INSTITUTE OF ENVIRONMENTAL SCIENCES AND TECHNOLOGY Contamination Control Division Recommended Practice 034.2 IEST-RP-CC034.2 __ ee HEPA and ULPA Filter Leak Tests INSTITUTE OF ENVIRONMENTAL SCIENCES AND TECHNOLOGY 5005 Newport Drive, Suite 506 ~— Rolling Meadows, IL 60008-3841 Phone: (847) 255-1561 « Fax: (847) 255-1699 it E-mail: iest@iest.org + Web: worw est ongThis Recommended Practice is published by the INSTITUTE OF ENVIRONMENTAL SCIENCES AND TECHNOLOGY to advance the technical and engineering sciences. Its use is entirely voluntary, and determination of its applicability and suitability for any particular use is solely the responsibility of the user. ‘This Recommended Practice was prepared by and is under the jurisdiction of Working Group 034 of the IEST Con- tamination Control Division. Copyright © 2005 by the INSTITUTE OF ENVIRONMENTAL SCIENCES AND TECHNOLOGY Second printing, July 2005 ISBN 0-9747313-4-X PROPOSAL FOR IMPROVEMENT: The Working Groups of the INSTITUTE OF ENVIRONMENTAL SCIENCES AND TECHNOLOGY are continually working on improvements to their Recommended Practices and Reference Documents. Suggestions from those who use these documents are welcome. If you have a suggestion regarding this document, please use the online Proposal for Improvement form found on the IEST website at ‘www dest org/proposal/form html INSTITUTE OF ENVIRONMENTAL SCIENCES AND TECHNOLOGY 5005 Newport Drive, Suite $06 Rolling Meadows, I. 60008-3841 Phone: (847) 255-1561 Fax (847) 255-1699 Email: iesiest org» Web: wares org TESTRP-CCOB2 [RSTITUTE OF Envinowuestal Samers aso TecmnoLnoy Copyrighted © mteok 3HEPA and ULPA Filter Leak Tests TEST-RP-CC034.2 CONTENTS SECTION 1 SCOPE AND LIMITATIONS. : ss sense REFERENCES a 7 2 3. TERMS AND DEFINITIONS 4 BACKGROUND AND PURPOSE 5 CHOICE OF TESTS 6 RECOMMENDED TEST PROCEDURES. TABLE 1 GUIDE FOR SELECTING APPROPRIATE LEAK TESTING STRATEGIES. vena I, 2. DESIGNATED LEAK SIZES FOR VARIOUS EFFICIENCIES OF FILTERS. : 0 E1 PROBABILITIES OF MISSING LEAKS AND FALSE INDICATIONS... — 35 FIGURE 1 EXAMPLE OF A SAMPLING PROBE. E1_ EXAMPLE OF A SAMPLING PROBE. 2. EFFECT OF THE NUMBER OF PARTICLES USED TO CHARACTERIZE A DESIGNATED LEAK..35 I1_ EFFECT OF MEDIA VARIABILITY ON LEAK DETECTION... APPENDIX A. CONSIDERATIONS IN LEAK TESTING B_ CALCULATION OF AEROSOL CHALLENGE...... © EXAMPLE CALCULATION OF SPATIAL AND TEMPORAL CHALLENGE UNIFORMITY. D E EXAMPLE CONDITIONS FOR FILTER LEAK SCAN TEST: AEROSOL PHOTOMETER METHOD...32 METHODS OF CALCULATION AND EXAMPLE CONDITIONS FOR FILTER LEAK SCAN TEST: DISCRETE-PARTICLE COUNTER TEST METHOD. se 38 DILUTION OF AEROSOL FOR PARTICLE COUNTERS... EXAMPLE CONDITIONS FOR FILTER LEAK SCAN TEST: DISCRETE PARTICLE COUNTER METHOD H__ MEASURING THE RESPONSE TIME OF PHOTOMETERS. 1 EXCESSIVE WIDESPREAD NON-SITE-SPECIFIC PENETRATION IN LEAK TESTING J BIBLIOGRAPHY. | a MEST. RP-ccusaa INSTITUTE OF ENVRONOENTAL SCENCES AND THANOLOGY Coprighed © material 5INSTITUTE OF ENVIRONMENTAL SCIENCES AND TECHNOLOGY Contamination Control Division Recommended Practice 034.2 HEPA and ULPA Filter Leak Tests IEST-RP-CC034.2 1 SCOPE AND LIMITATIONS 1.1 Scope ‘This Recommended Practice (RP) covers definitions, equipment, and procedures for leak testing high- efficiency particulate air (HEPA) filters and ultralow- penetration air (ULPA) filters in the factory as they are produced, at the job site before they are installed, and after they are installed in cleanrooms and in unidi- rectional-flow, clean-air devices. When used in con- junction with other RPs, including IEST-RP-CC0O1, TEST-RP-CC002, IEST-RP-CC006, IEST-RP-CC007, and IEST-RP-CC02I, this RP may be used to define the basis of an agreement between customer and sup- plier in the specification and procurement of HEPA, and ULPA filters, and in the testing of unidirectional flow, clean-air devices and cleanraoms. This RP also includes procedures for measuring the uniformity of the aerosol challenge approaching the filter under test. 1.2 Limitations ‘This document does not cover in-place testing of banks of filters in nuclear power or nuclear research applications, nor does it cover biological safety or containment cabinets except for scanning of the fil ters for leaks. This document does not cover Jeak testing of filters in high-temperature environments ‘This document provides values of acceptance limits” - for guidance; however, it is the responsibility of the customer and supplier to specify which leaktest method is used and the acceptance criteria for each application, NOTE: Testing in accordance with this RP may involve hazardous materials, operations, and equipment. This RP does not purport to address all of the safety problems associated with its use TESTRP-CC0342 STITUTE OF ENVIRONMENTAL SCIENCES ANDTHEHROLOGY Ic is the responsibitity of the user to consult and establish appropriate safety and health practices and to determine the applicability of regulatory limitations prior to use of this RP 2 REFERENCES ‘The following documents are incorporated into this RP to the extent specified herein, Users should apply the most recent editions of the references, 2.1 American Society of Mechanical Engineers (ASME) ASME N50: Testing of Nuclear Air Treatment Systems 2.2 Institute of Environmental Sciences and Technology (IEST) IEST-RP-CCOO1: HEPA and ULPA Filters IEST-RP-CC002: Unidirectional Flow Clean-Air Devices IEST-RP-CCO06: Testing Cleanrooms IEST-RP-CCO07: Testing ULPA Filters IEST-RP-CCO2I: Testing HEPA and ULPA Filter Media 23 ISO ISO 14644-1 Cleanrooms and associated controlled environments—Part I: Classification of air cleanliness 2.4 National Sanitation Foundation NSF/ANSI Standard 49: Class II (Laminar Flow) Bio- safety Cabinetry (Coprighd © motel 72.5 Sources and Addresses Referenced documents may be obtained from the following sources: ASME ‘ASME International ‘Three Park Avenue New York, NY 10016-5990, USA Phone: 212-591-7722 Fax: 212-591-7674 Web: www.asme.org TEST Institute of Environmental Services and Technology 5005 Newport Drive, Suite 506 Rolling Meadows, IL 60008-3841, USA Phone: 847-255-1561 Fax: 847-255-1699 Webs: ww iest.org Iso In the U.S.: Documents available from lest ‘5005 Newport Drive, Suite 506 Rolling Meadows, Illinois 60008-3841, USA ‘www iest org Outside the U.S.: Documents available from representative ISO member body organization NSF NSF Intemational 789 N. Dixboro Road P.O. Box 130140 Ann Arbor, MI 48113-0140, USA Phone: 734-769-8010 Fax: 734-769-0109 Web: www.nsforg 3 TERMS AND DEFINITIONS The following terms have special meaning in the con- {ext of this RP: aerosol A gaseous suspension of fine solid or liquid particles. aerosol photometer ‘An instrument that measures mass concentrations of aerosol using the forvard light scattering principle. For the applications described in this RP, the instru- iment should be capable of measuring aerosol concen- trations of up to 100 mg/m’ and have a threshold sensitivity capable of measuring 1 wgimn? calibrated equipment ‘Test equipment that has been calibrated in accor- dance with the manufacturer’s recommendations or accepted industry practices. challenge aerosol ‘An aerosol derived from the selected aerosol source Imaterial and used as the leak test challenge for filter ‘mediuim leak test. Challenge aerosols may be produced by a variety of methods, the choice determined by the aerosol type and the particle size characteristics count or number mean diameter of aerosol The count mean diameter is the average particle size of the number distribution of the aerosol. count or number median diameter of aerosol (CMD) ‘The median diameter is the 50th percentile of the umber distribution of the aerosol; ie., 50% of the particles are smaller than the count median diameter and 50% are larger than the count median diameter. clean-air device, unidirectional flow ‘A clean work station or other device (other than a cleanroom or clean zone) that incorporates one ot more HEPA or ULPA filters and one or more motor= blowers for the purpose of supplying unidirectional- flow, clean air to a controlled work space. clean zone ‘A defined space in which the concentration of air- bore particles is controlled to specified limits, designated leak AA leak that should be detectable during scanning of a filter. When using a photometer, a designated leak is characterized by a standard leak penetration, which is the photometer response when the sampling probe is stationary over a leak, divided by the photometer re- sponse upstream of the filter when using a 28.3-L/min (1-#imnin) photometer. When using a discrete-particle counter, a designated leak is characterized by a stan- dard leak penetration and by a designated number of counts chosen to establish statistical probabilities re lated to its detection. dituter ‘A device that reduces the particle concentration by the dilution ratio of the device. These devices typicelly mix a known volume of sample air with a known vol- ume of particle-fee or filtered air to achieve the dilu- tion. The ratio of the two volumes is the dilution ratio. discrete-particle counter An instrument, such as an optical particle counter or condensation nucleus counter, capable of resolving responses from individual particles DOP or DEHP (di-2-ethylhexyphthalate) ‘The mineral oil di octyl phthalate or DEHP (2-ethyl hexyl phthalate) CAS# 117-81-7. With reference to filter testing, DOP also refers to a poly disperse aero- sol ofthe described material 8 Copyrighted © mera WSTTTUTE OF ENVIROSOIENTAL SCIENCES AN TECIOLOGY TESTRPCOOA?field bench tests Leak tests performed on HEPA or ULPA filters after unpacking at the location of use prior to installation in the clean-air device or cleanroom. A portable test system is used, HEPA (high-efficiency particulate air) filter ‘An extended-medium, dry-type filter in rigid frame ‘when tested at rated airflow having a minimum parti- cle collection efficiency of 99.97% for 0.3-jm mass ‘median diameter particles of DOP when tested in accordance with MIL-STD-282. The foregoing is the traditional definition for the HEPA filter. Changes in the filter manufacture, end user requirements, and testing have introduced products that vary parameti- cally from filters that have been considered standard according to this definition. Additional information ‘on this topic is given in IEST-RP-CCOOL Laskin nozzle ‘Arnozaie used as part of a system to generate a hetero- geneous aerosol ffom a liquid, such as PAO, that uses 8 source of compressed gas as shown in NSF/ANS] Standard 49. Laskin-nozzle-generated aerosol_ ‘An aerosol generated by a Laskin nozzle from liquid diocty! phthalate or other oil. The aerosol has a light- scattesing mean diameter (mass median) of approxi- imately. 