Case Study: Design of a Calibration Reference Device for ANSI/ASHRAE Standard ME 4054W November 16, 2010 T. H. Kuehn SCREEN CONCEPTS

Transcription

1 Case Study: Design of a Calibration Reference Device for ANSI/ASHRAE Standard ME 4054W November 16, 2010 T. H. Kuehn IDENT OPP GATHER INFO DEFINE PROBLEM GEN CONCEPTS SCREEN CONCEPTS IMPLEMENT HANDOFF DEFINE THE PROBLEM Background UNIVERSITY OF MINNESOTA 1

2 Objective as defined by the members of ASHRAE TC 2.4 Develop a calibration reference device for use in ANSI/ASHRAE Standard 52.2 filter test facilities that can be replicated to serve as a check on a facility s capability to provide the proper performance values for filters being tested. Typical Ventilation System for a Commercial Building Receive filters from each location. Remove filter samples. Elute filter samples in PBS by hand shaking and vortex mixing. Use resulting eluate in molecular and culturable identification methods. 2

3 Primary performance values determined by the ASHRAE 52.2 standard method of test Pressure Drop: measured in inches of water at 2000 cfm of air flow Particle Size Removal Efficiency: measured as fraction of airborne particles removed versus particle size in 12 size bins ranging from 0.3 to 10 microns (using KCl) Example Filter Test Results (ASHRAE Standard Figure D-1) 3

4 Primary Test Output Minimum Efficiency Reporting Value (MERV) Based on average particle removal efficiency in three particle size bins: 0.3 to 1, 1 to 3, and 3 to 10 microns Ranges from 1 (low efficiency) to 16 (high efficiency) ASHRAE 52.2 Test Facility at U of MN Schematic Diagram of ASHRAE Standard 52.2 Test Facility at the University of Minnesota Department of Mechanical Engineering 8 4

5 ASHRAE 52.2 Test Facility at U of MN Photographs of ASHRAE Standard 52.2 Test Facility at the University of Minnesota Department of Mechanical Engineering 9 Why this project? Filter manufacturers were receiving different MERV values for the same filters sent to different test facilities. A Calibration Reference Device was desired so that test facilities could compare their measurements to accepted values using the same device 5

6 IDENT OPP GATHER INFO DEFINE PROBLEM GEN CONCEPTS SCREEN CONCEPTS IMPLEMENT HANDOFF DETERMINING CUSTOMERS Project Monitoring Subcommittee ASHRAE Technical Committee 2.4 (TC 2.4) ASHRAE Standards Committee for 52.2 Operators of Filter Test Facilities that Comply with Standard 52.2 Filter Manufacturers Purchasers and users of particulate air filters in buildings UNIVERSITY OF MINNESOTA Initial Product Design Specifications Number Description Metric Desired Value 1 Air flow rate cfm Pressure drop in. water <1.5 3 Particle size range µm Particle collection efficiency MERV 10 5 Cost $ <10,000 6 Weight lbs <50 7 Dimensions inches 24x24x2 8 Material stainless steel 6

7 CONCEPT GENERATION OPTION 1: INERTIAL IMPACTOR Advantages: Impactor theory well known, therefore easy to design Repeatable and accurate results Foolproof design Robust Disadvantages: Small nozzle sizes for low pressure drop requirement might increase manufacturing costs Potential for solid particle bounce 14 7

8 Impactor Example- Respirable Cut Impactor (Marple and McCormack, 1983) 15 Filters B and C Curves with Three Parallel-Flow Slit Impactor Efficiency Curves Particle Size Removal Efficiency, % Minimum Efficiency Curves Filter B Filter C Impactor Theory Particle Diameter, µm 16 8

9 OPTION 2: ELECTROSTATIC PRECIPITATOR (ESP) Advantages: Low pressure drop Robust Ease of cleaning Disadvantages: Would need to alter commercial unit design to meet requirements Charged particles upstream of ESP may alter collection efficiency characteristics Humidity effects may alter collection efficiency characteristics 17 ESP Calibration Using KCl at 1,750 CFM (provided by Air Quality Engineering) 100 Efficiency vs. Particle Diameter P=.261 H20 80 % Fractional Efficiency Particle Diameter ( µ m) 18 9

10 Two Reference Device Alternatives Selected Based Upon Literature Review, Further Analysis and Discussions with Committee Members 1. MULTIPLE ROUND NOZZLES, PARALLEL-FLOW IMPACTOR. 2. STAINLESS STEEL WIRE MESH. Impactor Design Considerations- Nozzle vs. Sharp-Edged Orifice Numerical Model: Re = 62,000 S/W = 1.0 T/W = 1.0 Flow Streamline Tapered Inlet Nozzle Impaction Plate 20 10

