Workplace Respiratory Protection Factors during Asbestos Removal Operations

Workplace Respiratory Protection Factors during Asbestos Removal Operations Abstract Numerous changes have been made to the French labour regulations in recent years relating to the prevention of risks of exposure to asbestos fibres for operators removing asbestos-containing materials. These changes refer to the method used to count fibres, the collective and personal protective devices to be used on these worksites, and the occupational exposure limit value, which was reduced to 10 f.L−1 on 2 July 2015. In this context, this study assessed the level of respiratory protection afforded by supplied-air respirators and powered air-purifying respirators by monitoring exposure for several operators on nine worksites. The levels of dustiness measured in personal samples taken outside masks showed significant evidence of potential exposure during removal of asbestos-containing plaster or sprayed asbestos, and when using abrasive blasting to treat asbestos-containing materials. For these tasks outside concentration regularly exceeds 25000 f.L−1. Measurements inside masks were generally low, under 10 f.L−1, except in some situations involving the removal of asbestos-containing plaster. This partial penetration of fibres inside masks could be due to the high loading linked to this material. The distributions of Workplace Protection Factors obtained for the two types of respiratory protective devices studied were broad, and the fifth percentile values equal to 236 and 104, respectively, for supplied-air respirators and powered air-purifying respirators. This work highlights once again the need to prioritize collective protection when seeking to prevent asbestos-related risks. asbestos exposure, powered air-purifying respirator, respiratory protection, respirator testing, supplied-air respirator, Workplace Protection Factor Introduction The French regulations define the collective and personal protective measures that should be implemented during operations to remove asbestos-containing materials. These measures are adapted depending on the expected level of dustiness. In addition to collective protective approaches (confinement of the work area, maintaining the area in a depressurized state, aspiration at source, humidification of the material to be removed, etc.), respiratory protective devices (RPDs) may be used in situations resulting in most exposure. RPDs include powered air-purifying respirators (PAPRs), which ensure a minimum flow rate of 160 l min−1 and supplied-air respirators (SARs), which can meet requirements up to 300 l min−1, both with full facepiece mask. The recommendations made in the regulations are based on the Assigned Protection Factor (APF) for these two types of RPD. According to EN 529:2005 (AFNOR, 2005), the APF is the level of protection that can be expected for 95% of well-trained wearers using properly fitted and functioning respirators. In such conditions, measure of the protection is defined by the Workplace Protection Factor (WPF). The WPF corresponds to the level of protection measured on an operator while wearing the RPD in an occupational setting. APF is determined from a distribution of WPF values measured on different operators and workplaces. This distribution generally follows a log-normal law and APF equals the fifth percentile of this distribution. The statistical analysis of the WPF distribution will allow to get an updated APF value for both types of RPD tested. The APF values currently used were determined following several INRS campaigns performed in the 1990s (Villa et al., 1994) and are equal to 60 for PAPRs and 250 for SARs. Since these values were determined, the changes to regulatory constraints and technological advances have led to the development of new RPD models, in particular with the development of pressure demand SARs and an increase in the air flow rates for both PAPRs and SARs. In parallel, new asbestos-containing materials have emerged on worksites which require new techniques for their removal. This is the case, for example, with asbestos-containing plaster, the removal of which started in the early years of this millennium. In addition, since 2012, the French regulations require personal exposure to asbestos fibres to be measured using analytical transmission electron microscopy rather than optical phase-contrast microscopy, the technique used previously (Villa et al., 1994). Finally, since 2 July 2015, the occupational exposure limit (OEL) for asbestos fibres has been reduced from 100 to 10 f.L−1. The objective of this study was, therefore, to determine WPFs for the most protective and most frequently used RPD in the French asbestos removal sector. Methods Modified masks First, a prototype modified mask had to be developed to allow sampling of asbestos fibres inside the mask. Several RPD models among those used for asbestos removal tasks were selected for the study based on their frequency of use and an expected high level of protection on the basis of their APF. Thus, one positive pressure demand SAR and two models of PAPRs (PAPR-1 and PAPR-2) were selected. PAPR-1 is composed of a full face mask and a fan worn at the belt and equipped with two P3 filters. PAPR-2 is composed of a full face mask and the fan is connected to the mask and is equipped with one P3 filter. The sampling system used inside masks was developed in the laboratory in collaboration with manufacturers. The location of the sample inside mask was imposed by two constraints: • the introduction into the mask of a three-pieces open-face cassette for the sampling of asbestos fibres complying with the recommendations of the standard XP X 43-269 (AFNOR, 2012) to guarantee identical sampling conditions between the outside and the inside of the mask and • the development of a system common to all mask models. In a previous study, Chazelet (2016) demonstrated