Andrew F. Oberta, MPH, CIH 

Does it take an air volume on the order of 3,000 L? 

How are sample volumes calculated? The required air volume for a sample to be analyzed by PCM is found by the following equation:
where
This equation is used to determine the required sample volume for PCM clearance on abatement projects. The AHERA regulations effectively dictate a sample volume of 1,200 L and a maximum flow rate of 10 L/min for TEM clearance, but these rules only apply to schools. For nonschool abatement projects, PCM clearance is acceptable and widely used. with samples often taken at 1,200 L or thereabouts at a nominal flow rate of 10 L/min. (Note: PCM clearance sample flow rates are not limited to 10 L/min.) This discussion is entirely about PCM clearance, not TEM clearance. 

Why take a 3,000 L PCM clearance sample? . Arguments have been advanced for a minimum PCM sample volume of 3,000 L based on two references: 1. NIOSH 7400 method 1 NOTE 1 on page 4 states:
A "quantifiable loading" according to page 1 of NIOSH 7400 is 100 to 1300 fibers/mm² of filter surface area. Using at least 3,000 L is intended to help achieve that objective. 2. Page 45 of the EPA "Silver Book" 2 solves the equation for the required volume as follows:
A footnote on page 45 says "The NIOSH 7400 Method lowers the minimum fiber loading to 5 f/100 fields." Actually, it is 5.5 f/100 fields. This page doesn't mention, however, that the NIOSH 7400 method uses a 25mm cassette with a filter area of 385 mm² and a 0.00785 mm² graticule instead of the 37mm cassette and variablearea Porton reticle of the P&CAM 239 method. 3 Solving the above equation with the NIOSH 7400 values gives a required volume of 270 L. I do not suggest using a sample volume this low for clearance but do we need an order of magnitude safety factor? 1 Asbestos and Other Fibers by PCM: Method 7400, Issue 2, dated 15 August 1994, NIOSH Manual of Analytical Methods (NMAM), Fourth Edition, 8/15/94


Why does it matter? • If the samples are collected toward the end of a normal working day, the analyst may have to remain onsite to count the fibers. • If the samples are sent to an offsite laboratory, extra charges for analysis after normal working hours may be incurred. Deadlines for express delivery to an outoftown laboratory may be missed, resulting in further delay in the availability of results. • The longer it takes to receive results, whether it means clearing the area and vacating the site or failing and recleaning, reinspecting and resampling. the more it costs the abatement project participants. • A delay could impact the operations of the facility when work must be performed during specified hours and days. • A delay could also impact the schedule and cost of the renovation contractors who come in afterwards. 

Figure 1 shows the time required to collect a 3,000 L air sample. Depending on how high the pump flow rates can be set (without exceeding the 16 L/min maximum in NIOSH 7400) the downstream impacts of taking a 3,000 L sample can add up. 


How much air volume DO you need? For the answer let’s look at what really happens inside the enclosure during clearance sampling. Figure 2 shows an enclosure with three sampling pumps, and a negative air machine and fan in the background. Figure 3 shows schematically what is happening to the air and fibers inside an enclosure where plaster has been removed from the ceiling and sealer (encapsulant to some) has been applied. During the socalled “settling period” fibers supposedly fall slowly to the floor as shown by the solid line. Actually they follow the air currents as shown by the dotted line and are scavenged by the negative air machine. The makeup air is assumed to be clean and not introducing fibers into the enclosure. (Click on "Makeup air" in Figure 3 for further discussion.) 




How is PCM clearance sampling done? There are, of course, numerous variations within the details of these steps in state regulations and project specifications. For the rest of this discussion, the following parameters will be assumed: 
Figure 4. Aggressive samplng with leaf blower 

Figure 5. Air sampling pump and fan 

What happens to airborne fibers during clearance sampling? We have already seen that they don't settle to the floor but are scavenged by the negative air machine. During clearance sampling as described above, the fiber concentration inside the enclosure is undergoing an exponential decay according to the following equation: where
The exponential decay of airborne contaminants has been demonstrated for industrial hygiene and ventilation applications using tracer gas and smoke particles. For a review of the relevant literature, click here. 

The above equation is for any airborne contaminant in a wellmixed ventilated space and does not depend on the volume of the space as long as the air exchange rate is maintained. This condition obviously does not exist in an enclosure with dead spaces such as side rooms, but it is reasonable to apply it to rooms where the pumps are placed and fans are circulating the air during clearance sampling. At the minimum of four air changes per hour required by OSHA, some state regulations and numerous specifications, the fiber concentration decays with time as shown in Figure 6. The concentration decreases by an order of magnitude every 36 minutes. At the end of that time, an enclosure where the concentration at the start of sampling was at 0.1 f/cc will be down to 0.01 f/cc, the level routinely used for PCM clearance. Figure 6 shows that not only are extended “settling periods” unnecessary, but that clearance sampling for more than two hours (at four or more air changes per hour) is not needed to achieve clearance at abatement sites. 
Figure 6. Exponential decay in ventilated space at four air changes per hour 

