This article was downloaded by: [Umeå University Library] On: 19 November 2014, At: 17:22 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Journal of Occupational and Environmental Hygiene Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/uoeh20
Use of Direct Versus Indirect Preparation Data for Assessing Risk Associated with Airborne Exposures at Asbestos-contaminated Sites a
Mary Patricia Goldade & Wendy Pott O’Brien
b
a
Office of Technical and Management Services , United States Environmental Protection Agency Region 8 , Denver , Colorado b
Ecosystems Protection and Remediation , United States Environmental Protection Agency Region 8 , Denver , Colorado Accepted author version posted online: 20 Sep 2013.Published online: 27 Nov 2013.
To cite this article: Mary Patricia Goldade & Wendy Pott O’Brien (2014) Use of Direct Versus Indirect Preparation Data for Assessing Risk Associated with Airborne Exposures at Asbestos-contaminated Sites, Journal of Occupational and Environmental Hygiene, 11:2, 67-76, DOI: 10.1080/15459624.2013.843779 To link to this article: http://dx.doi.org/10.1080/15459624.2013.843779
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
Journal of Occupational and Environmental Hygiene, 11: 67–76 ISSN: 1545-9624 print / 1545-9632 online DOI: 10.1080/15459624.2013.843779
Use of Direct Versus Indirect Preparation Data for Assessing Risk Associated with Airborne Exposures at Asbestos-contaminated Sites
Downloaded by [Umeå University Library] at 17:22 19 November 2014
Mary Patricia Goldade1 and Wendy Pott O’Brien2 1 Office of Technical and Management Services, United States Environmental Protection Agency Region 8, Denver, Colorado 2 Ecosystems Protection and Remediation, United States Environmental Protection Agency Region 8, Denver, Colorado
At asbestos-contaminated sites, exposure assessment requires measurement of airborne asbestos concentrations; however, the choice of preparation steps employed in the analysis has been debated vigorously among members of the asbestos exposure and risk assessment communities for many years. This study finds that the choice of preparation technique used in estimating airborne amphibole asbestos exposures for risk assessment is generally not a significant source of uncertainty. Conventionally, the indirect preparation method has been less preferred by some because it is purported to result in false elevations in airborne asbestos concentrations, when compared to direct analysis of air filters. However, airborne asbestos sampling in non-occupational settings is challenging because non-asbestos particles can interfere with the asbestos measurements, sometimes necessitating analysis via indirect preparation. To evaluate whether exposure concentrations derived from direct versus indirect preparation techniques differed significantly, paired measurements of airborne Libby-type amphibole, prepared using both techniques, were compared. For the evaluation, 31 paired direct and indirect preparations originating from the same air filters were analyzed for Libby-type amphibole using transmission electron microscopy. On average, the total Libby-type amphibole airborne exposure concentration was 3.3 times higher for indirect preparation analysis than for its paired direct preparation analysis (standard deviation = 4.1), a difference which is not statistically significant (p = 0.12, two-tailed, Wilcoxon signed rank test). The results suggest that the magnitude of the difference may be larger for shorter particles. Overall, neither preparation technique (direct or indirect) preferentially generates more precise and unbiased data for airborne Libby-type amphibole concentration estimates. The indirect preparation method is reasonable for estimating Libby-type amphibole exposure and may be necessary given the challenges of sampling in environmental settings. Relative to the larger context of uncertainties inherent in the risk assessment process, uncertainties associated with the use of airborne Libby-type amphibole exposure measurements derived from indirect preparation analysis are low. Use of exposure measurements generated by either direct or indirect preparation analyses is reasonable to estimate Libby-type Amphibole exposures in a risk assessment.
