1、690 ASHRAE TransactionsThis paper is based on findings resulting from ASHRAE Research Project RP-1376.ABSTRACTLaboratory measurements have been made to determinethe accuracy of several handheld instruments when used tomeasure the exhaust airflow rates during commercial kitchenairflow balancing. The
2、instruments included 4 in. (0.1 m) and2.75 in. (0.07 m) rotating vane anemometers, a hot filmanemometer, a velocity grid, and flow hoods with 2 2 ft(0.61 0.61 m) and 2 4 ft (0.61 1.22 m) hoods. Grease filterconfigurations included conventional baffle filters, cyclonefilters, and slot filters mounted
3、 in an 8 ft (2.4 m) canopy exhausthood. Exhaust airflow rate measurements were made atapproximately 2000, 3000, and 4000 scfm (0.94, 1.4, and1.9 m3/s) for each instrument and each filter type. Measure-ments were made with a gas fryer, half-size electric oven, anda gas underfired broiler idling (hot
4、tests) and turned off (coldtests) to determine the influence of hot appliances. Flow ratesdetermined using calibrated flow nozzles in the supply airsystem agreed to within a few percentage points of thosemeasured using the flow hoods. The other instruments gener-ally did not provide accurate results
5、 unless an appropriatecorrection factor, or K-factor, was used.INTRODUCTIONProper airflow balance is necessary in commercial kitch-ens to capture and contain the cooking effluent within theexhaust system, to allow for proper airflow rate and the result-ing minimal operating energy as the cooking loa
6、d changes,and to provide adequate indoor environmental air quality forthe cook staff and other building occupants. At present, the airbalancing industry and the commercial kitchen ventilationindustry each focus on different hardware, use different instru-mentation, and have developed their own test
7、protocols. TheInternational Mechanical Code (IMC) requires that properairflow rates be verified but offers no acceptable method toaccomplish this requirement. Other related documents includeANSI/ASHRAE Standard 111-2008, ANSI/ASHRAE Stan-dard 154-2003, ASTM F1704-05, and California EnergyCommission
8、Design Guides 1 and 2 (ASHRAE 2003, 2008;ASTM 2005; CEC 2004a, 2004b). The objective of this research is to provide a laboratory-developed/field-validated method of test (MOT) applicable tocommercial kitchens that would include recommended instru-mentation and test procedures to verify the supply an
9、d exhaustairflow rate balance and to measure the associated air pressuredifferences. This paper summarizes the laboratory testsconducted on three types of grease filters installed in a canopyexhaust hood. TEST FACILITYThe test facility is located at the Thermal EnvironmentalEngineering Laboratory in
10、 the Department of MechanicalEngineering at the University of Minnesota. It was initiallyconstructed for ASHRAE RP-745, and subsequently used forASHRAE RP-1375 to characterize the emissions from variouscommercial kitchen cooking appliances and representativefood products using an 8 ft (2.4 m) canopy
11、 exhaust hood,exhaust duct, and exhaust fan. Details of the initial configura-tion are contained in the ASHRAE RP-745 Phase II, FinalReport. Performance Evaluation of Handheld Airflow Instruments Applied to Commercial Kitchen Exhaust SystemsThomas H. Kuehn, PhD, PE Bernard A. Olson, PhDFellow ASHRAE
12、Kevin Campbell Andrew J. HawkinsonStudent Member ASHRAEThomas H. Kuehn is a professor, Bernard A. Olson is a senior research associate, and Kevin Campbell and Andrew J. Hawkinson areresearch assistants in the Department of Mechanical Engineering, University of Minnesota, Minneapolis, MN.LV-11-004 (R
13、P-1376)2011. American 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
14、ASHRAES prior written permission.2011 ASHRAE 691Significant modifications were required for the presentwork. Most were required for airflow handling and monitor-ing. Constraints on the length of the exhaust duct due to labo-ratory size prevented accurate direct measurements of theexhaust airflow rat
15、e. Thus, accurate flow rate measurementscould only be measured on the supply side. With an airtightfacility, the supply and exhaust flows rates would be equalwhen no heat sources were used and no additional flow wasadded. Several changes were made to the existing kitchen facil-ity, including efforts
16、 to make the entire kitchen airtight, relo-cating the exhaust duct collar on the hood to the right-rearcorner, and moving the hood so it was flush with the right-handwall. This wall acted as a symmetry plane so the hood wouldsimulate airflow entering a 16 ft (4.8 m) long hood; the 8 ft(2.4 m) hood w
17、as actually operational and tested while theother 8 ft (2.4 m) was the imaginary half. In addition, twoacrylic plastic windows were installed in the right and leftwalls near the front lower edge of the hood for optical evalu-ation of capture and containment of the cooking effluent. To provide measur