11 Impactor Design Considerations- Nozzle vs. Sharp-Edged Orifice Numerical Model: Re = 62,000 S/W = 1.0 T/W = 0.03 Flow Streamline Orifice Impaction Plate 21 Change to PDS The impactor with orifices requires 3 in. water to operate so the ASHRAE monitoring subcommittee allowed us to increase the allowable pressure drop from 1.5 to 3 in. water after consulting with a sample of test facility operators to make sure their fans could handle this increase in pressure drop. 11

12 Impactor Design Considerations- Cross-flow Parameter DnN/4Dc < 1.2 Where: Dn is the nozzle diameter N is the number of nozzles Dc is the nozzle cluster diameter 23 Collection efficiency curve for five parallel-flow nozzle impactor design showing sampling points of OPC 100 Parallel-FLow Impactor with Five Nozzle Sizes Particle Size Removal Efficiency, % um Cutsize Impactor Collection Efficiency OPC Mean of Particle Size Range 3.0 um Cutsize 1.66 um Cutsize 1.0 um Cutsize 0.71 um Cutsize Particle Diameter, µm 24 12

13 Scaled-Down Impactor Design- Nozzle Plate in. Thick Fabricating Holes in Nozzle Plate in. Stainless 1. Drilling: too time consuming and expensive for over 60,000 holes for full scale unit 2. Punching: we tried this and the material deformed, we also broke the smallest punch 3. Chemical etching: this worked the best, different sized holes can be fabricated from a single sheet of material, hole diameter standard deviation very low 13

14 Scaled-Down Impactor Design- Impaction Plate Wire Mesh Specifications Start with 0.04 mm/0.07 mm stainless steel wire mesh. Specifications based upon previous work by Palas from their presentation Influence on Measurement of Fractional Efficiency on Test Systems in Accordance with DIN (provided by Bruce McDonald). Sample obtained from G. Bopp USA, Inc. for testing

15 Wire Mesh Specifications 29 Wire Mesh Calibration by Palas 30 15

16 Reference Devices: Selection Two Small-Scale Reference Device Candidates Ready for Testing PARALLEL-FLOW IMPACTOR-ETCHED STAINLESS STEEL WIRE MESH Reference Devices: Calibration Method The calibration method for scaled-down reference device testing included: Generate monodisperse fluorescent particles using VOAG in the size range from approximately 0.7 to 10 um Wash components followed by fluorometric analysis to determine collection efficiency as a function of particle size Use Aerodynamic Particle Sizer for quality assurance 32 16

17 Parallel-Flow Impactor: Calibration Results Each nozzle set was calibrated independently at 10 cfm with oleic acid droplets Numerous particle sizes were run for each nozzle set to determine the 50% collection efficiency size For both 1.66 um cut stage and 1.00 um cut stage, the 50% collection efficiencies were approximately 30% larger than anticipated Collection Efficiency (percent) Collection Efficiency (percent) um Cut Stage Collection Efficiency 1.66um Cut Stage Effciency Linear (1.66um Cut Stage Effciency) *50% Efficiency at 2.13um 1 10 Aerodynamic Particle Diameter (um) 1.0um Cut Stage Collection Efficiency 1.0um Cut Stage Efficiency Linear (1.0um Cut Stage Efficiency) *50% Efficiency at 1.31um Aerodynamic Particle Diameter (um) Parallel-Flow Impactor: Calibration Results During calibration, efficiencies for the smallest nozzle set were considerably lower than expected Further calibration points were run with threequarters of the nozzles plugged at 2.5 cfm In comparison of the two data sets, it was determined cross-flow interaction reduced the efficiency of the 0.71 um cut stage Collection Efficiency (percent) Cut Stage Collection Efficiency Comparison: Unplugged Vs. 3/4 Plugged Nozzle Sets Unplugged Nozzle Set 3/4 Plugged Nozzle Linear (3/4 Plugged Nozzle) *50% Efficiency at 1.55um Aerodynamic Particle Diameter 17

18 Parallel-Flow Impactor: Summary Designed Cut: 0.71um Actual Cut: 1.55um Percent Difference: 119% Designed Cut: 1.00um Actual Cut: 1.31um Percent Difference: 31.0% Designed Cut: 1.66um Actual Cut: 2.13um Percent Difference: 28.3% Both 1.66 um and 1.00 um cut stages were determined to have approximately 30% larger cut points Effects of cross-flow greatly lowered the efficiency of the.71um stage Adjustments to nozzle sizing and spacing will be made, and with the addition of a body-impactor plate upstream of the parallel nozzle plate, all intended cut points should be achieved Wire Mesh: Specifications Three different Twilled Double Dutch Weave wire mesh sizes were inspected The 325 x 2300 wire mesh was determined to best meet the project requirements 36 18

19 Wire Mesh: PCL Calibration Results Calibration tests were run with solid particles at various sizes for the 325 x 2300 wire mesh Trends for solid particle testing were consistent with particle bounce phenomenon Further calibration points were run with liquid oleic acid particles which were tagged with uranine dye. In comparison of the two data sets, particle bounce was determined to be the cause of the difference between the collection efficiencies Collection Efficiency, % Twilled Double Dutch Weave Mesh 325x2300 Comparison Between Solid and Liquid Particle Collection Efficienies Solid Particles Liquid Particles Aerodynamic Particle Diameter (um) Wire Mesh: Summary 325 x 2300 fine wire mesh determined to best meet project requirements Particle bounce greatly reduced collection efficiency therefore liquid droplets were used for calibration Optimized pleated design would reduce pressure drop across wire mesh 19