Keywords
asbestos, exposure assessment, direct, indirect, risk assessment, Libby Amphibole
Address correspondence to Mary Goldade, Express Delivery, U.S. Environmental Protection Agency Region 8, 1595 Wynkoop Street, Mail Code: 8TMS-IO, Denver, Colorado, 80202; e-mail: goldade. [email protected]
INTRODUCTION
S
ix miles northeast of the town of Libby, Montana, is an extensive vermiculite ore deposit, which is co-located with a deposit of amphibole asbestos(1) referred as “Libby-type Amphibole” or LA. Between about 1920 and 1990, the Libby mine was a major source of vermiculite ore for the United States and for the world.(2) Numerous studies of workers from the Libby mine have demonstrated increased morbidity and mortality due to malignant and non-malignant diseases associated with exposure to LA.(3–14) Increased risk for the development of mesothelioma among manufacturing workers who expanded and processed Libby vermiculite has been reported, (15) as has increased morbidity in workers from an Ohio facility that used Libby vermiculite in the manufacturing of lawn care and other agricultural products.(16,17) Additionally, epidemiological investigations of Libby residents with no known occupational exposure to LA-contaminated vermiculite have documented respiratory symptoms, radiographic abnormalities, and adverse health effects.(11,12,18–21) In its initial investigative and subsequent cleanup activities, the U.S. Environmental Protection Agency (EPA) conducted evaluations for environmental exposures to asbestos contamination in Libby in accord with U.S. Occupational Safety and Health Administration (OSHA) regulations.(22,23) Breathing zone exposure measurements were collected among personnel
Journal of Occupational and Environmental Hygiene
February 2014
67
Downloaded by [Umeå University Library] at 17:22 19 November 2014
involved in initial investigation and cleanup activities in Libby. These breathing zone air samples, representing airborne LA exposure measurements, were analyzed by phase contrast microscopy (PCM); confirmation by transmission electron microscopy (TEM), though not required, was performed on a subset of samples. The goal of the monitoring was to measure breathing zone concentrations of airborne LA associated with disturbance of source materials in residential scenarios. Optimally, electron microscopy analysis of air samples collected for asbestos exposure monitoring is performed on the original filter media with as little additional preparation as possible using a direct transfer technique. For convenience, this approach is termed direct preparation analysis. Historically, the vast majority of airborne asbestos exposure measurements were collected for the purposes of monitoring occupational exposures in which asbestos was a known component of the production process. Under these conditions, the asbestos levels were generally the primary source of fibrous contamination. Given the lack of interfering particulates, collecting an air sample with sufficient air volume for adequate analytical sensitivity without filter overload was feasible, thus permitting direct preparation analysis. Often, exposure measurement in environmental settings is not as straightforward as in occupational settings. Sampling and analysis of residential airborne asbestos levels present significant challenges that prevent direct preparation analysis, owing to an abundance of interfering non-asbestos particulates (e.g., inorganic dust and soil particulates, vermiculite, human or pet hair, cigarette or wood smoke, yard organics such as leaves and grass, and so on), frequently at levels much higher than the asbestos itself. Overloading occurs when particles overlay or obscure one another. Direct preparation analysis of overloaded filters may result in inaccurate asbestos exposure measurements due to difficulty observing the presence of asbestos particles. Strategies to prevent filter overloading during sampling include reducing the air collection volume and/or reducing the sample collection time; however, representativeness of the samples regarding actual human exposure levels may be sacrificed when such strategies are employed. Alternatively, an indirect preparation analysis may be performed on overloaded filters, whereby a dilution by indirect transfer technique reduces the particulate loading to improve observation of individual asbestos particles. Experience at the Libby Asbestos Superfund Site has shown that even after modifying the air collection volume and/or sample collection time, many residential activities generate a large amount of insoluble inorganic dust, an interfering particulate. Dilution via indirect preparation analysis is often necessary to obtain an estimate of airborne asbestos concentration. However, concerns raised regarding the use of indirect preparations as opposed to direct preparations for estimating asbestos levels(24) prompted this research to compare the two preparation approaches. The primary concern stemmed from reports of an increase in the number of countable particles for indirect preparations that may affect the resulting exposure estimate.(24,25) A long-held hypothesis suggests ultrasonication 68
