High-Efficiency Particulate Air Filters

Content

• Introduction
• Notice Summary
• Applicability
• Related Events
• Significance of Events
• Event Causes
• Corrective Actions
• Hazard Reduction
• References

Introduction

This notice is one in a series of publications issued by the Office of Nuclear and Facility Safety to share nuclear safety information throughout the Department of Energy complex. For more information, contact Dick Trevillian, Office of Operating Experience Analysis and Feedback, Office of Nuclear and Facility Safety, U.S. Department of Energy, Washington, DC 20585, telephone (301) 903-3074. No specific action or responses are required solely as a result of this notice.

Safety Notices are distributed to U.S. Department of Energy Program Offices, Field Offices, and contractors who have responsibility for the operation and maintenance of nuclear and related facilities, and to other organizations involved in nuclear safety. Written requests to be added to or deleted from the distribution of Safety Notices should be sent to: BR Richard L. Trevillian, EH-33, Room E-460 GTN, U.S. Department of Energy, Washington, DC 20585.

The ESH Office of Information Management maintains a file of Safety Notices and supporting information. Copies can be obtained by contacting the Office of Information Management at (301) 903-0449 or by writing to the Office of Information Management, U.S. Department of Energy, EH-72/Suite 100, CXXI/3, Washington, DC 20585.

Notice Summary

This notice presents lessons learned related to specification, purchase, and application of Flanders and other High-Efficiency Particulate Air (HEPA) filters used in safety-related applications. It describes the purchase and use of Flanders HEPA filters that did not meet DOE qualification specifications at Los Alamos National Laboratory (LANL). The notice also describes the use of Flanders HEPA filters that had been down-rated by the manufacturer to flow rates below those being used at the Savannah River Site. The events described in this Safety Notice did not result in the release of contaminants or radioactive materials to the environment. However, the consequences of contaminants penetrating improperly rated or unqualified HEPA filters can be avoided by applying the lessons learned discussed in this notice.

Applicability

This Notice applies to all DOE facilities that specify, procure, test, and use HEPA filters. Personnel at facilities that use HEPA filters for protection against airborne radiological releases and for final filtration in hazardous systems should pay particular attention to this notice. No specific actions or responses are required solely as a result of this Notice.

Related Events

Savannah River Site

While reviewing replacement of HEPA filters in the 2-H evaporator cell ventilation system at the H-Area tank farm, facility personnel identified a discrepancy between the operating and rated flow capacity of the associated system filters. They first discovered and reported the condition on July 1, 1992.¹ The installed Flanders 50-16610 HEPA filters were rated at 1,500 scfm flow/1-inch water column (WC) pressure drop according to the 1989 Process Ventilation Design Guide.² However, the 1991 revision of the Guide rated the same filters at 1,000 scfm/1-inch WC. The filters had been de-rated by the manufacturer without the user facility or Waste Management Engineering being aware of the change. The 2-H evaporator cell ventilation system normally operates at 1,400 scfm through a dual-bank, parallel, HEPA filter arrangement. One filter serves as the primary filter and the second as a spare. The 1,000- scfm rated filter could not be used as either the primary or spare filter for a 1,400-scfm flow without increasing the risk of exceeding the single-filter differential-pressure limitation of 10 inches WC.

Investigators also determined that facility drawings showed the system design flow rate to be 800 scfm and not the actual fan-flow capacity of 1,400 scfm. Facility personnel believed the 1,400-scfm fan was installed later to assure sufficient negative pressure in the evaporator cell. Documentation of the fan change was not identified in the facility design drawings. Facility personnel shut down the 2- F evaporator unit as a result of these findings pending a complete evaluation.

Although airflow through the filters exceeded the flow rate at which they were tested, the differential pressure across the filters did not exceed the manufacturer recommendation, which was used as the performance parameter of the filters. The critical characteristics of HEPA filters, efficiency and differential pressure, were within MIL-F-51068F³ specifications, which are greater than 99.97 percent and less than 10 inches WC, respectively.

Los Alamos National Laboratory

In November 1992, LANL personnel reported⁴ that several HEPA filters procured from Flanders Filters did not meet DOE qualification requirements.^5,6,7,8 The filters involved were the poly- vinyl-chloride (PVC) coated aluminum-separator type with Flanders base part numbers starting with -007-V. LANL procured 180 Flanders filters of this type between July 1, 1991, and June 30, 1992. Procurement specifications required the filters to be on the Qualified Products List of MIL-F-51068^9,10 or pass equivalent tests specified in ASME N509-89, paragraph 5.1.⁷ During the early stages of procurement in July 1990, facility personnel were aware that the purchased Flanders filters were no longer on the Qualified Products List. They expected that qualification test data would be provided and continued procuring Flanders filters. Between August 1990 and August 1991, they attempted to obtain clear, up-to-date qualification test results from either Flanders or the Rocky Flats Filter Test Facility, but were unsuccessful. In March 1992, Flanders filters from LANL stock were sent to Rocky Flats for independent testing. Preliminary results from Rocky Flats indicated that some of the filters failed. In October 1992, Flanders personnel performed qualification tests on 16 Flanders HEPA filters from LANL stock. The tests were witnessed by two LANL representatives. Two of four filter models failed qualification testing. Following the disclosure of these undisputed test results, facility personnel submitted an Occurrence Report Notification in November 1992.⁴