0.7 wn, with a geometric standard deviation of about 8, Such an geal bas an approximate comt nen lame of 04 wo leak penetration ‘When using a photometer, measured leak penetration is the photometer response when the probe is station- ary over @ leak, divided by the photometer response ‘upstream of the filter. When using a discrete-particle counter, leak penetration is the concentration of parti- cles counted when the probe is stationary over a leak, divided by the concentration of particles that would be ‘counted upstream of the filter in the same time if the ‘counter could sample upstream. Measured leak pene tration depends on the photometer or diserete-partcle counter flow rate. The leak can be defined as the por- tion (percentage) of the test instrument’ sample flow that passes unfiltered through a defect in the filter or filter medium, ‘mass mean diameter of aerosol ‘The mass mean diameter is the average particle size of the mass distribution of the aerosol. Typically, because the mass of a particle varies as the cube of its diameter, most of the mass of an aerosol tends to be in the larger sizes, resulting in the mass mean being larger than its count or number mean, mass median diameter of aerosol (MMD) ‘The mass median diameter is the 50th percentile of the mass distribution of the aerosol; ie., 50% of the mass of the aerosol is made up of particles smaller than the mass median diameter and 50% of the mass of the aerosol is made up of particles larger than the ‘mass median diameter. most penetrating particle size (MPPS) ‘The particle size at which a given filter has its highest, penetration (or lowest efficiency); i., the worst-case particle size with respect to filtration efficiency. From filtration theory, efficiency is higher for particle sizes smaller or larger than the MPPS. In practice, the MPPS of a filter is considered to be within a measured size range typical in commercial instruments rather than at ‘one unique particle size. PAO (poly Alpha Olefin) ‘The mineral oil PAO CAS# 68649-12-7. With refer- ence to filter testing, PAO also refers to a poly dis- perse aerosol of the described material, photometer response time constant The time it takes for the photometer reading to reach 63% of its final value when there is a step change in aerosol concentration at the probe inlet (ie., the time constant of the exponential response of the photome- ter). The response time is a-function of the photome- ter design, the photometer scale used (e.., full scale equals 0.1%), and the sampling system, PSL. Polystyrene latex particles in a suspension that can be used to generate a monodisperse solid aerosol. PSL is useful for leak testing fillers where other aerosols may contaminate the filters in an unacceptable manner. representative upstream sample ‘A sample of the aerosol upstream of the filter that is being leak tested. This aerosol sample is used to set the 100% point of a photometer or to calculate values for discrete-particle counter scanning. In factory test systems, in field bench test systems, and in unidirec- tional, clean-air devices, a method of obtaining a rep- resentative upstream sample should be built into the device by the device manufacturer. sample acquisition time The time that the particle counter accumulates counts before it automatically resets to zero and starts count ing again scanning ‘A method for disclosing leaks in HEPA and ULPA filters whereby the inlet of the sampling probe of an aerosol photometer or discrete-particle counter is moved in a series of parallel, slightly overlapping, strokes across the test area at a distance of approxi- TEST RP-CCOD INSTITUTE OF ENYROWaNTAL SCIENCES AND TECHNOLOGY Copyighed Dmacral 9mately 2.5 cm (1 in.) from the filter face and at a rate based on the leak penetration to be detected and the Upstream concentration of the challenge aerosol standard leak penetration ‘The leak penetration that would be measured by a photometer or particle counter with a 28.3-Limin (L-f/min) sample flow rate when the sample prot stationary over the leak, super ULPA (ultrs-low-penetration air) filter ‘An extended-medium, dry-type filler in a rigid frame, made with a filter medium having a minimum particle collection efficiency of > 99:9999% (ie,, a maximum particle penetration of < 0,0001%) for particles of the ‘most-penetrating particle size (MPPS) when tested at the average medium velocity in the fiter in accordance with the methods of IEST-RP-CC021 thermal generator A device that produces a liquid aerosol by means of ‘an evaporation-condensation process in the absence of controlled nucleation. The particle size distribution and mass median diameter (MMD) of an aerosol pro- duced by a thermal generator are measurably smaller than those of an aerosol from a Laskin nozzle. thermally generated aerosol ‘An aerosol generated by quenching (condensing) vapor that has been evaporated from mineral oils such as DOP or PAO by heat. The aerosol has-a.mass ‘median diameter (MMD) of approximately 0.3 Hm with a geometric standard deviation of about Such an aerosol has a count median diameter {cmp} smaller than 0.2 jim. This aerosol is also often re- ferred to as “hot” DOP or PAO. ULPA (ultra-low-penetration air) filter A throwaway, extended-medium, dry-type filter in a rigid frame, having a minimum particle-collection efficiency of 99.99% (that is, a maximum particle penetration of 0.001%) when tested in accordance With the methods of IEST-RP-CC007. unidirectional airflow Ai that flows in a single pass in a single direction through an air device or clean zone with generally parallel streamlines. In past cleanroom literature, this was referred to as laminar airflow. work zone The volume within the clean-air device or cleanroom that is designated for clean work and for which test- ing is required. The work zone volume is defined by an entrance plane and an exit plane (normal to the airflow where there is unidirectional airflow), 4 BACKGROUND AND PURPOSE HEPA filters and ULPA filters such as those used in cleanrooms and clean benches are tested for particle removal efficiency and resistance to airflow when they are manufactured, Frequently, these filters are scanned or probed for leaks as an additional precau- tion. This scanning is performed by the filter manu- facturer before the filters are packaged and shipped. The scan test may be repeated as a field bench test When the filters are unpacked at the customer's loca- tion before the filters are installed. If product is ex posed or located immediately downstream of the HEPA ot ULPA filter(s), a scan test is often per- formed after the filters are installed to verify the in- tegrity of the filter and its installation. Typically, the scanning is repeated at regular time intervals to verify the continued integrity of the filtration system. If the filters are remote to the cleanroom (clean zone), atest to determine the total leakage of the filter installation is adequate Leak testing HEPA and ULPA filters is one impor- tant part of verifying their performance, Other tests, such as those described in IEST-RP-CCO01, IEST= RP-CC002, IEST-RP-CC006, and IEST-RP-CCO007, are also important for characterizing the performance ‘of HEPA and ULPA filters, clean-air devices, or cleanrooms. ‘The choice of which test procedures are required de- pends on the performance level of the filters, clean- air device, or cleanroom, and on the intended applica- tion. Specific test requirements are the result of an agreement between customer and supplier. This RP provides a set of recommended test proce- dures for testing HEPA and ULPA filters for leaks. ‘Two methods are provided for scanning the down- stream surface of HEPA and ULPA filters for leaks, One method is provided for determining if there are leaks in inaccessible HEPA and ULPA filters. Sec~ tion 5 provides guidance for selecting the appropriate leak-test method, When purchasing HEPA or ULPA filters, ora clean- air device or cleanroom containing HEPA or ULPA filters, the appropriate sections of this document should become part of the purchase order to define the expected performance of the filters as received. 