treatments used in indirect preparations may cause some asbestos particles to disaggregate into smaller particles;(24,26–29) however, others have proposed alternative theories.(30,31) Several studies have investigated the effect of indirect preparation on estimates of asbestos concentration. Generally, the magnitude of any difference between direct and indirect preparation for an asbestos sample depends on two key factors: the mineralogy of the asbestos (chrysotile versus amphibole mineralogy), and preparation steps taken in the two transfer techniques.(24,26–34) Studies of amphibole minerals demonstrated that the relationship between the two methods was comparable, usually within one order of magnitude.(29,35) In contrast, comparison studies for chrysotile reported a wider range of difference in measured chrysotile levels for the two preparation techniques, anywhere from one to three orders of magnitude.(24,27,29–31) Given the concerns regarding the effect of indirect preparation analysis on exposure measurements as well as the expense associated with planning and conducting representative exposure sampling in residential locations, mobilizing the sampling crew, and analyzing the samples by TEM, the authors sought to obtain empirical data specific to LA regarding the comparability of direct and indirect preparation analyses. That is, the overall purpose of this study was to determine whether estimates of airborne LA exposure concentrations differed significantly for direct versus indirect preparation techniques, which could limit confidence in the use of indirect preparation data in exposure assessments intended for use in risk assessment at the Libby Asbestos Superfund Site. METHODS
T
he air samples used in this study represent exposure monitoring conducted during EPA’s investigation and cleanup efforts in the Libby area. The archived breathing zone samples originally were collected during activities that disturbed LAcontaminated dust, soil, or other source materials and then analyzed by PCM in accord with OSHA regulations.(22,23) Because PCM cannot discriminate between asbestos and other non-asbestos particles, data generated by PCM may represent an overestimate of airborne asbestos concentrations, which may require confirmatory analysis by TEM for regulatory purposes.(22,23) Although the time-weighted averages (TWAs) of samples selected for this study did not require confirmatory analysis as per OSHA regulations, EPA routinely performed confirmatory testing using TEM to better understand LA levels and particle size distributions of LA in air. Due to the large number of TEM analyses performed at Libby and the resulting need to gain efficiencies in analytical costs, EPA’s voluntary confirmatory testing relied on the TEM preparation and analysis method generally described in the Asbestos Hazard Emergency Response Act (AHERA), but modified to record the individual particle dimensions rather than simply binning the number of particles as required by AHERA.(36) Thus, the initial group of archived samples available for possible inclusion in the evaluation consisted of 77 breathing zone samples
Journal of Occupational and Environmental Hygiene
February 2014
Downloaded by [Umeå University Library] at 17:22 19 November 2014
originally prepared using the direct preparation procedure and analyzed by TEM according to the modified AHERA methodology. From these archived samples, a pool of candidate air filters was identified for this retrospective study. Originally collected between the years 2000 and 2005, the filters were evaluated using the direct preparation analysis and thus known to have measureable quantities of LA. Indirect preparations were generated using portions of the same air filters from which direct preparations were produced, thereby enabling paired comparisons of direct and indirect preparation analyses. Direct transfer preparation and TEM analysis were performed at two satellite locations of a parent laboratory. For the subsequent indirect transfer preparations and TEM analyses, variability associated with laboratory preparation and analysis was controlled to the extent possible by restricting these activities to only one of the satellite laboratories performing the original direct preparation and analysis. In addition, the single laboratory performing the TEM analysis for indirect preparations used the same particle counting procedures employed in the analysis of the direct preparations. Selection of Breathing Zone Samples for Reanalysis Thirty-one of the 77 archived airborne exposure monitoring samples that met study criteria were selected randomly for