Scan testing filters is critical

To the Editor:

In the “Ask the Experts” section of your May edition of Cleanrooms magazine, I found some information misleading. The question was, ” How much cleaner will my fab be if I use 99.99999 percent ULPA filters?” The response seemed to assume comparison of a scan-tested 99.999 percent filter versus a non-scan tested filter. I agree that specifying the efficiency alone is not enough to ensure acceptable cleanliness. Scan testing of each filter intended for unidirectional flow cleanroom service is critical. The response also seemed to imply that 99.99 percent filters may be just as good. We know that using 99.99 percent filters is generally consistent with Class 100 environments while 99.9995 percent can achieve a Class 10 and 99.99999 percent can achieve Class 1 assuming they are all scan tested to the appropriate level in the factory and the ceiling.

The response also indicates that UL-586 is the traditionally defined method for efficiency testing when actually it is a test method to measure a HEPA (99.97 percent) filter`s ability to provide a certain amount of its efficiency after exposure to flame and heated air. It originated from the U.S. military specification MIL-F-51068 (no longer active) which included a section on heated air testing for HEPA filters intended for nuclear containment service. UL-586 requires the HEPA filter to be tested for efficiency before and after the heated air exposure according to Mil-Std-282 which is the traditional “hot-DOP” method of efficiency testing where the filter is challenged with a near monodispersed aerosol of about 0.3 microns. This efficiency test is not suitable for ULPA filters since their efficiency is beyond the sensitivity of the photometer in the Mil-Std 282 test.

Today, ULPA filters are tested for efficiency using discrete particle counters according to a method that is now defined in IES-RP-CC-007.1 published by the Institute of Environmental Sciences. IES-RP-CC-001.3 provides the recommended efficiency test and scan test for both HEPA and ULPA grades. ULPA filters 99.99999 percent efficiency provide a considerable problem for even the IES-RP-CC-007.1 method since it requires a 95 percent confidence level on the particle counts downstream of the filter. For efficiency levels this high, the time required to test is usually not economical since the downstream counts are so few. Users of these filters typically rely on efficiency testing of samples of the media before assembly into the filters followed by stringent scan testing of each filter at the factory and in the ceiling. The scan test leak criteria for filters of this penetration is often smaller than the traditional 0.01percent also.

Stephen Klocke

Director of Engineering

Flanders Filters, Inc.,

Washington, NC