10 Copighed © materiel [STITUTE oF FRONTAL SCIENCES AND TECHNOLOGY TESTRP-COW25 CHOICE OF TESTS This section provides guidance for selecting leak-test methods, materials, and criteria, Selection of a spe- cific method, aerosol material, and criterion is ulti- mately the responsibility of the user. The test method, aerosol material, and criteria should be an explicit part of the agreement between the customer and sup- plier for the filters, clean-air device, cleanroom, or testing services. Table 1 summarizes the selection process. NOTE: The test procedures deseribed in this RP involve introducing an aerosol into a HEPA or ULPA filter, a clean-air device or cleanroom, or into the environment of @ clean-air device or cleanroom, HEPA and ULPA filters, clean-air devices, and cleanrooms are used to provide clean environments. Introduction of the aerosols, ‘may contaminate the filter, the elean-air device fr cleanroom, equipment in or around the clean- air device of cleanroom, product in or around the clean-air device or cleanroom, or the surrounding area. The subject of which, if any, aerosols are acceptable should be part of the agreements be- tween customer and supplier. Aerosols to be used when testing clean-air devices should be specified in agreements with contractors testing the clean-air device. Examples include but are not limited to: * DOP * DOS, dioctyl sebacate, also known as DEHS, di(2-ethylhexyl)sebacate ‘© mineral oil * oleic acid + PAO (patents conceming the use of PAO for testing filters are listed in Appendix J) * PSL The use of ambient acrosols is not acceptable. The size distribution, concentration, and material of am- bient aerosols vary from location to location and ‘rom time to time. The variability of ambient aerosols ‘means that the leak-test results are not repeatable. ‘The concentrations are normally too low. The un- known characteristics of ambient aerosol make it difficult to establish leak criteria. It should also be bbome in mind that repeated leak testing with oil acro- TesTR-ccwe2 INSTITUTE OF FavinOMMENTAL SCINGHS AND TEOHNOLOGY sols may lead to filter loading. Increase in resistance to airflow due loading may be observed in some ccases, especially when newer filter media such as expanded polytetrafluoroethylene (PTFE) membranes are used in the construction of the filter NOTE: Some particles will penetrate the filter medium even in the absence of any leak. The par- ticles penetrating the unbroken medium will be detected by the photometer or particle counter used to scan the filter. The reading or counts from the particles penetrating unbroken filter medium form a background level. In order for a scanning. procedure to be practical, the designated-leak penetration criterion must be substantially greater than the background level. I is recommended that the designated-leak penetration. be -minimum of ive times the background level. ‘The efficiency of HEPA and ULPA filters is a strong _function_of particle size and airflow velocity. There- fore, the background level is a strong function of the aerosol and filter flow rate used for scanning. There are numerous combinations of particle size, flow rate, designated-leak criteria, filter efficiency, and scanning methods that work. Listing all of the possibilities with their related recommended practices is outside the scope of this RP. Details on selected situations are given in Appendix A. In an unusual situation, one may tse the basis of the methods outlined inthis RP to de- ‘velop an appropriate leak testing practice. 6 RECOMMENDED TEST PROCEDURES ‘The operating instructions provided by the test equipment manufacturer should be followed. All test instruments should be calibrated per manu- facturers’ instructions. The accuracy of all measure- ments is +10% of reading unless otherwise specified. When leak testing filters, it is important. thatthe roper flow rate through the filters. be established prior to testing. In factory or field bench tests, the uniformity of airflow approaching the filter must be established as part of qualifying the test system. When testing filters in installations, airflow volume and velocity tests and balancing should precede the challenge uniformity tests and leak tests, IEST-RP- CC002 and IEST-RP-CC006 provide examples of airflow uniformity and volume flow tests Copyrighted Omatenal ‘11Table I—Guide for selecting appropriate leak testing strategies. Condition Aerosol Photometer Particle Counter using PSL* Filter Type i __ : _ [ HEPA (Type AD Yes! To total leakage test method published HEPA (Type B) _ Yes No total leakage test method published HEPA (Type C) - Yes: - Caution? HEPA (Type D) Yes : Cation’ HEPA (Type E) [ No test 7 No test_ ULPA (Type F) 96.990% Yast Yest | ULPA (Type F) 98.9905%_ Caution i Yes! ‘Super ULPA (Type G) No. 2 Yes! HEPA (Type H) _ Yes? —| No otal eakage test method published HEPA (Type!) : Yes? No total leakage fest method published [HEPA (Type J) Yes! : Caution HEPA (Type K) 7 Yes® vestimte UselApplication : _ - ISO Giass 6 andless dean Yes Nlotrecommended 180 Class § Yes _ Yes 180 Class 4 Caution’ : Yes 7 TSO Glass 3 and deaner No z= Yes Remote Mount Fillers we fea ye No tlk ot metiod published “V"Bank Firs eee Noto leakage test method published ‘Semiconductor, disk dive, opis Caution’ 6 Yes? | Pharmaceutical Yes_ Caution — Test Method/Location Factory yes Yer Bench test Yes Yes: | Manual or Automatic Scanning __Yes Yes _j Particle Counter : _ 7 Counter must have continuous or ong 2-1 count defines leak (Np) Not Applicable duration count mode with beep output a pet each particle detected ‘Aerosol Material ; Poly Styrene Latex PSL io Yes, PAO CAS# 68649-12-7 Yes a Caution? DOP CASH 117-81-7 Yes Caution? Other CustomeriSuppler Agreement | Customer/Suppher Agreement” —_] NOTES: + By agreement between customer, supplier and fier testing agency, other aerosol material may be used 2 Test using total leakage test method (section 6.2.3). Fler may not pass a filer media scan test. > Leak citeria must be less stringent than 0.01%, especially if particle size near the most penetrating particle size (MPS) is used ‘Membrane type ULPA fiters (PTFE or ePTFE) cannot accept ci-based aerosols. § Photometer must be sensive enough to detect below designated leak levels. ®iffiter use appicaon does not permit testing using oil-based aerosols photometer cannot be used. Must use particle ‘counter with PSL spheres, ? Suitabity depends on U.S. Food & Drug Administration (FDA) or other regulatory acceptance. & Factor tes method shouldbe based on customer use needs and methods of tet fr fldcertication * Liquid aerosol ntoducton ino particle counters should be done with caution as optics contamination is an operational concern that may impact testing accuracy. 12 Conpigited © mart TISTIFUTE OF EAviRONSENTAL ScENCRS AND TEOHNOLOGY TEST RP-CCO2 os6.1 Challenge aerosol uniformity and steadiness To perform an accurate filter leak tet, itis necessary that the challenge aerosol concentration be uniform over the entire upstream face of the filter(s) being tested (spatial uniformity), It is also necessary that the aerosol concentration remain constant throughout the entire duration of the test (temporal uniformity) An upstream sample port must be provided in the system that will allow for the determination of an acceptable challenge concentration. This is addressed in Appendix C. ‘This testis a prerequisite to using the leak detection techniques described in section 6.2 6.1.1 Qualification test for factory test systems, field bench test systems, ducted terminal filters with aerosol injection ports, and unidireetional- flow, clean-air devices This testis a qualification test that applies to leak-test systems used in the filter production factory, to field bench test systems, and to unidirectional-flow, clean- air devices. It is performed once to prove a design, and should be redone only if there are changes tothe test system or changes to the design of the unidirec- tional-flow, clean-air device. The test does not apply to cleanrooms ‘The test system or clean-air device design and con- struction should provide a method for introducing the challenge aerosol in a manner that gives a uniform challenge to the filter, to the seal between the filter and the clean-air device, and to the supporting struc- ture. The uniformity ofthe challenge upstream of the HEPA or ULPA filter should be verified. The manu facturer of devices should also provide a method to obtain a single-point, representative sample of the aerosol upstream of the filter. Hereafter, the aerosol measurement at that single representative sample Jocation is called the representative upstream aerosol measurement or upstream sample port. The external connection to the upstream sample port should be located so it will be accessible inthe final installation of the device If the test system or clean-air device manufacturer provides documentation of the uniformity of the aerosol challenge per the test methods herein, then: '* The user need not repeat those tests as long, as the device remains unaltered and the aerosol is injected per the test system or clean-air device manufacturer's instruction. * The manufacturer does not need to provide multiple aceess ports to sample upstream of the filter per the sampling grid described in this procedure Each model of factory or field test system should be tested for uniformity. The clean-airsievice manufae- turer can qualify a line of geometrically similar de- vices by testing the smallest and largest devices inthe line. Each unit produced does not need to be tested If the test system or clean-air