indirect preparation analysis. For the selected samples, original breathing zone sample collection occurred primarily at outdoor locations in the Libby area. Twenty-six of the 31 airborne exposure measurements were collected during activities taking place outside. The remaining five breathing zone samples were collected at indoor locations in Libby. The most common interfering particulates for samples collected outdoors in Libby are insoluble inorganic dust particulates. The sample collection medium consisted of an electrically conductive cassette assembly that included a 50 mm extension cowl and mixed cellulose ester (MCE) membrane filter having a 25 mm diameter and 0.8 μm pore size. To ensure samples contained measurable levels of LA and to avoid selection of filters having few or no LA particles, a minimum LA loading criterion was established. Filters selected for indirect preparation analysis required LA loading of 100 structures per square millimeter (s/mm2) or more, based upon results from the direct preparation analysis. Actual LA loading for selected membrane filters ranged from about 120 s/mm2 to 1100 s/mm2 and averaged about 260 s/mm2. Only intact archived filters with at least one-half of the original membrane filter were candidates for indirect preparation analysis. Original Direct Transfer Sample Preparation Shortly after sampling, the two satellite laboratories performed the original direct transfer sample preparation and TEM analysis of the exposure monitoring samples. The direct transfer technique generally retains the spatial orientation of airborne particulates by collapsing the filter and creating a film of the intact filter surface. For these samples, the laboratory analysts placed portions of each MCE membrane filter onto a
microscope slide with the sample face upward and collapsed the filter using acetone vapor. The surface of the cleared MCE membrane filter was then etched using a low-temperature plasma asher, carbon coated, and mounted onto 200-mesh copper grids for TEM analysis.(36) A minimum of two grids was prepared for each air sample. Indirect Transfer Sample Preparation The indirect transfer technique enables analysis of overloaded air filters by disaggregating particulates that obscure or overlie one another. One of the two satellite laboratories that originally performed direct preparation analysis also performed indirect transfer sample preparation for each of the 31 original archived membrane filters. For each sample, laboratory analysts suspended approximately a one-half portion of the original MCE membrane filter into 100 milliliter (mL) of fiber-free deionized water. In cases where only half the original filter was available in archive, the entire filter was consumed. The suspension was then treated with ultrasonication (100 W for 15 minutes) from which the laboratory analyst volumetrically transferred an aliquot (25–100 mL) of the suspension onto a secondary membrane filter (MCE, 25 mm diameter with 0.8 μm pore size) using a disposable filtration unit.(36–38) The secondary filter was prepared in a similar fashion as the primary filter for the direct transfer technique. Because about one-half of the primary membrane filter was used, the effective filter areas for the primary and secondary filters were about one-half of 385 (i.e., about 192.5) square millimeters (mm2) and 201 mm2, respectively. Portions of each secondary MCE membrane filter were placed onto a microscope slide with the sample face upward and collapsed using acetone vapor. The surface of the cleared secondary MCE membrane filter was then etched using a low-temperature plasma asher, carbon coated, and mounted onto 200-mesh copper grids for TEM analysis.(36) A minimum of two grids was prepared for each membrane filter. In this retrospective study, some preparation steps commonly employed in indirect preparation could not be applied. Typically, when the cassette is first opened for indirect preparation analysis, the filter is removed and the interior of the cassette is washed to remove particulates that may adhere to its sides. For this study, the archived filter had already been analyzed via direct preparation, which involves opening the cassette and removing a portion of the filter. Because the filter cassette had been opened and disturbed previously, cassette washing was not performed as it could not result in a reliable estimation of concentration for particles adhering to the inner sides of the cassette. In addition, the samples were not low-temperature ashed prior to dilution to ensure consistency with the indirect preparation analyses of authentic field samples at Libby. While ashing serves to remove any organic particulates present, 26 of 31 samples were collected to monitor outdoor airborne exposures. The majority of particulate interference for samples in Libby is insoluble inorganic material, which the ashing step cannot resolve.