The two filters failed the rough-handling and wet-overpressurization tests of military specification MIL-F-51068F³ but passed the aerosol-penetration and air-flow-resistance tests. The rough- handling and wet-overpressurization qualification requirements indicate filter penetration efficiency after exposure to abnormal operating conditions. The 24" x 24" x 11-1/2" filters were used primarily for final filtration in facility exhaust air-cleaning systems. The failures also affected qualification of filters of the aluminum separator design with LANL stock numbers GS-2703 (8"x8"x5- 7/8"), GS-2704 (12" x 12" x 5-7/8"), GS-2705 (18" x 18" x 11-3/4"), and GS-2716 (8" x 8" x 8-1/4").

Test and operational data indicated that no deficiencies were related to the capability of the Flanders filters to meet penetration (ability to remove aerosols) and air flow requirements⁸ under normal operating conditions. In-place filters which do not meet performance requirements are replaced.

Significance of Events

These events are significant because unqualified or de-rated HEPA filters could be procured and installed in critical ventilation systems that provide confinement of toxic or nuclear materials, creating the potential for release of hazardous or radioactive materials to the environment. Installation of unqualified HEPA filters has occurred without awareness of the facility operator and without subsequent evaluation of the impact on facility safety. The use of improperly rated or unqualified HEPA filters caused no undesirable releases of hazardous or radioactive material to the environment at Savannah River and LANL.

Event Causes

Savannah River Site

Investigators determined that the direct cause of the event was failure to disseminate changes to filter capacity test ratings in the Process Ventilation Guide to personnel acquiring, installing, and operating ventilation equipment. Waste Management Engineering did not receive information from the Procurement Cycle System, which specifies who should receive technical changes. Failure to disseminate test-flow parameter changes to users resulted in air-flow rates in excess of the design capacity of the filters. Investigators further determined that the root cause of the occurrence was a misunderstanding of documented flow rate.

Los Alamos National Laboratory

LANL investigators determined that the direct and root cause of the event was defective or failed material, because two filters failed the rough-handling and wet-overpressurization-resistance requirements specified in MIL-F-51068F.³ However, procurement process problems contributed to the event and were addressed by subsequent corrective actions.

Procurement of Flanders HEPA filters proceeded for over a year without formal part qualification documentation with the expectation that all of the filters were or would be qualified and that the filter manufacturer (or the DOE filter test facility) would provide the necessary qualification data. The filters were procured even though they were not listed on the Qualified Products List⁹¹⁰ and there was no documentation showing the filters had passed alternate qualification tests.^3,7,8,11 Flanders Filter Inc. elected to not submit their nuclear (Type B) filters to an approved test facility for requalification every 5 years as required by MIL-F- 51068F; therefore, on January 1, 1990, these filters were removed from the Qualified Products List. Units purchased after that date were technically unqualified for use in radionuclide environmental protection applications according to DOE Nuclear Standard NE F 3-42, paragraph 4.¹² Throughout the procurement period, Flanders agreed to provide test data to LANL that would demonstrate qualification of the product. After repeated attempts, LANL received only a portion of the required data.

Corrective Actions

Savannah River Site

At the time of the initial inspection, the 1-H and 2-H evaporators were shut down. Upon discovering that the flow rate through the HEPA filters in the 2-H evaporator cell ventilation system exceeded the filter rating, the 2-F evaporator was also shut down pending evaluation.

After an initial evaluation, facility personnel decided to operate the 1,400-scfm air flow through two parallel 1,000-scfm HEPA filters, which provided a total capacity of 2,000 scfm. This precluded the convenient practice of changing out one filter while continuing to run on the second. However, there was ample opportunity to change filters during facility shutdown for other reasons.

Investigators also discovered contaminated corrosion in ductwork downstream of the HEPA filters on the 2-H Evaporator ventilation system. The ductwork was decontaminated and two additional parallel 1,000-scfm HEPA filters were temporarily added to the outlet of the original exhaust stack. The extra filters assured capture of corrosion that could have broken free from the old ducting. The contaminated components have since been replaced.

Savannah River personnel planned and implemented the following additional corrective actions.

1. The Procurement Cycle System database was updated to list the Waste Management Group as users of Flanders filter 50-16610, thus initiating notification to the group of changes associated with filter procurement including qualification data.