How does exponential decay affect clearance sampling? Because the fiber concentration is decreasing with time, the filter collects fibers at a decreasing rate over the sampling period. If 100 fibers are collected during a sampling period of more than 1.80 hours, 90% (90 fibers) will have been collected during the first 0.60 hour (36 minutes), 90% of the renaining 10 fibers, or 9 fibers, during the second 0.60 hour and the remaining fiber during the third 0.60 hour. After 1.80 hours there are no more airborne fibers left to collect. After that, the pumps are sucking clean air. They are looking for "A little man who wasn't there." Table 1 shows the air volume collected at a nominal flow rate of 10 L/min if the pump runs long enough (4.80 hours) to collect a 2,880 L sample. Running the pump at a higher flow rate would not change the percent of the total number of fibers collected during each 0.60 hour interval. 
Table 1. Exponential decay of fiberconcentration during clearance sampling


What about fiber density? To meet the 100 f/mm² fiber density on which the NIOSH 7400 method bases the 3,000 L requirement for low fiber concentrations, one would have to count 78.5 fibers in 100 fields. With an air volume of 3,000 L this gives a concentration of 0.0128 f/cc. Thus, meeting the 100 f/mm² fiber density requirement would mean collecting enough fibers to fail clearance at 0.01 f/cc. In practice, the number of fibers collected on clearance samples is much less, and according to the exponential decay relationship they are all collected in less than two hours, regardless of the actual number of fibers collected. A strict interpretation of the 100 f/mm² fiber density requirement would mean rejecting any sample that did not comply, including most clearance samples. This obviously is not done, nor should it, regardless of the sample volume. 

Any other arguments against a 3,000 L clearance sample? As if the reasons in Why does it matter? aren't enough, consider the following equation for fiber concentration: Assuming that all 100 fields are counted for a clearance sample, the variables in this equation are Fibers and Air volume. We have seen that, because of exponential decay, the number of Fibers will not increase if the pumps are run for more than two hours. Therefore, the numerator in this equation will not increase. However, running the pumps longer increases the Air Volume in the denominator, resulting in a decrease in the value of C. This is misleading. The concentration, C, should be based on the Air volume collected while fibers are being collected. Including the Air volume while the pump is sucking clean air artificially reduces the concentration, C, suggesting that the pumps should run as long as possible  even beyond the five hours to collect a 3,000 L sample  to achieve clearance. This would not only be unnecessary but a corruption of the clearance sampling process. 

What about different air changes and sample volumes? Thus far the examples have used four air changes per hour, the minimum allowed by OSHA and other agencies. Actual air exchange rates will vary and higher rates may be required by regulations, specifications or to maintain a pressure differential between inside and outside the enclosure. Table 2. Sampling times and air volumes for different air exchange rates
The columns on the left of Table 2 give the time required to achieve a 90% reduction in airborne fiber concentration. The times in the “Recommended time and volume” columns are four times those required for a 90% reduction, meaning that 99.99%  effectively all  of the airborne fibers would be removed (see Table 1). For example, at 5.0 air changes per hour, the sampling time would be 115 minutes. (Note: Times have been rounded off and the air volumes are given at 10L/min as a reference value, not a requirement.) Note that, for a specific air exchange rate, the sample volume does not affect the percent of airborne fibers collected during an interval or over the total sampling time. The air exchange rate determines the time for a 90% reduction and, by extension, the recommended sampling time. As long as the air exchange rate is known and maintained, the actual flow rate, and therefore the air volume, can vary from Table 2. In jurisdictions where a minimum air volume for clearance samples is specified by regulations, or to comply with a specification, it may be necessary to increase the flow rate to collect the required air volume in the time shown in Table 2. If the required volume cannot be collected without exceeding the 16 L/min maximum flow rate allowed by NIOSH 7400, it will be necessary to run the pumps longer than the times in Table 2. In the latter event, the pumps will be sucking clean air for part of the time. In an extreme case, collecting a 3,000 L sample at 16 L/min would take 187.5 minutes, which would exceed the 72 minutes required for an 8 ac/hr exchange rate by 115.5 minutes – almost two hours. 

How do I use this sampling strategy? 1. Make sure that all of the bulleted conditions above are clearly described in the specification. 

Are there any other issues to consider? I contend that airborne fibers behave in a manner similar to smoke particles in the same size range. Because the concentration of these smoke partciles has been shown to follow an exponential decay (click here), I contend that the concentration of airborne fibers will do the same. If this hypothesis is challenged and needs to be proven empirically, I have drafted an approach to doing so. The variability in PCM analysis of air samples at the low fiber levels typical of clearance sampling could be raised in support of a 3,000 L sample. The way regulations and specifications are written, a clearance sample either passes or fails, and the 95% Upper Confidence Limit is not used as the criterion. The reported concentration would have to be 0.007 f/cc or lower to be sure the 95% UCL did not exceed 0.01 f/cc. The effect of sample volume on the 95% UCL is minimal – doubling the number of fibers collected only decreases the 95% UCL by 9%, so not much is gained in statistical confidence. Collecting as many fibers as possible is advisable if it can be done by increasing the flow rate, but as I’ve shown it doesn’t help to run the pumps after all the fibers have been removed by the negative air machines. 

What is the point of all this? Time is money on abatement projects and both are wasted by requiring excessive air sample volumes. There is no need for sampling times exceeding 2.4 hours, much less the five hours it could take to collect a 3,000 L sample (at 10 L/min). Air volume should be increased by increasing the flow rate consistent with the capability of the pumps being used, not the sampling time. Time and money can be saved by following the recommended sampling times in Table 2, with the usual caveat that adjustments may be required to accommodate regulations, specifications or projectspecific conditions. I would appreciate feedback to andyobe@aol.com from anyone who tries the sampling strategy in Table 2 or has comments on the exponential decay hypothesis underlying this approach. 