Journal of Occupational and Environmental Hygiene
February 2014
69
TABLE I. Paired Whole Population ComparisonsA: Airborne Concentrations for Direct and Indirect Preparation Analyses of 31 Air Samples for Libby-type Amphibole (LA)
LA Size Range Total Length ≥5 μm Fraction Length
Journal of Occupational and Environmental Hygiene Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/uoeh20
Use of Direct Versus Indirect Preparation Data for Assessing Risk Associated with Airborne Exposures at Asbestos-contaminated Sites a
Mary Patricia Goldade & Wendy Pott O’Brien
b
a
Office of Technical and Management Services , United States Environmental Protection Agency Region 8 , Denver , Colorado b
Ecosystems Protection and Remediation , United States Environmental Protection Agency Region 8 , Denver , Colorado Accepted author version posted online: 20 Sep 2013.Published online: 27 Nov 2013.
To cite this article: Mary Patricia Goldade & Wendy Pott O’Brien (2014) Use of Direct Versus Indirect Preparation Data for Assessing Risk Associated with Airborne Exposures at Asbestos-contaminated Sites, Journal of Occupational and Environmental Hygiene, 11:2, 67-76, DOI: 10.1080/15459624.2013.843779 To link to this article: http://dx.doi.org/10.1080/15459624.2013.843779
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
Journal of Occupational and Environmental Hygiene, 11: 67–76 ISSN: 1545-9624 print / 1545-9632 online DOI: 10.1080/15459624.2013.843779
Use of Direct Versus Indirect Preparation Data for Assessing Risk Associated with Airborne Exposures at Asbestos-contaminated Sites
Downloaded by [Umeå University Library] at 17:22 19 November 2014
Mary Patricia Goldade1 and Wendy Pott O’Brien2 1 Office of Technical and Management Services, United States Environmental Protection Agency Region 8, Denver, Colorado 2 Ecosystems Protection and Remediation, United States Environmental Protection Agency Region 8, Denver, Colorado
At asbestos-contaminated sites, exposure assessment requires measurement of airborne asbestos concentrations; however, the choice of preparation steps employed in the analysis has been debated vigorously among members of the asbestos exposure and risk assessment communities for many years. This study finds that the choice of preparation technique used in estimating airborne amphibole asbestos exposures for risk assessment is generally not a significant source of uncertainty. Conventionally, the indirect preparation method has been less preferred by some because it is purported to result in false elevations in airborne asbestos concentrations, when compared to direct analysis of air filters. However, airborne asbestos sampling in non-occupational settings is challenging because non-asbestos particles can interfere with the asbestos measurements, sometimes necessitating analysis via indirect preparation. To evaluate whether exposure concentrations derived from direct versus indirect preparation techniques differed significantly, paired measurements of airborne Libby-type amphibole, prepared using both techniques, were compared. For the evaluation, 31 paired direct and indirect preparations originating from the same air filters were analyzed for Libby-type amphibole using transmission electron microscopy. On average, the total Libby-type amphibole airborne exposure concentration was 3.3 times higher for indirect preparation analysis than for its paired direct preparation analysis (standard deviation = 4.1), a difference which is not statistically significant (p = 0.12, two-tailed, Wilcoxon signed rank test). The results suggest that the magnitude of the difference may be larger for shorter particles. Overall, neither preparation technique (direct or indirect) preferentially generates more precise and unbiased data for airborne Libby-type amphibole concentration estimates. The indirect preparation method is reasonable for estimating Libby-type amphibole exposure and may be necessary given the challenges of sampling in environmental settings. Relative to the larger context of uncertainties inherent in the risk assessment process, uncertainties associated with the use of airborne Libby-type amphibole exposure measurements derived from indirect preparation analysis are low. Use of exposure measurements generated by either direct or indirect preparation analyses is reasonable to estimate Libby-type Amphibole exposures in a risk assessment.