2. Site personnel evaluated the impact of exceeding HEPA filter design flow rates and determined that the fundamental performance requirement for qualification and acceptance (MIL-F-51068F³ and Nuclear Standard NE F-3-45⁹ of a new HEPA filter is reduction of particle concentration by 99.97 percent at rated air flow with a pressure drop not exceeding 1-inch WC (1.3 inches WC for flow capacities of 125 scfm or less). The fundamental structural requirement for qualification is the capability of a preconditioned wet filter to withstand an inlet-to-outlet differential pressure of 10-inches WC for a specified time. The structural requirement assures that the filter will not deteriorate, blow through, or otherwise fail and allow contaminants to pass. In normal filter usage, as particles or moisture build on the filter media, the differential pressure typically increases to levels greater than 1- inch WC. The operator must then decide the differential pressure at which the filters must be replaced. The high-operating-alarm limit for the evaporator ventilation system was set at 3.5-inches WC to trigger scheduling of a filter change. For additional assurance, a high-high alarm was set at six inches.

3. Site personnel evaluated the 2-H and 2-F evaporator ventilation systems to determine acceptable air flow limits for existing HEPA filter systems.

4. Facility personnel installed gauges to measure evaporator cell vacuum at various flow rates at the H- and F-Area tank farms. At 800 scfm, the measured cell vacuum of about 0.14-inches WC was sufficient to satisfy the required minimum of 0.1 inches. This information permitted reduction of the facility flow rate to within the capacity of one 1,000-scfm HEPA filter in any future modification. This decision was pending an evaluation of an overall facility upgrade.

5. Facility personnel developed a ventilation test and checkout plan for evaporator ventilation systems for the H-Area tank farm. Managers assigned actions to develop a similar plan for the F-Area tank farm.

6. Site personnel continued to evaluate all tank farm ventilation systems to determine if operation was in accordance with the design basis. The evaluation included walkdowns, review of documentation, and upgrade of facility drawings.

Los Alamos National Laboratory

In October 1992, the Industrial Hygiene and Operational Safety Group requested and received from Materials Management the procurement record of the Flanders HEPA filters. The procurement records identified which groups had requested filters since January 1990, and revealed that approximately 180 of the HEPA filter models of concern were procured after the Qualified Parts List expiration date. Those remaining in the Materials Management warehouse were impounded. Facility personnel initiated a laboratory-wide search for in-service, unqualified filters. Investigators addressed environmental, safety, and health concerns of specific filter installations.

In the Plutonium Processing Facility at Technical Area 55, none of the subject HEPA filters were installed between nuclear materials storage or processes and the environment. In the Chemistry and Metallurgy Research Facilities, subject filters were not installed at the final stage. Filters were located at a number of other sites throughout the laboratory. Industrial Hygiene, Operational Safety Group, and Materials Management personnel assessed the potential impact at each site and initiated replacement according to usage.

Facility personnel revised the HEPA filter procurement standard to ensure qualification of all filters in accordance with MIL-F-51068 and inclusion of test data with filter shipments.

Hazard Reduction

Personnel can be confused by or misinterpret HEPA filter specifications when procuring and using the filters. An example is the difference in rated flow of the same filter in MIL-F- 51068F³ and Nuclear Standard NE F 3-45.⁸ NE F 3-45 is more stringent than MIL-F-51068F in terms of permitted pressure drop at test flow rates for filters rated above 1,000 scfm. Consequently, a filter procured under NE F 3-45 could be rated at 1,000 scfm; but when procured under MIL-F-51068F, the same filter could be rated at 1,500 scfm. Personnel involved in procurement of HEPA filters for DOE facilities should be aware of the applicable procurement specification details. In spite of such inconsistencies, the use of standardized specifications for procurement of HEPA filters has contributed to an increased level of safety at DOE facilities and should be applied to procurement of other safety- related equipment when possible.

In recent years, the DOE Filter Test Facilities have changed policies for reporting acceptance test results under NE F 3-45, which should help assure procurement and receipt of qualified filters. In the past, if a filter exceeded the maximum pressure drop at the nominal rated flow, the Test Facility was permitted by paragraph 5.4 of NE F 3-45 to report the measured (reduced) airflow at maximum pressure drop as the rated air flow. The customer, purchaser, or user had the option to approve, by waiver, the down-rated capacity and accept the filters. With the current policy, if the maximum pressure drop is exceeded at rated flow, the Test Facility rejects the filter, initiates a rejection notice, and disposes the rejected filters according to an agreement between the purchaser and the manufacturer. Generally, rejected filters are discarded at the Test Facility or returned to the manufacturer.

There has also been confusion as to the rated flow capacity of a filter. The Test Facilities now define rated flow as that called out in the procurement documentation. Flow capacity ratings in manufacturer literature, government standards, or specifications that are not called out in the procurement documentation are not used to select acceptance-test flow rates.