device manufacturer does not provide documentation of the uniformity of the challenge aerosol, then: * The manufacturer should provide multiple access ports to enable the user to sample up. stream of the filter per the sampling grid de- scribed in this procedure, © Verification of the uniformity of the chal- lenge is the responsibility of the user, Ifa representative upstream sample is not available, see Appendix B. To achieve a homogeneous challenge, the aerosol is typically introduced upstream of the fan. Due to the high concentration of the aerosol and the extreme turbulence caused by the fan, challenge droplets col- lide with each other and with the fan blades, forming larger droplets. Localized accumulation of droplets around the inner periphery of the fan housing results from the radial momentum imparted by the fan blade and the influence of established airflow patterns. The ‘extent of accumulation will depend on the concentra- tion of the challenge aerosol and the duration of test- ing. The accumulation of material implies that the fan is separating and not mixing atleast some of the par- ticles. Introducing the aerosol upstream of the fan does not ensure a uniform challenge. Alternative techniques that can be demonstrated to provide ac- ceptable uniformity of challenge concentration may be used. Regardless of the method used to obtain a uniform aerosol challenge, acceptable uniformity of challenge concentration should be demonstrated. 8) Apparatus for the aerosol injection and challenge ‘uniformity test 1) Aerosol source. 2) Aerosol photometer or particle counter. 3) Ifa particle counter is used, a suitable, cali- brated dilution device for use in measuring ‘the upstream challenge concentration. Test RP-coma [NSTTUT OF EXVIROWHENTAL SCIENCES AND THARNGLOGY Copyrighted Omateral 13,4) Sampling probe and access ports to sample upstream of the filter. The sampling probe can be cither square or rectangular, with inlet air velocity approximately equal to the velocity of the air entering the filter. One possible type of probe is a piece of tubing with inside diameter (ID) equal to the ID of the sample line to the particle detector with one end sealed. A hole on one side of the tube is used to obtain the sample. The edge of the hole should be within 6 mm (0.25 in.) of the end of the tube to allow sampling along the periphery of the filter. The hole should be oriented to face into the airflow approaching the filter. To demonstrate the uniformity of the aerosol chal- lenge, the device manufacturer may use the apparatus for either the aerosol photometer filter scan test method (section 6.2.1) or the discrete-particle counter filter scan test method (section 6.2.2). Prior to performing the challenge aerosol uniformity and steadiness test, the system should be checked in accordance with tests described in IEST-RP-CC002, TEST-RP-CC006, and IEST-RP-CC007 to verify that airflow velocity, volume, and balance (as specified) and, if appropriate, the uniformity of these parame- tets are within specified limits b) Procedure ‘Sampling locations are in each comer of the filter and six samples uniformly spaced over the face of each filter. The comer sample location is within 12 mm (05 in.) of the comer ofthe filter. The location of the six uniformly spaced sampling locations is determined by dividing the face area thatthe air flows through into squares of rectangles with the minimum possible as- pect ratio (Le. a8 close to square as possible). The sampling locations are the center of each square or rectangle. All samples are to be taken within 12 mm (05 in.) of the upstream face ofthe filter. In addition to sampling at the locations described in the previous paragraph, a sample should be taken at the port provided for the representative upstream sample. 1) Turn on airflow through device. 2) Turn on aerosol generator. 3) Measure concentration or counts at the rep- resentative upstream sample port. 4) Adjust aerosol source output as necessary. 5) Allow aerosol concentration or count to sta- bilize to a range that is within +10% of the ‘mean or target value, Record the aerosol 14 Copyrighted © material [NSTTUTE oF ENVRONMGNTAL SCIENCES AND THIOOLOOY concentration or count at the representative "upstream sample port. 6) Measure and record the aerosol concentra- tion or count at each of the other sample fo- cations upstream of the filter. 7) Measure and record the aerosol concentra- tion ot count at the representative upstream sample port to ensure that the aerosol source has remained stable. The concentration or count should be within +10% of the value obtained at the beginning of the test. ©) Reporting Calculate (an example is given in Appendix C): 1) Average of representative samples from beginning and end of test. 2) Average concentration of the other sample locations. 3) Standard deviation of the other sample locations. 4) Relative standard deviation (standard devia- tion/average obtained in step 2) 5) The relative deviation of each data point as abs [average value-point value)/average value]. 6) Ratio of average concentration at representa- tive upstream sample port to average con- centration atthe other sample locations. 4) Acceptance Suggested acceptance criteria (an example is given in ‘Appendix C) are: 1). Relative standard deviation less than 20% 2) Maximum relative deviation of any single point (50%). 3) Ratio of concentration at the representative Upstream sample port to average concentra- tion between 0.75 and 1.25. Based on these acceptance criteria and a 95% confi- dence level, the maximum error due to nonuniform upstream aerosol challenge and the bias in the repre- sentative upstream sample is a 0.038% leak passing as a 0.01% leak. If the bias in the representative sam- ple is corrected, there is a 95% confidence level that the maximum error due to nonuniformity of the dis- tribution of the challenge aerosol would be a 0.02% Jeak passing as a 0.01% leak, TEST-RP-COO26.1.2 Qualification test for cleanroom installation leak tests ‘The test aerosol should be introduced into the airflow upstream of the filter or filters in a manner that will produce a uniform challenge concentration over each of the filters, seals, and supporting structure, Where the-cleanroom design and construction permit, means should be provided to test one filter at a time, This procedure will minimize the exposire of the filters to the challenge aerosol When the cleanroom is constructed with multiple units, where each unit is a unidirectional-flow, clean- air device as described in LEST-RP-CC002, then the challenge uniformity of each unit should be tested according to section 6.1.1. Examples are cleanrooms built from multiple fan filter units oF multiple termi- nal units with replaceable filters. Terminal units with nnon-replaceable filters that have means for injecting aerosol into each individual unit should be tested according to section 6.1.1 For cleanrooms constructed from multiple fan filter units or multiple terminal units that do not have the means for injecting aerosol, it is useful to demon- strate the uniformity of the challenge on a typical unit according to section 6.1.1. The test would be con- ducted to determine a location to inject the challenge ‘aerosol that provides a uniform challenge. For exam- pple, determine the minimum distance between the injection into the supply duct and the terminal unit that provides a uniform challenge. Thereafter, when the aerosol is injected into the individual supply duct at the prescribed minimum distance upstream of the unit, it would be sufficient to measure the upstream aerosol at a representative sample port ‘Where construction dictates that multiple filters must be exposed to the challenge aerosol simultaneously, the aerosol should be introduced at a location that will provide a uniform concentration to all of the filters. ‘The uniformity of the challenge should be verified to the extent possible and practical 2) Apparatus for the challenge aerosol uniformity and steadiness test 1) Aerosol souree to be used during leak scan test 2) Aerosol photometer or particle counter to be used during leak scan tet. 3) If @ patticle counter is used, suitable cali- brated dilution device for use in measuring the upstream challenge concentration, TEST-RP-COOa TNSTTTUTE OF ENVROIENTAL SCIENCES AND THOBOLOGY 4) Access to sample upstream of the filters. The fiequency and number of locations for taking upstream counts are a matter for agreement between the customer and the supplier. To demonstrate the uniformity of the aerosol chal- lenge, the tester should use the apparatus to be used to scan the filters. Prior to performing the challenge aerosol uniformity and steadiness test, the system should be checked in accordance with tests described in IEST-RP-CC002, TEST-RP-CC006, and IEST-RP-CCO07 to verify that airflow velocity, volume, and balance (as specified) and, if appropriate, the uniformity of these parame- ters is within specified limits b) Procedure 1) Tun on airflow in the cleanroom, 2) Tum on aerosol generator. 3) Measure concentration or counts at the agreed-upon locations. 4) Adjust aerosol source output as necessary to obtain concentration or counts that are t0 be used for scanning the filters. 5) Allow aerosol concentration or count to sta- bilize to a range that is at least equal to the specified challenge concentration but no more than twice the specified challenge con- centration. At each location, take four con- secutive measurements at I-min intervals. Record the aerosol concentrations or counts for each measurement at each of the agreed- ‘upon locations. 