18、ed amounts of makeup air that repre-sented supply air from remote diffusers and transfer air fromother parts of the building, existing vinyl mesh screens thatformed the back wall of the facility were replaced by twogeneral air supply units. These were designed to be displace-ment air units with low
19、discharge velocity and formed the rearwall of the test kitchen. One or both could be removed to allowfor installation and removal of appliances, makeup air units,hoods, and any other large piece of equipment. Air wassupplied through an 8 in. (0.2 m) diameter flexible ductconnected at the top center
20、of each chamber. The air passedthrough an internal diffuser section 30 in. (0.76 m) long, andthen discharged through perforated panels that cover theremaining 79 in. (2 m) height of the chamber. The two sectionswere clamped together and to the outward face of the steelframe of the test kitchen with
21、a gasket interface. A sectionview of the test facility is shown in Figure 1.Air was supplied to the test facility through two sources,the general supply units described above and various dedi-cated makeup air units. A perimeter makeup air unit installedin front of the hood as shown in Figure 1 was u
22、sed in the testsreported here. A single fan served the entire supply air system using avariable frequency drive (VFD) connected to a 7.5 hp (5.6 kW)belt driven fan for airflow control. From a common 16 in.(0.41 m) diameter 72 in. (1.83 m) straight duct, two takeoffducts using 45 degree elbows provid
23、ed air to two parallel16 in. (0.41 m) diameter legs. One leg supplied air to thegeneral supply air units that formed the back wall of the testfacility, and the other supplied the various dedicated makeupair units. In both legs, a Model 2350 Metering Venturi Nozzle,manufactured by Lamba Square, was u
24、sed to monitor airflowrates. A full calibration of each Venturi nozzle was conductedby the Colorado Engineering Experiment Station, Inc. (CEESI),Figure 1 Schematic of commercial kitchen test facility used for flow measurement tests.692 ASHRAE Transactionsby sending the entire supply duct system to t
25、he calibration facil-ity. These nozzles are located 80 in. (2.0 m) downstream fromthe last elbow, and 68 in. (1.7 m) upstream of the dischargebranches. Dampers are located just upstream of the dischargebranches for flow balancing. Additionally, three holes weredrilled 20 in. (0.5 m) upstream of the
26、nozzle flanges in both legsfor velocity traverse measurements using a Pitot tube or a hotfilm anemometer. The combination of the VFD, flow-meteringnozzles, and dampers enabled accurate flow monitoring andcontrol to ensure overall accuracy during testing. Additional modifications were made to the exh
27、austsystem. A 16 in. (0.41 m) diameter, 90 degree elbow connectsthe hood to two 45 degree elbows. A straight section of ductwas welded to the exhaust fan, being a 5 hp (3.7 kW) belt-driven fan controlled using a VFD that discharges horizontallyoutside the building. Three commercial kitchen cooking a
28、ppliances were posi-tioned under the exhaust hood. These were set to idling condi-tions during some of the tests to simulate the thermalconditions that exist just before cooking is initiated. A 96,000Btu/h (28 kW) gas charbroiler was positioned next to the right-hand wall under the exhaust duct coll
29、ar. This represented aheavy-duty appliance. A 5.5 kW half-sized electric convectionoven was placed on a stainless-steel stand next to the char-broiler. The oven could accommodate nine 21 14 1 1/2 in.(0.53 0.36 0.04 m) trays and represented a light-duty appli-ance. An 80,000 Btu/h (23 kW) gas deep fa
30、t fryer was locatednear the left-hand end of the hood and represented a medium-duty appliance. The vat had a 50 lb (23 kg) capacity. Eachappliance was calibrated according to its correspondingASTM standard to provide appropriate idle conditions duringthe hot-air balance tests and the necessary condi
31、tions for full-load cooking during the capture and containment tests. A data acquisition system was established to monitor thevarious static and dynamic variables associated with the testfacility. Five calibrated Rosemount 3031 pressure transducerswere used to monitor the pressure differences across
32、 eachVenturi nozzle, between the interior of the test kitchen facilityand the laboratory, and between the high-pressure side of eachVenturi nozzle and the laboratory. Additionally, severalMagnehelic pressure gauges also displayed pressure differ-ences as backup. Each transducer transmitted an analog
33、 volt-age signal to a Keithley 2700 Multimeter/Data AcquisitionSystem. Six thermocouples sensed temperatures just down-stream of both Venturi nozzles, the ambient laboratorytemperature, the exhaust temperature just upstream of theexhaust fan, the ambient temperature inside the test facility(located