Except for penetration and flow resistance requirements, current policy allows the user facility to waive some of the remaining NE F 3-45 requirements that were not met by specific filters and to take delivery. DOE Nuclear Standard NE F 3-45⁸ and others^12,14,15,16 apply in accordance with paragraph 4 of NE F 3-42¹² to HEPA filters intended for radionuclide environmental protection applications.

In January 1993, Martin Marietta Energy Systems released a Radiation Safety Bulletin¹⁷ on the use of wood-frame HEPA filters in high moisture applications. Wood frames can deteriorate enough to disintegrate during removal, increasing the risk of radiation exposure and contamination. Wood frame filters in high-moisture applications should be replaced with metal-frame filters if allowed by safety documentation. Safety documentation that prohibits this exchange should be modified, if possible, to allow the replacement.

The following practices should reduce safety and compliance problems associated with procurement and use of HEPA filters.

• When procuring HEPA filters, especially for radionuclide applications, both the purchasing and user organizations should verify that the offered parts are listed in the Qualified Parts List or that successful completion of qualification tests are documented and data is current.

• If the Qualified Parts List is used for procurement, the filter type listing should be tracked to a specific manufacturer part number. The filter type listing may be general.

• End users and purchasers should be designated to receive all notifications regarding HEPA filter qualification, ratings, or other indicators of fitness for use.

• Initial and periodic in-place penetration and flow resistance tests should be performed to ensure satisfactory post-installation performance and to determine when filter changes are required.

• A consistent rating definition for HEPA flow capacity should be established. For example, the three DOE Filter Test Facility sites at Hanford, Oak Ridge, and Rocky Flats test according to the filter flow capacity specified in procurement documentation.

• Configuration management of heating, ventilating, and air handling systems that include HEPA filters should be maintained. Design and operating conditions of such systems should be documented and installed filters should be compatible with system and facility design. DOE-STD-1073-93, Guide for Operational Configuration Management Program¹⁸, contains program criteria and implementation guidance for an operational configuration management pro-gram for DOE facilities. As stated in the Standard, the objective of an operational configuration management program is to establish consistency among design requirements, physical configuration, and facility documentation and to maintain this consistency throughout the life of the facility.

• Qualify before you buy.

References

1. ORPS Report SR--WSRC-HTANK-1992-0026, "2-H Evaporator Ventilation Flowrate Exceeds HEPA Filter Rated Flowrate (U)," July 1, 1992; July 20, 1992; August 31, 1992; February 3, 1993.

2. DPSPM-GEN-101, Savannah River Site SRP Process Ventilation Design Guide, Section 9, 1989.

3. MIL-F-51068F, Military Specification, Filters, Particulate (High-Efficiency Fire Resistant), August 11, 1988.

4. ORPS Report ALO-LA-LANL-ESHSUPT-1992-0010, "HEPA Filters Not Meeting DOE Qualification Requirements," November 6, 1992; November 13, 1992; September 15, 1993.

5. DOE 6430.1A, General Design Criteria, pp. 15-39, April 6, 1989.

6. ANSI/UL 586-1990, High-Efficiency, Particulate, Air Filter Units, October 18, 1990.

7. ASME N509-1989, Nuclear Power Plant Air-Cleaning Units and Components, 1989.

8. DOE Nuclear Standard NE F 3-45, Specifications for HEPA Filters Used by DOE Contractors, October 1988.

9. QPL-51068-4, "Qualified Products List" of Military Specification MIL-F-51068, Filters, Particulate, High-Efficiency Fire Resistant, January 24, 1992

10. QPL-51079-5, "Qualified Products List" of Military Specification MIL-F-51079, Filter Medium, Fire-Resistant, High- Efficiency, January 24, 1992

11. MIL-F-51079D, Filter Medium, Fire-Resistant, High-Efficiency, March 14, 1985.

12. DOE Nuclear Standard NE F 3-42, Operating Policy of DOE Filter Test Program, October 1988.

13. Memorandum, P. Worthington (EH-12) to R.L. Trevillian (EH-11), subject: "Comments on the Draft Notice on Flanders Filters," December 22, 1993.

14. DOE Nuclear Standard NE F 3-41T, In-place Testing of HEPA Filter Systems by the Single-Particle, Particle-Size Spectrometer Method, December 1981.

15. DOE Nuclear Standard NE F 3-43, Quality Assurance Testing of HEPA Filters, February 1990.

16. DOE Nuclear Standard NE F 3-44, DOE Filter Test Facilities Quality Program Plan, July 1986.

17. Radiation Safety Bulletin, "Wood Frame HEPA Filters," Martin Marietta Energy Systems, January 1993.

18. DOE-STD-1073-93, Guide for Operational Configuration Management Program, November 1993.