6) When testing large cleanrooms, it may be appropriate to repeat the measurements pe- riodically during the scanning. 7) When testing large cleanrooms, the aerosol generator may be moved to several locations to challenge various portions of the filters. ‘These challenge uniformity tests should be repeated after each such move ©) Reporting Caleulate and report 1) The average concentration or count of the four consecutive measurements at each location. 2) Average of the location averages. Copyrighted Omawnal 154d) Acceptance Suggested acceptance criteria: 1) At each location, the four consecutive readings should be within a range defined by the loca- tion average +20% ofthe location average. 2) All of the location averages should fall be- tween the specified challenge and two times, the specified challenge. 6.2 HEPA and ULPA filter leak test These tests are performed to confirm that the HEPA or ULPA filters are free of damage and small leaks. ‘When these tests are performed on filters after instal- lation, they also confirm that the filter system is properly installed by verifying the absence of bypass leakage in the installation. These tests are particularly important for filters used in clean-air devices and cleanrooms intended to create clean areas classified at ISO Class 5 (according to ISO 1464-1), The tests are performed by introducing a challenge aerosol upstream of the filters and scanning immedi- ately downstream of the filters and support frame or sampling in » downstream duct. These procedures detect small holes and other defects in the filter me- dium and frame seal, bypass leaks in the filter frame and gasket seal, and leaks inthe filter mounting, ‘Two scanning techniques are presented along with recommendations for two aerosol challenge methods and two detection instruments. A total leakage method is provided for determining if there are leaks in duct-mounted or air-handler-mounted HEPA. and ULPA filters, where the filters are inaccessible. Scanning HEPA and ULPA filters may not be appro- priate in cases where access to the downstream face ‘of the filter i difficult or impossible. The presence of non-removable grilles or “diffusion membranes” downstream of the filter(s) will make location and repair of leaks difficult or impossible. In these cases, the decision to scan the filters and the choice of the scanning methods and criteria should be carefully considered by the customer, supplier, and tester. The smallest leak to be detected during a leak testis called the designated-leak size. The designated-leak size establishes an upper limit for the size of leaks that are not detected during scanning and remain in the filter system after the scanning and repairs are completed. The choice of the size of the designated leak depends on the efficiency of the filters, on the type of test (factory, field bench, or post installation) and on the intended use of the clean-air device or cleanroom. When HEPA filters are scanned with the Photometer test method and aerosol generated by a Laskin nozzle, itis customary to define the desig- nated leak to have 2 standard leak penetration of 0.0001 (0.01%). For ULPA filters, itis appropriate to use smaller values to define the designated leak. See Table 2 for other values. Using smaller designated leak sizes will cause the scan test to take longer or will require higher challenge aerosol concentrations The customer needs to balance the added time, cast, atid challenge aerosol deposited on the filters with the benefit of finding very small leaks ‘The description of the two scanning methods includes discussion of factors such as photometer response time and counting statistics. Consideration of such factors and the calculations described in the proce- dures are important for detecting small leaks (i.c., leaks near the designated-leak size). If the procedures described in sections 6.2.1 and 6.2.2 are not followed, the usual result is that leaks near the designated-leak size are missed while larger leaks are still detected, Prior to performing the leak test, the system should bbe checked in accordance with tests described in TEST-RP-CC002, IEST-RP-CC006, and IEST-RP- €C007 to verify’ that airflow velocity, volume, and balance (as specified) and, if appropriate, the uni- formity of these parameters are within specified lim- its. The challenge aerosol uniformity and steadiness test should be completed prior to doing the leak test. 6.2.1 Aerosol photometer filter scan test method This section establishes the basis for choosing the conditions for conducting the filter leak test with an aerosol photometer. The choice of specific test condi- tions is a matter for agreement between the customer and the supplier. This agreement should specify the following: ‘© The aerosol photometer to be used and its response time constant. © The filter exit airflow velocity at which the leak testis to be conducted. Unless specified otherwise, the exit airflow velocity is 0,45(£0.05) msec (90[10] fVimin). © The challenge aerosol material and genera- tion method. Unless specified otherwise, the ‘aerosol generator is one or more Laskin noz~ les operated at 138 Pa (20 Ibfin.) * For cleanrooms, the frequency and loca- tions at which the challenge concentration is verified. © The standard Jeak penetration of the desig- nated leak. Unless specified otherwise, the 16 Copyrighted © materi [STITUTE OF ENVINOSMENTAL SCIENCES AND TECHNOLOGY TeSTRP-COMGdesignated leak is 0.0001 (0.01%) when scanning HEPA filters. For ULPA filters, it is appropriate to use smaller values to define the designated leak. See Table 2 for other values. © The maximum scan rate. Unless specified otherwise, the maximum scan rate when us- ing a square probe should nat exceed 3 m/min (5 cmisec) (10 f/min, 2 in/sec) With a rectangular probe, the maximum area scan rate should not exceed 0,093 m*/min (1.0 ft/min) unless otherwise specified This section provides a method for determining the sampling probe inlet size. It also relates the linear scan rate, the photometer response time constant, the sampling probe inlet size, the standard leak penetra- tion of the designated teak, and the threshold that indicates a possible leak during scanning. Examples ning HEPA Filters with Photometers” (see Appendix J, Bibliography) This test procedure has been an industry standard ‘method for defining defects in filters or filter systems, NOTE: As presented in this document, the pro- cedure has been modified to account for the re- sponse time of the photometer and to account for rectangular probes. ‘The test provides both qualitative and quantitative results in identifying leaks, The test can easily be re: produced. The generation of a reasonably consistent ‘concentration of liquid aerosol to challenge most sys- tems is not difficult. A concentration of approximately 10 pg/L to 90 pg/L of air is an adequate challenge, NOTE: When the test system or clean-air device is located in a nou-clean environment or in @ par- tial-coverage cleanroom, aerosol from the sur- are provided in Appendix D. The basis of the rela- rounding space may interfere with the leak tests, tions presented in this section can be found in “Scan- Table 2—Designated leak sizes for various efficiencies of filters. Sade Metin | Sanda ea Peaaict Dasgneseatorasoyar | “Deal loa triste 7 Son To pelea Ts tig va Sree | Momma atae bs eran “arn, (Poe tae Type C(HEPA) | 99.99% (0.01%) 0.3 um (2) 0.01% 0.01% 1 TypeD (HEPA) | 9.999% (0.001%) 0.3 um (2) 0.005% 0.01% 10 TypeF(ULPA) | 99.96% (001%) | O1o02um | 0.005%: 5 001% 10 [Type F (uuPay | 99.9905% (c0005%) | 0.1t002um | O00 : 5 (0.005% 10 Ter Ura) —| war omor | Ofnoaym | oon | “oman Type G (Super ULPA) |>99.9990% (<0.001%) MPPs? 0.001% 10 0.002% 20. Tmesceray [oosmoony —OSP02K" | coin 1 onm | 1 | | Teokteear [aman [S29 omen | ve | oom | 2 NOTES: * Overall orintegral minimum efciency (maximum penetration) 2 Thermally generated aerosol with nominal 0.3 jm mass median clameter. I has been found n practice that he number mean size forthe thermaly generated aerosol is well under 0.2 um and lose tothe MPPS of typical HEPA fers 3 Mia are tested at MPPS proro fier assembly. These filers are not tested for overall efciency “Recommended iter type ifn stu or field bench test wl uiize a Laskin nozzle generator aerosol and photometer. if thermal generators ized, the % penetraton values should be se at 0.05% and the mpi of overal or infogral minimum efciency value woul be 5, The par- lice size range selected s that which yield the minimum eficiency (contains the it’s MPPS). § Recommended fit ype it inst orld bench test wil uilize a thermal generated aerosol and photometer. This file provides some margin {or measurement ferences (factory tin stu) along with the uiizalon of afield aerosol hal appreaches the MPPS, The parce size range selected is that wich yields the minimum efficiency (contains te it’s MPS) TEST-RP-CCOBA2 [NSTITUTE OF ENVIRONHENTAL SCRNCES AND TECHNOLOGY Conprighed Omaeral 17a test b) test Apparatus for the aerosol photometer filter scan 1) Laskin nozzles or other aerosol source com- patible with an aerosol photometer. 