34、near the general supply air unit), and the dedicatedmakeup air unit discharge. All thermocouples were calibratedand also connected to the Keithley multimeter. The data werethen transferred from the Keithley 2700 to a PC computeroperating LabView 7.0. The program read all inputted data,adjusted for i
35、nitialized offsets, converted the raw signals usingthe appropriate calibration equations, and produced a data logand graphical representation of the system features. HANDHELD INSTRUMENTATIONThe following instruments were loaned to the researchgroup by three different manufacturers in “like-new” cond
36、i-tion with fresh calibration documentation:4 in. (0.1 m) rotating vane anemometer (4 in. RVA)2.75 in. (0.07 m) rotating vane anemometer (2.75 in.RVA)Hot-film anemometerTwo velocity grids Two flow hoods with 2 2 ft (0.61 0.61 m) and2 4 ft (0.61 1.2 m) hoodsDetailed instrument specifications are prov
37、ided in theAppendix.PROCEDUREThis section discusses the procedures and instrumenta-tion used to measure the airflows through different styles offilters in exhaust hoods. Some styles of filters could notaccommodate all of the different measurement devices; there-fore, the tests methods for each filte
38、r style are outlined. Baffle FiltersThe baffle filters installed in the exhaust hood are shownin Figure 2. The filter bank consisted of 5 filters, two 18 in.(0.46 m) wide and three 13.75 in. (0.35 m) wide. All of thefilters were slid as far to the right as possible, minimizing gapsbetween them. The
39、filters did not completely fill the hoodopening, leaving a gap on the left. This was closed using a steelplate that was sealed using aluminum tape. A dimensioneddrawing of the filter bank is shown in Figure 3.The airflow rate through the baffle filters was determinedusing a velocity grid, 2 2 ft (0.
40、61 0.61 m) and 2 4 ft(0.61 1.22 m) flow hoods, 2.75 in. (0.07 m) and 4 in.Figure 2 Baffle filters as tested in the exhaust hood.2011 ASHRAE 693(0.10 m) RVAs, and a hot-film anemometer. For the velocitygrid, one data point was taken per filter with the legs orientedtoward the corners of each filter.F
41、or the 2 2 ft (0.61 0.61 m) flow hood, four airflow ratemeasurements were taken along the filter bank adjacent toeach other along the entire length of the hood. A tight seal wasmaintained above and below the filters against the housing.Because the width of the filters is less than the width of thefl
42、ow hood, data were not obtained for each filter. When the2 4 ft (0.61 1.22 m) flow hood was used, two measure-ments were taken, one on each half of the filter bank.For the two RVAs and the hot-film anemometer, discretepoint velocity measurements and a traverse were performed oneach filter. For the d
43、iscrete point measurements, each filterwas divided into 9 equal areas (3 3 matrix) with the velocitymeasured in the center of each area. The traverse patternconsisted of 3 legs for each filter. For the 4 in. (0.10 m) RVAand the hot-film anemometer, a 30 second time constant wasused. The pattern was
44、traversed for 15 seconds each way,resulting in a 30-second total traverse per filter. The longesttime constant available for the 2.75 in. (0.07 m) RVA was 16seconds. Therefore, a 16-second traverse was performed eachdirection, and the two values averaged. The traverse patternused for the 18 in. (0.4
45、6 m) wide filters is shown in Figure 4.Previous velocity measurements with baffle filters usinga 4 in. (0.10 m) RVA reported by Gordon and Parvin (1992)recommended a standoff distance of 2 in. (0.05 m). Therefore,this distance was also used in the present measurements whenusing both RVAs and the hot
46、-film anemometer. The distancebetween each RVA housing and the filter surface was main-tained by taping a short piece of ruler to the exterior surface ofthe housing. The same ruler was taped to the support rod of thehot-film anemometer approximately 2 in. (0.5 m) from theprobe. The velocity grid was
47、 equipped by the manufacturerwith 1.5 in. (0.04 m) standoff legs. These were used for allmeasurements made with the velocity grid.Tests were conducted in triplicate for each instrument atexhaust flow rates of approximately 2000, 3000, and 4000scfm (0.94, 1.4, and 1.9 m3/s) with the appliances turned
48、 off.The entire set of tests was then repeated with the appliancesturned on at idling conditions.The pressure difference was also measured between theinterior of the test kitchen and the plenum behind the filterbank. The kitchen pressure tap was located on the right-handwall that served as the symme
49、try plane for the hood 48 in.(1.2 m) above the floor and 24 in. (0.6 m) in front of the hood.The pressure tap behind the filter bank was located on top ofthe hood 3 in. (0.076 m) from the rear and 18 in. (0.46 m) fromthe left-hand end of the hood. The same pressure tap locationswere used for the cyclone and slot filters.Cyclone FiltersThe cyclone filters tested are shown in Figure 5. The filterbank consisted of five filters, two 19.5 in. (0.5 m) wide andthree 15.5 in. (0.4 m) wide. All of the filters were slid as far tothe right as possible, minim