Keywords
asbestos, exposure assessment, direct, indirect, risk assessment, Libby Amphibole
Address correspondence to Mary Goldade, Express Delivery, U.S. Environmental Protection Agency Region 8, 1595 Wynkoop Street, Mail Code: 8TMS-IO, Denver, Colorado, 80202; e-mail: goldade. [email protected]
INTRODUCTION
S
ix miles northeast of the town of Libby, Montana, is an extensive vermiculite ore deposit, which is co-located with a deposit of amphibole asbestos(1) referred as “Libby-type Amphibole” or LA. Between about 1920 and 1990, the Libby mine was a major source of vermiculite ore for the United States and for the world.(2) Numerous studies of workers from the Libby mine have demonstrated increased morbidity and mortality due to malignant and non-malignant diseases associated with exposure to LA.(3–14) Increased risk for the development of mesothelioma among manufacturing workers who expanded and processed Libby vermiculite has been reported, (15) as has increased morbidity in workers from an Ohio facility that used Libby vermiculite in the manufacturing of lawn care and other agricultural products.(16,17) Additionally, epidemiological investigations of Libby residents with no known occupational exposure to LA-contaminated vermiculite have documented respiratory symptoms, radiographic abnormalities, and adverse health effects.(11,12,18–21) In its initial investigative and subsequent cleanup activities, the U.S. Environmental Protection Agency (EPA) conducted evaluations for environmental exposures to asbestos contamination in Libby in accord with U.S. Occupational Safety and Health Administration (OSHA) regulations.(22,23) Breathing zone exposure measurements were collected among personnel
Journal of Occupational and Environmental Hygiene
February 2014
67
Downloaded by [Umeå University Library] at 17:22 19 November 2014
involved in initial investigation and cleanup activities in Libby. These breathing zone air samples, representing airborne LA exposure measurements, were analyzed by phase contrast microscopy (PCM); confirmation by transmission electron microscopy (TEM), though not required, was performed on a subset of samples. The goal of the monitoring was to measure breathing zone concentrations of airborne LA associated with disturbance of source materials in residential scenarios. Optimally, electron microscopy analysis of air samples collected for asbestos exposure monitoring is performed on the original filter media with as little additional preparation as possible using a direct transfer technique. For convenience, this approach is termed direct preparation analysis. Historically, the vast majority of airborne asbestos exposure measurements were collected for the purposes of monitoring occupational exposures in which asbestos was a known component of the production process. Under these conditions, the asbestos levels were generally the primary source of fibrous contamination. Given the lack of interfering particulates, collecting an air sample with sufficient air volume for adequate analytical sensitivity without filter overload was feasible, thus permitting direct preparation analysis. Often, exposure measurement in environmental settings is not as straightforward as in occupational settings. Sampling and analysis of residential airborne asbestos levels present significant challenges that prevent direct preparation analysis, owing to an abundance of interfering non-asbestos particulates (e.g., inorganic dust and soil particulates, vermiculite, human or pet hair, cigarette or wood smoke, yard organics such as leaves and grass, and so on), frequently at levels much higher than the asbestos itself. Overloading occurs when particles overlay or obscure one another. Direct preparation analysis of overloaded filters may result in inaccurate asbestos exposure measurements due to difficulty observing the presence of asbestos particles. Strategies to prevent filter overloading during sampling include reducing the air collection volume and/or reducing the sample collection time; however, representativeness of the samples regarding actual human exposure levels may be sacrificed when such strategies are employed. Alternatively, an indirect preparation analysis may be performed on overloaded filters, whereby a dilution by indirect transfer technique reduces the particulate loading to improve observation of individual asbestos particles. Experience at the Libby Asbestos Superfund Site has shown that even after modifying the air collection volume and/or sample collection time, many residential activities generate a large amount of insoluble inorganic dust, an interfering particulate. Dilution via indirect preparation analysis is often necessary to obtain an estimate of airborne asbestos concentration. However, concerns raised regarding the use of indirect preparations as opposed to direct preparations for estimating asbestos levels(24) prompted this research to compare the two preparation approaches. The primary concern stemmed from reports of an increase in the number of countable particles for indirect preparations that may affect the resulting exposure estimate.(24,25) A long-held hypothesis suggests ultrasonication 68
