Analysis of Vinyl Floor Tiles by PLM vs.TEM
Karen Slayer, Laboratory Manager, Asbestos Analytical Services
Wendy Bunner M.Sc., Quality Control Manager, Asbestos Analytical Services
September, 2009
ABSTRACT
The inability to detect fine asbestos fibres (<0.25 µm in diameter) is a known limitation of the PLM (Polarized Light Microscopy) method of analysis due to the resolution limit of the microscope.
In samples such as vinyl floor tiles, this limitation of PLM analysis is further compounded by the interfering binder material which adheres to the asbestos fibres and may prevent analysis of these fibres.
A positive finding of asbestos in a vinyl floor tile can be considered a conclusive result; however, if asbestos is not detected by PLM, there may be doubt as to the accuracy of the result for the reasons given above. The EPA/600/R-93/116 Method suggests using TEM (Transmission Electron Microscopy) as an alternative to PLM to identify asbestos in difficult samples such as vinyl floor tiles.
The Pinchin laboratory decided to test the ability of our analysts to correctly report a negative result on vinyl floor tiles by PLM by sending a random selection of vinyl tile samples to a second laboratory for re-analysis by TEM. The results of the comparison indicate that the Pinchin Laboratory had 99% success in correctly identifying the presence/absence of asbestos in 83 vinyl floor tiles. The study confirms that if analysts are properly trained in analysis of vinyl floor tiles by PLM and perform the analysis following the specified protocols, the analysis of vinyl tile using PLM alone is reliable.
BACKGROUND
The inability to detect very fine fibres (< 0.25µm) is a known limitation of the PLM (Polarized Light Microscopy) method due to the resolution limit of the microscope. Detection of fine fibres in some samples, such as vinyl floor tiles, is further hampered by the interfering binder material. These limitations are cited in the EPA Method 600/R-93-116 (Method for the Determination of Asbestos in Bulk Building Materials) which further states that TEM (Transmission Electron Microscopy) analysis is “particularly applicable to bulk materials that contain a large amount of interfering materials that can be removed by ashing and/or dissolution and contain fibres that are not resolved by PLM techniques”.
In “Advances in Environmental Measurement Methods for Asbestos” (Michael E. Beard and Harry L. Rook, Editors) it is concluded that “The EPA test method is quite capable of producing accurate qualitative results for almost all sample types, with the possible exception of some particularly challenging floor tiles.” A positive finding of asbestos in a vinyl floor tile can be considered a conclusive result; however, if asbestos is not detected by PLM, there may be doubt as to the accuracy of the result for the reasons given above.
The Pinchin Environmental asbestos laboratories have traditionally added the following disclaimer to the certificates of analysis by PLM for vinyl floor tiles when asbestos is not detected by the PLM method.
“Vinyl floor tiles may contain very fine asbestos fibres which are not visible using the PLM method. For confirmation of the absence of asbestos, analysis by Transmission Electron Microscopy (TEM) is necessary.”
Since the EPA method is used in both the NVLAP (National Voluntary Laboratory Accreditation Program through the U.S. National Institute of Standards and Technology) and RTI (Research Triangle Institute) quality control programs, similar comments are commonly added to reports by laboratories accredited through NVLAP Lab Code 101270-0 and NVLAP Lab Code 200795-0.
The combination of low asbestos concentration, small fibre size and the interfering matrix/binder certainly causes vinyl floor tiles to be more difficult (though not impossible) to analyze by PLM compared to most other sample types, however, the question arises “is TEM really necessary for the confirmation of asbestos in vinyl floor tiles?” To that end, Pinchin has performed a program of comparison testing between the PLM and TEM methods on a randomly selected series of vinyl floor tile samples.
Most analytical methods rely on sophisticated equipment to provide analytical results. In contrast, the sensitivity and accuracy of PLM analysis for asbestos greatly depends on the skill and experience of the analyst (as stated in NIOSH Method 9002), and the amount of time devoted to the analysis. Pinchin Environmental Ltd. requires all of its’ analysts to complete a rigorous in-house training program under close supervision by either the Quality Control Manager or Laboratory Manager, which includes completing a minimum of 200 training samples. Analysts are then sent to the U.S. for an intense, one week course in Polarized Light Microscopy presented by a leading authority on the PLM analysis of bulk building materials. We do not impose quotas on the number of analyses required in a day by analysts, but allow them to take the appropriate time, along with peer advice if necessary, to perform a thorough analysis on each and every sample. Therefore, we believe the results presented in the following comparisons are the best possible that may be obtained by the PLM method.