2) Aerosol photometer with logarithmic or tin- ‘ear read-out (section 3), with a sample flow rate of 28.3 Limin (1 ft/min) and a known response time constant 3) A sampling probe with an inlet opening of square oF rectangular configuration, provid- ing an inlet airflow velocity within +10% of the average exit airflow velocity of the fer(s) being scanned when operating. at the sample flow rate of the photometer (see Fig- ure 1). The dimension of the opening of the probe inlet parallel to the direction of scan- ning (D,) should be no less than 6 mm (nominal 0.25 in.). The dimension of the ‘opening of the probe inlet perpendicular to the direction of scanning (1V,) is determined from the sample flow rate. The transition section of the probe, the portion between the inlet and the connection to the instrument tubing, should have an overall length, 7, at least equal to the maximum dimension of the probe inlet Procedure for the aerosol photometer filter scan This test is performed by introducing the specified challenge aerosol upstream of the filters and search- ing for leaks by scanning the downstream side of the filters with the photometer probe. ‘The sampling probe inlet size should be calculated from the photometer sample flow rate and the filter exit airflow velocity so that the probe inlet air veloc- ity approximates the average filter exit airflow veloc- ity within 10%, where: D, )» = probe dimension parallel to the sean direction incm(or in) F, = flow rate of the photometer V = average exit airflow velocity of the filter W, = probe dimension perpendicular to the scan direction in em (or in.) NOTE: Consistent units should be used. Figure [Example of a sampling probe. Linear scan rate is related to area scan rate by: 4 where S, = linear sean rate in em/sec (or in/see) [/| A, = area sean rate ‘The aerosol photometer filter scan test should be per- formed as follows: 1) The design airflow velocity should be estab- lished and uniformity confirmed prior to performing this test 2) The aerosol should be introduced into the air supplied to the filter or filters under test in a ‘manner that will produce a uniform chal- lenge concentration over the surface of each filter. Minimize exposure ofthe filters to the challenge aerosol. 3) The aerosol challenge should be verified as uniform as required in section 6.1 4) The concentration of the challenge aerosol should be measured at the representative up- stream sample location, using a photometer ‘whose sensitivity is adjusted to a baseline of 100 e/L in accordance with the manufac- turer's instructions or calibration curve. A reading of 10% to 90% (which corresponds to 10 pg/L to 90 ug/L of ait) on the photometer should be obtained for correct challenge con- 18 Cortes Smal [ISITUTE oF ENVIRONMENTAL SCENCES 9 TECHNOLOGY TeSTRP-COmMa acentration. The upstream concentration should be stable over time such that four consecutive readings at I-min intervals are within 15% of the average of the readings If a representative upstream sample is not available, see Appendix B. If the reading is high, it may be desirable to reduce the concentration of the challenge aerosol in order to limit exposure of the fil ter under test to the challenge material. Con- versely, if the reading is low, the amount of aerosol will need to be increased in order to ‘maintain sensitivity to small leaks. ‘After the correct reading has been obtained, adjust the photometer sensitivity, gain, or span for a reading of 100%, or full scale, while sampling the upstream aerosol 5) The entire face of each filter should be scanned for leaks, using slightly overlapping, strokes of the probe and moving the probe at arate not exceeding the maximum scanning, speed. Also scan the perimeter of each filter (o locate leaks in the bond between the filter pack and the frame and to locate leaks in the seal between the frame and the support structure. The probe should be held ap- proximately 25 mm (1 in.) from the filter ‘media during scanning. NOTE: Non-pane! filter configurations such as V Banks limit access to the fiter media. In these cases the total leakage test method (section 6.2.3) is recommended Possible leaks while scanning are indicated by a pho- tometer response equal to or greater than the penetra- tion response calculated by: where: Pq = maximum penetration indicated as the probe is scanned over a leak the size of the designated Teak 1L,= standard leak penetration of a designated leak, expressed as a fraction of the upstream concen- tration (e.g, 0.01% = 0.0001) robe dimension parallel to the scan direction in ‘em (or in.) ‘5, = maximum linear scan rate in emisec (or in/see) TT, = response time constant of the photometer as if being used for scanning (ie., with the same range setting, the same tubing, ete.) If the value calculated for Py, is less than three times the minimum sensitivity of the photometer, or if Py is less than 0.1 times Z,, then a slower maximum scan speed S, must be chosen and a new P, calculated, If an operator watches a meter during scanning to detect leaks, then any meter indication equal to or greater than Py, is considered an indication of a leak. If alarm settings are used to signal the operator o automatic scanning system, then the alarm level should be set to Py. Any indication of a leak equal to oo greater than P, should be cause for sustained resi- dence time of the probe at the leak location. The size and location of the leak are identified by the position of the probe that maintains the maximum sustained reading on the photometer. If these procedures regarding P, are not followed, the result will generally be that leaks near the desig- nated-leak size are missed while larger leaks are still detected. NOTE: Prolonged exposure of filters to the challenge aerosol should be avoided. ©) Reporting for the aerosol photometer filter scan test For HEPA filters, unless specified otherwise, the des- ignated-leak size is 0.01%. That is, report all leaks for which the stationary reading exceeds the following: 1) For a linear read-out photometer: a reading ‘greater than 0.01% of the upstream chal- fenge aerosol concentration or as otherwise agreed 2) For a logarithmic read-out photometer: a reading greater than 0.01% on instruments with a direct-reading scale or equivalent reading on the instrument calibration curve (one minor scale division), or as otherwise agreed. For ULPA filters, it is appropriate to use smaller val- ues to define the designated leak. See Table 2 for other values. )__ Repairs forthe filter scan test ‘The repairs should be limited as follows: 1) A material acceptable to the owner of the application should be used. 2) Factory repair should not block or restrict tore than 196 ofthe filter face area, and no single repair should exceed 13 om’ (2 in.) TEST. RP-CCI2 INSTITUTE OF ENVIROMENTAL SCENGES AND TECHNOLOGY Coprighed Omormiat 193) Field repair should not block or restrict more ‘than an additional 3% of the filter face area, and no single repair should have a lesser dimension exceeding 3.8 cm (1.5 in), After the repair is complete and suitable cure time has been allowed for the patch to set, check for leaks in the vicinity of the repair. Repairs to filter leaks may be made by procedures acceptable to the cus- tomer and the supplier. ©) Acceptance for the aerosol photometer filter scan test Filter integrity requirements are a matter for agree- ‘ment between the customer and the supplier. 6.2.2 Diserete-particle counter filter sean test method This section gives one possible set of conditions for conducting the filter leak test with a discrete-particle counter and alternative source particle challenge Appendix E provides the methods for calculating other conditions for conducting the filter leak test with a discrete-particle counter and alternative source aerosol challenge. Appendix D provides two exam- ples of other conditions. Appendix F addresses dilu- n of aerosol, which is typically needed when igher particle concentrations are required upstream, The choive of specific test conditions is a matter for agreement between the customer and the supplier. This agreement should specify: 20 ‘The discrete-particle counter to be used, its flow rate, the particle size range of interest, and the particle counter sample time to be used, ‘The filter exit airflow velocity at which the leak testis to be conducted. ‘The challenge aerosol material, size distribu- tion, and concentration, For cleanrooms, the frequency and locations at which the upstream challenge concentra- tion is verified The standard leak penetration of the desig- nated leak. To use the equations in Appen- dix E, the designated-leak penetration is expressed in terms of a fractional standard leak penetration (not a percent. The number of particle counts that char- acterize the designated leak at the specified test conditions. Copyrighed © moter [NSITTUTE OF ERVROSOHENTAL SCIENCES 8 THABOLOGY ©The probe dimension in the direction of the scan. (While not an essential part of the cus- tomer-supplier agreement, this dimension needs to be chosen in order to calculate the probe dimension perpendicular to the scan direction.) NOTE: When the test system or clean-air device is located in a non-clean environment or in par- tial-coverage cleanrooms, aerosol from the sur- rounding space may interfere with the leak tests, It is recommended that the discrete-particle counter test method be limited to applications where the background particle count in the vicin- ity of the test is no more than 10% of the count that characterizes a designated leak. NOTE: The penetration through the unbroken filter medium is considered to be the background count. ) Apparatus for the discrete-particle-counter filter scan test ‘Apparatus for this test should consist of 1) A suitable aerosol source to provide appro- priate challenge aerosol concentration. 