treatments used in indirect preparations may cause some asbestos particles to disaggregate into smaller particles;(24,26–29) however, others have proposed alternative theories.(30,31) Several studies have investigated the effect of indirect preparation on estimates of asbestos concentration. Generally, the magnitude of any difference between direct and indirect preparation for an asbestos sample depends on two key factors: the mineralogy of the asbestos (chrysotile versus amphibole mineralogy), and preparation steps taken in the two transfer techniques.(24,26–34) Studies of amphibole minerals demonstrated that the relationship between the two methods was comparable, usually within one order of magnitude.(29,35) In contrast, comparison studies for chrysotile reported a wider range of difference in measured chrysotile levels for the two preparation techniques, anywhere from one to three orders of magnitude.(24,27,29–31) Given the concerns regarding the effect of indirect preparation analysis on exposure measurements as well as the expense associated with planning and conducting representative exposure sampling in residential locations, mobilizing the sampling crew, and analyzing the samples by TEM, the authors sought to obtain empirical data specific to LA regarding the comparability of direct and indirect preparation analyses. That is, the overall purpose of this study was to determine whether estimates of airborne LA exposure concentrations differed significantly for direct versus indirect preparation techniques, which could limit confidence in the use of indirect preparation data in exposure assessments intended for use in risk assessment at the Libby Asbestos Superfund Site. METHODS
T
he air samples used in this study represent exposure monitoring conducted during EPA’s investigation and cleanup efforts in the Libby area. The archived breathing zone samples originally were collected during activities that disturbed LAcontaminated dust, soil, or other source materials and then analyzed by PCM in accord with OSHA regulations.(22,23) Because PCM cannot discriminate between asbestos and other non-asbestos particles, data generated by PCM may represent an overestimate of airborne asbestos concentrations, which may require confirmatory analysis by TEM for regulatory purposes.(22,23) Although the time-weighted averages (TWAs) of samples selected for this study did not require confirmatory analysis as per OSHA regulations, EPA routinely performed confirmatory testing using TEM to better understand LA levels and particle size distributions of LA in air. Due to the large number of TEM analyses performed at Libby and the resulting need to gain efficiencies in analytical costs, EPA’s voluntary confirmatory testing relied on the TEM preparation and analysis method generally described in the Asbestos Hazard Emergency Response Act (AHERA), but modified to record the individual particle dimensions rather than simply binning the number of particles as required by AHERA.(36) Thus, the initial group of archived samples available for possible inclusion in the evaluation consisted of 77 breathing zone samples
Journal of Occupational and Environmental Hygiene
February 2014
Downloaded by [Umeå University Library] at 17:22 19 November 2014
originally prepared using the direct preparation procedure and analyzed by TEM according to the modified AHERA methodology. From these archived samples, a pool of candidate air filters was identified for this retrospective study. Originally collected between the years 2000 and 2005, the filters were evaluated using the direct preparation analysis and thus known to have measureable quantities of LA. Indirect preparations were generated using portions of the same air filters from which direct preparations were produced, thereby enabling paired comparisons of direct and indirect preparation analyses. Direct transfer preparation and TEM analysis were performed at two satellite locations of a parent laboratory. For the subsequent indirect transfer preparations and TEM analyses, variability associated with laboratory preparation and analysis was controlled to the extent possible by restricting these activities to only one of the satellite laboratories performing the original direct preparation and analysis. In addition, the single laboratory performing the TEM analysis for indirect preparations used the same particle counting procedures employed in the analysis of the direct preparations. Selection of Breathing Zone Samples for Reanalysis Thirty-one of the 77 archived airborne exposure monitoring samples that met study criteria were selected randomly for indirect preparation analysis. For the selected samples, original breathing zone sample collection occurred primarily at outdoor locations in the Libby area. Twenty-six of the 31 airborne