c.
If the percent downstream concentration relative to ambient concentration is >1%, the cassette is poorly sealed. A value of <0.2% can routinely be achieved with a good assembly technique. The only exception appears to be for fibrous (e.g., glass, ceramic) filters without binder. The leak test still works in that a leak is detected, but fibers released from the filter cause overestimation (as much as 20-fold) of the leak rate.
The approximate cost required to setup the pressure drop leak test system is about $500 (the cost of the micromanometer), while the cost to setup the optical or particle count systems is >$1900. Although particle count systems are more expensive, they require less initial setup time and can be used for other industrial hygiene purposes, e.g., environmental monitoring and respirator fit testing (PortaCount, TSI, Inc. St. Paul, MN). The particle leak test does not provide a direct measure of the airflow through the leak because particles contained in the sampled aerosol can be collected on the edge of the filter, on the cassette walls, or on the backup pad as illustrated in Figure 2. The particles lost from the filter can bounce or be re-entrained from the filter surface or they can stick if the particles are liquid. Therefore it is difficult to use the leak test results to predict the amount of material lost during sampling with a leaking cassette. Under a worst case scenario, the loss from the filter can be about 4 times the percent leakage by the particle count leak test. Thus, with a 1% measured leak rate, about 4% of larger (> 2 µm) solid particles can be lost in or through the leak. Filters normally used for industrial hygiene sampling are efficient enough to collect particles from the sampled air with very high efficiency (>99.9%) [1]. Any particles detected downstream are assumed to be from leakage. The test will not work if low efficiency filters or polycarbonate straight-pore (Nuclepore) filters are used, because these will allow some particles to pass directly through the filter. Glass or ceramic fiber filters without binder will release fibers inside the leak and result in an overestimate of the leak size. With relatively large leaks using these fiber filters, the downstream concentration can be larger than the upstream concentration. Assembly of press-fit cassettes should be performed using a press. This allows an even, selected, and repeatable pressure to be applied across the surface of the cassette. Too high a pressure will cut the filter or crack the cassette, while too low a pressure will result in a bypass leak around the filter. Approximately 70 lbs pressure was found to produce a good seal in 25-mm acrylic-copolymer cassettes [2], but optimum pressure must be established for each cassette/filter/backup pad combination. A pneumatic press designed for cassetteclosing (APCSCLSR-2, about $2200, Omega Specialty, Chelmsford MA) has been used successfully for this purpose, though other pneumatic or hydraulic presses can also be used. Tests of several manufacturer’s pre-assembled cassettes indicated these are well-sealed. However, hand assembly of cassettes and even press-assembled cassettes of unusual combinations of filters and backup pads produced high rates of leakage [1]. In these cases, a quality control process including leak testing should be implemented to ensure proper cassette assembly. The appropriate sample size for testing batches of items are described in MIL-STD-105E [5]. If leak test is treated as a pass/fail test of cassette quality, tables are available to determine
3/15/03
180
NIOSH Manual of Analytical Methods