20 Wire Mesh: Summary Particle bounce greatly reduced collection efficiency therefore liquid droplets were used for calibration As ASHRAE Standard 52.2 requires the use of solid KCl particles, we recommend that the wire mesh option be dropped and the remaining effort be focused on the parallel stage impactor design Parallel-Flow Impactor: Designed Efficiency Pleated design impactor theoretical collection efficiency curves and actual data based upon scaled-down impactor calibrations showing sampling points of the OPC Particle Size Removal Efficiency, % Impactor Th. Collection Efficiency OPC Mean of Particle Size Range 0.84 um Cut Stage Data 1.88 um Cut Stage Data 4.70 um Cutsize 4.70 um Cut Stage Data 1.88 um Cutsize 0.84 um Cutsize Particle Diameter, µm 40 20

21 Full-Scale Reference Device Recommend 3-stage parallel flow impactor with cut sizes of 0.84, 1.88 and 4.7 microns (not enough room for five stages without significantly increasing size and weight) Recommend pleated design (W) to increase surface area for nozzle plates and impactor plates (increases depth of unit in airflow direction) 41 Revised Product Design Specifications Number Description Metric Desired Value 1 Air flow rate cfm Pressure drop in. water <3.0 3 Particle size range µm Particle collection efficiency MERV 10 5 Cost $ <10,000 6 Weight lbs <50 7 Dimensions inches 24x24x12 8 Material stainless steel 21

22 Parallel-Flow Impactor: Design 2,000 CFM Full-Scale Pleated Design End View 0.84 µm Cut Stages 4.70 µm Cut Stages 0.84 µm Cut Stages 0.84 µm Cut Stages 1.88 µm Cut Stages Reference Devices: Selection 2,000 CFM Full-Scale Pleated Design Dimensions: 23 5/8 long x 23 5/8 wide x 7 1/8 deep Isometric View Estimated weight: 50 lbs Material: 316 Stainless-steel (Aluminum for prototype) 22

23 Recommendations Build and test full-scale 2,000 cfm pleated impactor reference device Abandon wire mesh concept because of particle bounce problems 45 Parallel-Flow Impactor: Design 2,000 CFM Full-Scale Pleated Design Orifice Centerline Spacing to Diameter Ratio Number Design Orifice Number of Total Flow, of Stages Cutsize, µm Dia., in. P, in. H2O Orifices cfm , , Total: 2, Number of Strips Note: The orifice spacing in the vertical direction had to be decreased by 15% from the small-scale design for the 0.84 and 1.88 um cut stages and 20% from design for the 4.70 um cut stage to fit all the nozzles into the full-scale design

24 Parallel-Flow Impactor: Design 2,000 CFM Full-Scale Pleated Design Top View 4.70 µm Impaction plate slot Support µm Impaction plate slot 0.84 µm Impaction plate slot Support-2 Support-3 Support µm Impaction plate slot 1.88 µm Impaction plate slot Parallel-Flow Impactor: Design 2,000 CFM Full-Scale Pleated Design IsometricView Frame Top and Bottom Plates Support-3 Support-2 Support-4 Support-1 24

25 Parallel-Flow Impactor: Design Top and Bottom Plate Dimensioned Drawing Parallel-Flow Impactor: Design Photo of Full-Scale Prototype (Front View) Photo of Full Scale Impactor: Front View 25

26 Parallel-Flow Impactor: Design Photo of Full-Scale Prototype (Top View) Photo of Full Scale Impactor: Top View Full-Scale Reference Device: Calibration Test Results with Polydisperse KCl aerosols: 52 26

27 Summary Full-scale pleated impactor designed, manufactured and tested Pressure drop 20% more than anticipated at 2000 cfm (3.7 in. water) MERV rating less than anticipated with low collection efficiency for the large particles (MERV = 8) Some construction details need to be changed: Impaction plate mounting and end plate seals Mott material attachment 53 Summary Number Description Metric Desired Value Measured Value 1 Air flow rate cfm Pressure drop in. water < Particle size range µm Particle collection efficiency MERV Cost $ <10,000 15,000 6 Weight lbs < Dimensions in. water <24x24x x23.8x12 8 Material stainless steel stainless steel/al 54 27

28 Summary ASHRAE Project Monitoring Subcommittee will discuss results and make a decision whether to adopt this as the desired Calibration Reference Device for use at ANSI/ASHRAE test facilities. This will occur at the ASHRAE Winter Annual Meeting, Las Vegas, January If yes, the University of Minnesota has agreed to fabricate copies at cost and ship them to purchasers. If no,???? 55 28