exposure measurements were collected during activities taking place outside. The remaining five breathing zone samples were collected at indoor locations in Libby. The most common interfering particulates for samples collected outdoors in Libby are insoluble inorganic dust particulates. The sample collection medium consisted of an electrically conductive cassette assembly that included a 50 mm extension cowl and mixed cellulose ester (MCE) membrane filter having a 25 mm diameter and 0.8 μm pore size. To ensure samples contained measurable levels of LA and to avoid selection of filters having few or no LA particles, a minimum LA loading criterion was established. Filters selected for indirect preparation analysis required LA loading of 100 structures per square millimeter (s/mm2) or more, based upon results from the direct preparation analysis. Actual LA loading for selected membrane filters ranged from about 120 s/mm2 to 1100 s/mm2 and averaged about 260 s/mm2. Only intact archived filters with at least one-half of the original membrane filter were candidates for indirect preparation analysis. Original Direct Transfer Sample Preparation Shortly after sampling, the two satellite laboratories performed the original direct transfer sample preparation and TEM analysis of the exposure monitoring samples. The direct transfer technique generally retains the spatial orientation of airborne particulates by collapsing the filter and creating a film of the intact filter surface. For these samples, the laboratory analysts placed portions of each MCE membrane filter onto a
microscope slide with the sample face upward and collapsed the filter using acetone vapor. The surface of the cleared MCE membrane filter was then etched using a low-temperature plasma asher, carbon coated, and mounted onto 200-mesh copper grids for TEM analysis.(36) A minimum of two grids was prepared for each air sample. Indirect Transfer Sample Preparation The indirect transfer technique enables analysis of overloaded air filters by disaggregating particulates that obscure or overlie one another. One of the two satellite laboratories that originally performed direct preparation analysis also performed indirect transfer sample preparation for each of the 31 original archived membrane filters. For each sample, laboratory analysts suspended approximately a one-half portion of the original MCE membrane filter into 100 milliliter (mL) of fiber-free deionized water. In cases where only half the original filter was available in archive, the entire filter was consumed. The suspension was then treated with ultrasonication (100 W for 15 minutes) from which the laboratory analyst volumetrically transferred an aliquot (25–100 mL) of the suspension onto a secondary membrane filter (MCE, 25 mm diameter with 0.8 μm pore size) using a disposable filtration unit.(36–38) The secondary filter was prepared in a similar fashion as the primary filter for the direct transfer technique. Because about one-half of the primary membrane filter was used, the effective filter areas for the primary and secondary filters were about one-half of 385 (i.e., about 192.5) square millimeters (mm2) and 201 mm2, respectively. Portions of each secondary MCE membrane filter were placed onto a microscope slide with the sample face upward and collapsed using acetone vapor. The surface of the cleared secondary MCE membrane filter was then etched using a low-temperature plasma asher, carbon coated, and mounted onto 200-mesh copper grids for TEM analysis.(36) A minimum of two grids was prepared for each membrane filter. In this retrospective study, some preparation steps commonly employed in indirect preparation could not be applied. Typically, when the cassette is first opened for indirect preparation analysis, the filter is removed and the interior of the cassette is washed to remove particulates that may adhere to its sides. For this study, the archived filter had already been analyzed via direct preparation, which involves opening the cassette and removing a portion of the filter. Because the filter cassette had been opened and disturbed previously, cassette washing was not performed as it could not result in a reliable estimation of concentration for particles adhering to the inner sides of the cassette. In addition, the samples were not low-temperature ashed prior to dilution to ensure consistency with the indirect preparation analyses of authentic field samples at Libby. While ashing serves to remove any organic particulates present, 26 of 31 samples were collected to monitor outdoor airborne exposures. The majority of particulate interference for samples in Libby is insoluble inorganic material, which the ashing step cannot resolve.
Journal of Occupational and Environmental Hygiene
February 2014
69
TABLE I. Paired Whole Population ComparisonsA: Airborne Concentrations for Direct and Indirect Preparation Analyses of 31 Air Samples for Libby-type Amphibole (LA)
LA Size Range Total Length ≥5 μm Fraction Length