2) A discrete-particle counter having a known sample flow rate and the capability to detect, the particle sizes designated in the challenge aerosol. The particle counter should have a very low count rate when sampling particle- free air. This is particularly important when ‘one count is used to characterize the desig- nated leak. To use the method described in section 6.2.2b, the discrete-particle counter should have an audible alarm that beeps for each particle that is counted and have a counting mode that is continuous or at least long compared with the time required to scan a filter. To use the method described in Appendix E, the particle counter should have essentially zero reset time between known sampling time increments. This ‘method can be used with discrete-particle counters with an adjustable alarm that can be set to the particle count used to character- ize the designated leak. The method in Ap- pendix B can also be used with automated scanning, systems that record the counts in ‘each sampling period as the probe is moved across the face of the filter. 3) A sampling probe with an inlet opening of square of rectangular configuration, provi ing an inlet airflow velocity within 10% of the average exit airflow velocity of the fil- TESTR-COMMter(3) being scanned when operating at the sample flow rate of the discretepartcle counter (see Figure 1). The dimension of the ‘opening of the probe inlet parallel to the di- rection of scanning (D,) should be no less than 6 mun (nominal 0.25 in). The dimen- sion of the opening of the probe inlet per- pendicular to the direction of scanning (W,) is determined from the sample flow rate The transition section of the probe, the por- tion between the inlet and the connection to the instrument tubing, should have an over- all length, 7, atleast equal to the maximum dimension ofthe probe inlet W, 4) A suitable calibrated dilution device to meas- lure the upstream challenge concentration, b) Example set of conditions for conducting the filter leak test Specified conditions: + The particle counter flow rate is 28.3 Limin (1 f/m), © The particle counter sample acquisition time is continuous, ‘© The filter exit airflow velocity is 0.45(40.05) misec (90{+10] f¥imin ). * The challenge aerosol concentration is 4.2 x 10 particles/m’ (1.2 x 10° particles/ft). This is the concentration of the aerosol in the size range of interest that the particle counter can detect. © The standard leak penetration is 0.0001 (0.01%), NOTE: This isthe historical leak definition used for HEPA filters and may not be appropriate for scanning ULPA filters. See Table 2 for other values. © The number of particle counts that characterize a designated leak is one. The particle counter is set up to give an audible beep for each particle counted, and each beep is interpreted as a possible leak. NOTE: Using a single count to characterize the designated leak is applicable only to ULPA fil- ters. The scanning conditions described here will be difficult to apply to HEPA filters and to situations where the challenge concentration is higher. In those cases, substantial counts due to the normal penetration through the media will ‘cause false alarms. For handling such cases, re- fer to Appendix E. ‘+ The dimension ofthe sampling probe in the scan direction is 12.5 mm (0.5 in. Calculated conditions (see Appendix E for the meth- ‘ods of calculating these values based on the specified test conditions) ‘* The calculated sampling probe dimension perpendicular to the scan direction is 84 mm 2 in), + The linear scan rate of the probe should be 1.5 m/min (5.0 fin) or less when using the given conditions and moving the probe paral lel to the 12.5-mm (0.5-in) direction, Allow: ing for some overlap of adjacent scans, this is equivalent fo an area scan rate of approxi- ‘mately 0.12 m/min (1.2 ft/min). ‘The sampling time criteria are not applica ble, However, the probe is over a point on the filter for 0.5 se. NOTE: When there is a sample time (7), it is necessary t0 check the sample time criteria to en- sure that the counts from a point leak are not split between consecutive sample times. If the probe is moving so slowly that the counts from a point leak are counted in more than one sample time, then it will be possible to miss leaks that are approximately the same size as the threshold leak. (Large leaks will still be obvious.) When the method described here is used with manual scanning, itis likely that the operators ‘will mentally sum the counts that they hear from a region of the filter. Hence, there is no mini ‘mum scan speed because, in essence, the opera- tors will automatically inerease the sample time as they slow down the scan. + Under these calculated conditions, the prob- ability of missing a leak that has a standard penetration of 0.00015 (1.5 times the desig- nated leak) is 22%. The probability of miss- ing a leak that has a standard penetration of 0.0002 (two times the designated leak) is 14%, The probability of a false indication from a leak that has a standard penetration of 0.00005 (one-half penetration of the des- ‘ignated leak) is 40%. When scanning in this manner, any particle count is inerpreted as a possible leak. Every count requires that the operator return to the site of the count and search for the location that yields the maximum sus- tainable count rate. Large leaks will be obvious, but for leaks that are near the designated leak size, a ‘6see count should be taken. Counts of 12 or more TEST-RP-CO042 INSTITUTE OF ENviROWMENTAL SCLENCES AND TREANOLOGY Copied Dmatrial 2dindicate a leak that is equal to or larger than the des- ignated-leak size when using the conditions specified in this section. NOTE: In situations where leaks close to the designated-leak size are to be detected, it may be desirable to verify the upstream challenge con- centration for leaks of questionable validity ‘The probability of missing leaks is established by the criteria during scanning, not by the stationary reading, after the possible leak has been identified, Returning, to the site of a count and counting for a longer period of time helps identify the location of the leak and identify counts that might have been the result of normal penetration through undamaged filter media for counts that might have been the result of a leak that is smaller than the designated leak. The probabil- ity of a false indication from a leak that is 0.00005 (one-half the penetration of the designated leak) is 2.0% during the 6-sec count. ‘The upstream challenge concentration will not be exactly the specified value. This RP allows the con- centration to be no less than the specified concentra- tion and up to two times the specified concentration, If the challenge concentration is twice the specified value, then the scanning will be sensitive to leaks that are one-half the threshold size. ©) Procedure for the discrete-particle-counter filter scan test This test is performed by introducing the specified challenge aerosol upstream of the filters and search- ing for leaks by scanning the downstream side of the filters with a particle counter probe. 1) The design airflow velocity should be estab- fished and uniformity confirmed accor to section 6.1 prior to performing this test. 2) Introduce the aerosol into the air supplied to the filter or filters under test in a manner that will produce a uniform challenge concentra- tion over each fiter’s surface. Minimize ex- posure of the filters to the challenge acrosal 3) Verify that the specified upstream aerosol challenge is present by measurement of the upstream particle concentration at the sample port provided for measuring the representa- tive upstream counts. The upstream concen- tration should be stable over time such that each of four consecutive readings is within 15% of the average of the readings. The up- stream concentration should be uniform over 4) Reporting for the discrete-particle-counter filter sean test the area immediately upstream of the filter area under test according to section 6.1 4) Ifa representative upstream sample is not available, see Appendix B. 5) A suitable dilution system is required to avoid exceeding the concentration capability of the particle counter. NOTE: It is not necessarily obvious when the concentration exceeds the capability of a particle counter. The counter may register numbers well lunder its limit when in fact it is only counting a fraction of the particles. 6) All counts taken from the representative up- stream sample port should be at least equal (o the specified challenge concentration but no more than (wice the specified challenge con- centration. Lower concentrations reduce the sensitivity to small leaks, while higher con- centrations increase the sensitivity to small Jeaks. The frequency and number of locations for taking upstream counts are a matter for agreement between the customer and the supplier. 