METHOD OF STUDY AND RESULTS
Positive identification of the presence of asbestos in a vinyl floor tile by PLM may be considered a conclusive result (this is about one third of the floor tile samples we test in our laboratory); however, due to the issues given above, particularly that of fine fibre size, obtaining a negative result may not be absolutely conclusive.
To determine our laboratories’ accuracy in correctly obtaining a negative result for floor tiles, i.e. the risk of a false negative; a random selection of vinyl floor tile samples, analyzed by various analysts using PLM, were chosen from a period dating from January, 2007 to August, 2009 and sent to ALS/DataChem in Cincinnati, Ohio for re-analysis by TEM following the standard EPA Method 600/R-93/116. Of the total eighty-three samples in the study, sixty-seven were identified as “None Detected” and sixteen were found to positively contain asbestos by PLM in our laboratory. The results were compared to those obtained by ALS/DataChem for the same samples and are shown on the following chart:
The bars are labelled as:
-/- a negative result by both PLM and TEM
+/+ a positive result by both PLM and TEM
-/+ a negative result by PLM and a positive result by TEM
+/- a positive result by PLM and a negative result by TEM
DISCUSSION OF RESULTS
Of the sixty-seven samples having a negative result by PLM, only one was found to contain Chrysotile asbestos by TEM, in the range of 1-3%. The remaining sixty-six samples had no asbestos content. All sixteen of the samples that were positive by PLM were also positive by TEM. There were no samples that were found to contain asbestos by PLM which came back negative by TEM. This study indicates there is a 99% chance of correctly identifying a floor tile as non-ACM by PLM in our laboratory. In the one case where our laboratory had not detected the presence of asbestos, the asbestos was present in a very low concentration.
According to NVLAP Guidelines (Handbook 150-3, section 5.6.2) “the laboratory shall have a qualitative error rate of less than 1% which is calculated by dividing the number of qualitative errors (false negatives, false positives, and asbestos types) by the total number of Q.A. samples”. The Handbook goes on to state that “the laboratory shall identify problem samples, such as floor tiles, that are difficult to analyze qualitatively and shall have specific written procedures to deal with problem samples to reduce the errors to less than 1%” (section 5.6.3).
The study demonstrates that these concerns over incorrect analysis are minimized at the Pinchin Laboratories through our extensive Quality Management System, rigorous analyst training requirements, and the adequate time spent on each sample to perform a thorough analysis, thereby leading to a high level of confidence in the results. The Pinchin error rate is generally one or two errors per year detected in thousands of Q.A. samples (13,122 in 2008), a fraction of which are vinyl floor tiles. Therefore, our total error rate is substantially lower than 1% and our method of analyzing vinyl floor tiles, as shown in the study, is very effective.
In conclusion, although identifying asbestos in floor tiles using the PLM method may not be 100% accurate, we have confirmed our performance is very high in correctly reporting a negative result. While fibres below the resolution of the PLM microscope may be present in vinyl floor tiles, if asbestos is indeed present in the sample, there will likely be enough larger fibres that a properly trained analyst could identify them and report a positive result, although the quantity estimated may be lower than that determined by TEM. That is to say, in most cases, it is unlikely that all of the fibres present would be so fine that they would be missed and a negative result reported.
Under most regulations, assessment of any test method to be utilized as a control standard would certainly be subject to an error rate of greater than 2%. We consider the results of this study to be conclusive confirmation that the use of PLM alone (as specifically required in IRSST Method 244-2, for bulk analysis in the province of Quebec) is appropriate without further testing by TEM. Unfortunately, we are unable to comment on the analytical accuracy or percentage of false negatives for other laboratories which do not employ the quality measures followed in our laboratories.
It should be noted that standard regulations in some Provinces (e.g. Ontario) and some States in the US, require the analysis of a minimum of three samples of a homogeneous material in order to consider it a non-asbestos-containing material. If in fact three separate samples are analysed of the same tile type this would effectively eliminate the possibility of false negatives since a consistent error would have to be made on all three samples of a specific vinyl tile, a very unlikely event.
REFERENCES
Beard, M.E. & Rook, H.L., Advances in Environmental Measurement Methods for Asbestos, ASTM STP 1342, 1999.
Institut de recherché en santé et en sécurité du travail du Québec: Caractérisation des fibres dans le poussières dèposées ou dans les matériaux en vrac, irsst 244-2, 1999.
NIOSH Method 9002, Issue 2, 15 August, 1994.
National Institute of Standards and Technology, Technology Administration, U.S. Department of Commerce, NVLAP Handbook 150-3, Bulk Asbestos Analysis, 2006 Edition.
Perkins, R.L. & Harvey, B.W., Method for the Determination of Asbestos in Bulk Building Materials, EPA/600/R-93/116, dated July 1993.
