increased makeup air velocity increasesairentrainmentintothefireplumeandcreatesgreatervolumesofsmoke,resultinginadeepersm
42、okelayer.ThepresentstudysimulatessimilarCFDmodelconfigurationsfromtheseprevi-ous studies using FDS 4.0 to confirm consistent results usingFDS 6.0. The present study obtains relatively similar resultswhile applying original design considerations such as extend-ing the boundary domain and using the sm
43、oke layer interfacediagnostics.The present study confirms the past study results thatexceeding 1 m/s (197 fpm) does disturb the plume and createsa greater volume of smoke production. However, the presentstudy also confirms it is possible to mitigate the smoke layerdescent by modifying the mechanical
44、 exhaust rate by apredictable amount.Carleton University Elevation ComparisonThe results from this study agree with the results of theCarleton University study (Hadjisophocleous and Zhou 2007)andidentifytheelevationofthemakeupairventasakeyinflu-ence to the overall smoke production within the plume.
45、Thisstudy distinguishes that the concentration of carbon dioxide(CO2) within the smoke layer is greater when the makeup airinjection is at the base of the flame. In the few cases where theelevation of the makeup air vent was raised but still within orright above the flame region, the impact on the C
46、O2concen-tration within the flame was reduced. This comparison simplyidentifies that with more of the increased makeup air directedat the combustion zone, a greater volume of smoke isproduced,causingthesmokelayerheighttodescendbelowitsoriginal design height. It is important to note that, in the pres
47、-ent study, makeup air is naturally entrained into the compart-ment from every wall at a low elevation. The volume ofmakeup air is plentiful for the flame, which isolates the duct-mountedmakeupairventvelocityastheonlychangingfactor.Published in ASHRAE Transactions, Volume 122, Part 2 ASHRAE Transact
48、ions 13The study by Hadjisophocleous and Zhou (2007) alsocompares the impact of the elevation of the makeup air vent.However, in their study, the makeup air vent is located on thewall and further from the base of the fire (2.5 m 8.2 ft, asopposed to 1 m 3.3 ft in the present 1 MW 950 Btu/s firestudy
49、). The arrangement in the Carleton University studycausessubstantialturbulentmixingwithinthecompartment,asthe clean air injected at higher elevations is pulled downwardto feed oxygen to the fire source. Overall, the current studyrecognizes the dynamics of the mixing and distribution of heatandCO2intheatriumandtheroleofelevationofinjectedcleanair. The C