7) Scan the entire face of each filter for leaks, using slightly overlapping strokes of the probe and moving the probe at a rate not ex- ceeding the linear scan rate determined ac- cording to section 6.2.2b. The probe should be held approximately 25 mm (1 in.) from the filter media during scanning. A separate pass should be made over the perimeter of each filter, at the determined linear scan rate, in order to locate leaks in the bond be- tween the filter pack and the frame and in the seal between the filter frame and the clean-air device or grid structure. NOTE: Non-panel filter configurations such as V Banks limit access to the filter media. In these cases the total leakage test method (section 6.2.3) is recommended. Any detected particle count equal to or greater than the number of particle counts that characterize the designated leak should be cause for sustained residence time of the probe at the leak location. The location is identified by the position of the probe that sustains the maximum reading on the parti- cle counter. 22 Comrighed Smmeriat WSTITUTE OF BVRONSENTAL SCIENCES An TEORNOLOGY TEST-R-CCuBA2Record all leaks that exceed the designated-leak size established by agreement between the customer and the supplier. ) Repairs for the filter scan test ‘The repairs should be limited as follows: 1) A material acceptable to the owner of the application should be used 2) Factory repair should not block or restrict ‘more than 1% of the filter face area, and no single repair should exceed 13 cm’ (2 in. 3) Field repair should not block or restrict more than an additional 3% of the filter face area, and no single repair should have a lesser di- mension exceeding 3.8 em (1.5 in). After the repair is complete and suitable cure time has been allowed for the patch to set, check for leaks in the vicinity of the repair. Repairs to filter leaks may be made by procedures acceptable to the cus- tomer and the supplier. 4) Acceptance forthe disrete-particle-counter filter scan test Filter integrity requirements are a matter for agree- ment between the customer and the supplier 6.2.3 Aerosol photometer total leakage test method This procedure may be used for installations where HEPA ot ULPA filters are inaccessible or the con- figuration of the installation does not permit stistac- tory scan testing of the filter. This method is less sensitive than the methods described in sections 6.2.1 and 6.2.2. 4) Apparatus for the aerosol photometer total leak- age test 1) Laskin nozzle aerosol source as described in seetion 3, of other aerosol source compatible with an aerosol photometer. 2) Aerosol photometer with linear read-out as described in section 3, with a nominal pling flow rate of 28.3 L/min (1.0 ft/min) ) Procedure for the aerosol photometer total leak- age test This test is performed by introducing the challenge aerosol upstream of the HEPA or ULPA filters in- stalled in ducts remote to the cleanroom. The filtered air from the duct is then measured and compared with the upstream concentration to determine the total leakage of the filter installation. The design airflow velocity should be established and uniformity con- firmed according to section 6.1 prior to performing this test. NOTE: Sampling should be performed at @ suf- ficient distance downstream from the filters to provide for mixing of the aerosol. Sampling is typically performed in the ductwork but, if ths is not practical, the air may be sampled at the sup- ply air diffusers as it enters the cleanroom. The aerosol photometer total leakage test should be performed as follows 1) Test the filters and the installation at an av- erage airflow rate that is within +20% of the ‘operating design airflow rate 2) Introduce the aerosol upstream of the filters in a manner that will produce a uniform concentration throughout the airstream. The upstream concentration of challenge aerosol should be adjusted t0 10 pg/L to 20 ug/L of air. The photometer should then be adjusted to read 100% 3) Measure the aerosol concentration in the duct downstream from the filters in accor- dance with section 10 of ASME NS10. 4) Measure the concentration in several loca- tions around the perimeter of the duct within 2.5.em (1 in.) of the duct wall NOTE: If any sample taken according to section 16.2.3b(3) varies more than 20% from the arith- metic mean of all of the samples taken, or ifthe wall area samples taken according to section 6.2.3b(4) indicate high concentrations streaming along the walls, this method cannot be used ‘unless mixing baffles can be installed to produce uniform mixture, or a sample location can be chosen that exhibits uniformity ©) Reporting for the aerosol photometer total leak- age test 1) Calculate the arithmetic mean of the down- stream readings. 2) Calculate the standard deviation of the arithmetic mean, 3) Caleulate and report the average filter bank penetration as described in section 3 of ‘ASME NSI0. )_ Repairs for the aerosol photometer total leakage test HEPA and ULPA filters may be repaired provided that the size of the repair or repairs does not block or restrict more than 5% of the filter's face area (not [NSTHTUTE OF ENVIRONMENTAL SGRNCESANOTHOBOLOGY Coprighed Smatenal 23including the frame). After repair is complete and suitable cure time has been allowed for the patch to set, retest the system. Repairs to filter installation feaks may be made by procedures acceptable to the customer and the supplier. NOTE: Permissible repairs in this section for remote, duct-mounted filters differ from the permissible repairs described in sections 6.2.1 and 6.2.2. For filters where the downstream flow directly enters the workspace to create a clean 2A Copyrighted © material [ISTITUTE OF ENVRONSHENTAL SCENCES AND THGINOLOGY work zone, distortion ofthe flow downstream of the filter by the repair may be a concern. In re mote, duct-mounted filters, total blockage is the concern. Moreover, it may be very difficult to locate a leak due tothe inability to scan the filter. ©) Acceptance for the aerosol photometer total leakage test The total leakage of the filter installation is @ matter for agreement between the customer and the supplier. TEST RP-COW2APPENDIX A—CONSIDERATIONS IN LEAK TESTING It is appropriate to scan HEPA filters using the acro- sol photometer and Laskin nozzle sean test method. The diserete-particle counter sean test method may be uused to scan HEPA filters in some circumstances. ‘When a single count is used to characterize the des- ignated leak, the counter method cannot be used with HEPA filters. When using discrete-particle counters to scan HEPA filters, itis not appropriate to use the same criteria for a leak as when scanning HEPA fil- ters with photometers. This is particularly true ifthe aerosol is near the most-penetrating particle size. Al_ TESTING BASED ON FILTER EFFICIENCY It is appropriate to scan ULPA filters using the dis- crete-particle counter scan test method. The aerosol photometer scan test method may be used in some Circumstances for 99.99% efficient filters. It is not applicable to higher efficiencies. tis appropriate to scan super ULPA filters using the THANOLOGY TESTRP-COB?APPENDIX B—CALCULATION OF AEROSOL CHALLENGE ‘This appendix provides a method to determine the approximate concentration of the challenge aerosol This method should be used only if it is impractical (o obtain a representative upstream sample. This method may produce less-accurate results in deter- ‘mining the upstream concentration and the size of leaks. The aerosol concentration is calculated from the previously known aerosol generator output and the airflow though the clean-air device. Values are provided for typical Laskin nozzle output when using DOP. Other aerosol generators or materials may be used if similar output information is available Bi CALCULATION WHEN GIVEN THE OUTPUT INTO VOLUME OF AIRFLOW Given: Aerosol generator output: C1 = ug/L or NoJ/L per nozzle when discharged into a flow rate of Q, m/min eee Airflow through the clean-air device of Q; m’/min Catan Cancenvaon Cyn wet of NodL ia fw rt eee o=ne, 2 Q Ite plore ed en Adjust the photometer so that full scale represents Crpw/L using calibration data from the photometer manufacturer. Ifa particle counter is used, then: Use the concentration C; in the equation in Appendix E13 to calculate the linear scan rate. B2 CALCULATION WHEN GIVEN OUTPUT PER UNIT TIME Given: Aerosol generator output: E= ug/min or NoJmin per nozzle N= nozzles in use Airflow through the clean-air device of Q; m'/min Caleutate: Concentration C; in ug/L or NoJL in flow rate Q» using N nozzles NE ** (@)(0000) where 1000 converts m’ to L. Ifa photometer is used, thea: Adjust the photometer so that full scale represents Crnw/L using calibration data from the photometer ‘manufacturer, Ifa particle counter is used, then: Use the concentration C; in the equation in Appendix E13 to calculate the linear scan rat. B3 EXAMPLE USING LASKIN NOZZLES AND A PHOTOMETER ‘The aerosol in this example is produced by a Laskin- nozzle generator using two nozzles operated at 138 kPa (20 Ibvin.) pressure with DOP. Other liquids may give different results. A concentration of ap- proximately 10 ug/L, of air is generated in a flow of 38 m’/min (1350 fP/min). A compressed-air source capable of delivering 75 L/min of air (volume at stan- dard conditions) at 138 kPa (2.7 std ft/min at 20 bin.) is required for each Laskin nozzle. This, concentration is equivalent to approximately 3 x 10” droplets/L (10° droplets/f) of air. Conditions given: Aerosol generator output: Cy = 10 nQIL per nozzle when discharged into aflow rate of O; Q = 38m'%min N ‘two nozzles in use Airflow through the clean-air device of Q, = 56 m’min Calculation: . C,=(2)(10ng/1){ 28 Ama | ayer 56m Fmin Adjust the photometer so that full scale represents =14 g/L using calibration data from the photometer ‘manufacturer. TesrRP-comaa TnetiTer or EvasosonraL SNES ano TECHNOLOGY Conical Smonrat 29