that aerosol sampling through the visor of these different models of SAR and PAPR gave a measurement result identical to that measured at the level of the mouth of the wearer. In the same paper (Chazelet, 2016), measurements of Co/Ci on wearers in laboratory showed that respiratory protection was not affected by the introduction of this sampling in the visor provided that the RPDs were used in accordance with the manufacturer’s recommendations: fit of the mask checked and supply of sufficient breathing air. In order to guarantee reliable measures, a quantitative fit test on all the wearers included in the study must be carried out. As the RPDs have undergone significant changes, its compliance with the European directive 89/686/EEC has had to be verified. Certification tests have ensured this compliance AFNOR, 1998. Equipping operators Full equipment of an operator consisted of two sampling lines for asbestos fibres, one drawing a sample from inside the respirator and one taking a sample from the person’s breathing zone outside the respirator (Fig. 1). Figure 1. View largeDownload slide Operator’s equipment. Figure 1. View largeDownload slide Operator’s equipment. The outside respirator sampling line included a personal sampling pump (Gilair Plus, Sensidyne), made decontaminable, a PVC tubing, and an antistatic open cartridge (37 mm, electrically conducting three-pieces open-face cassette, SKC Omega Specialty Division) placed in the personal breathing zone of the operator. This setup is in accordance with the recommendations set out in standard XP X 43-269 (AFNOR, 2012) and ISO 13794 (1999) for sampling of asbestos fibres, which require a sampling flow rate of 3 l min−1 and the use of an electrically conducting three-pieces open-face cassette [37 mm in diameter and fitted with a buffer and cellulose ester membrane filter (pore size: 45 µm)]. The inside respirator sampling line consisted of a second personal sampling pump (Gilair Plus, Sensidyne), made decontaminable, connected to an antistatic sampling cartridge connected in an air-tight manner to the visor of the modified respiratory mask. The same type of membrane as the first sampling line is used. Measurement campaigns in work situations Description of the worksites Asbestos removal operations in France are subjected to strict regulations, in particular in relation to the operators’ work organization. An asbestos removal operator’s working day is typically broken down into three phases of work in a confined area, with each phase lasting 90 min. These work phases are preceded by suiting-up and followed by decontamination. Between shifts, recovery periods are scheduled. These worksites were selected to allow analysis of a variety of procedures and expected dustiness levels. On each site, two to three operators fitted with the sampling systems illustrated in Fig. 1 were monitored over 3 days. In total on all worksites 25 operators participated in the study. On the first industrial site (A), the operators removed various asbestos-containing materials (surface coatings, sound-proofing, putty) from railway wagons, mainly by pneumatic scraping and sandblasting. On the industrial sites B and C, the operators carried out the removal of sprayed asbestos after watering of the latter. This removal was carried out by scraping with a spatula. The difference between these two workplaces stems from the configuration of the premises, a lot of small areas for the first and a wide area for the second. The organization of work was also different with several operators performing removal and maintenance tasks simultaneously on the first worksite and on the second worksite one operator successively performing the removal and maintenance tasks. Operations of removal of asbestos-containing plaster were followed at worksites D and F in identical configuration but with different removal technique: wet scraping for the first one and scraping; remote-controlled machinery and very high pressure for the second. At the worksite E, the operators carried out thermal insulation removal operations. In the first two workplaces where the PAPRs were evaluated (G and H), operators removed adhesive of floor coverings and tiles by sanding and grinding. On the worksite I, the asbestos-cement pipes were disassembled. Respirators tested The positive pressure demand SAR is tested on worksites A to F. The model of PAPR-1 is evaluated at worksites G and H and the model PAPR-2 at worksite I. Fit test protocol During the first day of the campaign, quantitative fit tests were performed on the operators selected for participation in the study. Operators donned their modified masks without air supply and performed a protocol consisting of eight exercises [seven exercises recommended by OSHA (2009) plus one step exercise]. Fit factors were monitored with a Portacount (TSI). This test was considered successful, and the operator was included in the measurement campaign, only if the fit factor for each of the test exercises exceeded a threshold of 2000 as recommended by HSE (2012). Description of the samplings The sampling campaign per se was performed on the 3 days following these tests. Two or three operators were equipped with the sampling lines according to the site. Samples were only collected while the operator worked in the contaminated area. The sampling duration for air inside the mask corresponded to the full time spent in the contaminated area, i.e. around 90 min per work shift. Several personal samples were collected outside the RPD in the personal breathing zone of the operator to cover the duration of the period. The sampling duration was adjusted depending on the expected dustiness level for the process being studied; it was generally between 30 min and 1 h, and thus two samples collected during each work shift. The flow rate for the two personal sampling pumps was set to 3 l min−1, as calibrated using a bubbling flow meter (Gilibrator) before use. The flow rate was recorded throughout the sampling period (Gilair + pump, Sensidyne). To ensure that sampling was performed as defined in the protocol, an INRS agent was present in the work area throughout. The numbers of inside and outside samples collected on each worksite are given in Table 1. Table 1. Numbers of asbestos samples realized during the study. Worksite ID Number of outside respirator samples Number of inside respirator samples A 60 27 B 42 21 C 36 27 D 52 27 E 47 25 F 41 18 G 36 18 H 32 18 I 17 18 Worksite ID Number of outside respirator samples Number of inside respirator samples A 60 27 B 42 21 C 36 27 D 52 27 E 47 25 F 41 18 G 36 18 H 32 18 I 17 18 View Large Table 1. Numbers of asbestos samples realized during the study. Worksite ID Number of outside respirator samples Number of inside respirator samples A 60 27 B 42 21 C 36 27 D 52 27 E 47 25 F 41 18 G 36 18 H 32 18 I 17 18 Worksite ID Number of outside respirator samples Number of inside respirator samples A 60 27 B 42 21 C 36 27 D 52 27 E 47 25 F 41 18 G 36 18 H 32 18 I 17 18 View Large End-of-sampling protocol A protocol was developed to close the cartridges as soon as the sampling pump was switched off, followed by cleaning of the casing with a wipe and placing them in a closed transport case. The transport case was double-bagged before removing it from the worksite (at the end of the campaign) and subsequently opened under a fume hood in the analysis laboratory. This protocol aimed to limit the risk of contamination of samples as far as possible once sampling cartridges were removed from the work area. Blank measurements Five measurements were performed inside respirators on worksites B, C, D, G, and I to serve as field blanks. These measurements were used to determine the limit of detection (LoD) for the analytical method. During these tests, the pumps associated with these samples were not switched on. Sample analysis For samples taken inside the mask, the analytical sensitivity was set to 1 f.L−1, which is 1/10th of the new OEL applicable in France. In order to reach this analytical sensitivity, whole sampling filters for in-mask samples were treated. For the outside sampling filters, fractions of each filter to cover the whole shift were treated together to return only a single concentration value. The filter fractions treated varied between 1 and 1/8 depending on the overall dustiness level. Grids were prepared by an indirect method (ISO 13794, 1999). As usual in France (Kauffer et al., 1996), for this preparation all or part of the sampling filter (or filters) was totally calcinated in an oxygen plasma oven so as to recover only fibres and mineral particles. Hydrochloric acid (1 N) treatment was used to eliminate any acid-soluble residues. The calcination residues were then re-dispersed and filtered on a pre-carbonated polycarbonate filter (with a pore diameter of 0.2 µm, according to the French standard NF X43-050). A second carbon deposit was then applied to trap the matter between the two layers. The filter was selectively dissolved to allow sample transfer onto microscopy grids. Fibres were then observed and analysed by analytical transmission electron microscopy, based on their morphological, chemical, and structural criteria. To comply with the regulatory exposure controls, both WHO-type asbestos fibres (length L ≥ 5 µm, diameter 0.2 µm < D < 3 µm, L/D ≥ 3) and thin fibres (L ≥5 µm, D ≤ 0.2, L/D ≥ 3) were counted. Counting was stopped when 100 asbestos fibres had been counted or when the analytical sensitivity of the technique was reached. Statistical analysis All statistical analyses were based on validated pairs of external and internal fibre counts. Fifteen couples of inside/outside measurements were excluded from the analysis for the following reasons: one measurement missing, incorrect use of the RPD (operator wearing a PAPR for which the battery indicated an inadequate charge level), or failure to respect the rules for wearing RPD (poorly shaved operator). No data were excluded based on the fibre counts recorded. In total, 132 pairs of external and internal samples were retained to determine the WPFs for SARs, and 47 pairs were retained for PAPRs. The analyses were performed in two steps: first, descriptive analyses are presented for both outside and inside fibre concentrations for the different RPDs and worksites. Secondly, the internal and external fibre counts were modelled using the standard statistical model for count data that is random effect Poisson regression fitted using maximum likelihood (Rabe-Hesketh and Skrondal, 2012). The hypotheses on which this model is based are: the Poisson distribution of asbestos counts, the log-normal distribution of exposure concentrations in workplaces, and an assumed log-normal distribution of the WPFs (see Supplementary Material, available at Annals of Occupational Hygiene online). Note that for this model zero counts are data like any other data. Results Outside samples The results from analysis of the personal samples taken outside the RPD are expressed as concentrations, Co, for the 25 different operators monitored in the nine companies. For the worksites A to F where SAR was used, the mean geometric outside concentration varies between 79 and 32011 f.L−1 (Table 2). For the G to I worksites where PAPRs were used, mean geometric outside concentration varies between 20 and 817 f.L−1 (Table 2). Table 2. Results of outside fibre concentration. Worksite Process Respirator Co (f.L−1) Geometric mean Minimum value Maximum value A Removal of surface coatings (shot peening), sound- proofing (abrasive blasting), putty (scraping) SAR 1075 25 154941 B Removal of sprayed asbestos (manual scraping) SAR 32 011 7606 84731 C Removal of sprayed asbestos (manual scraping) SAR 8511 1828 36074 D Removal of asbestos-containing plaster (pneumatic scraping) SAR 8448 414 235838 E Removal of insulation (manual scraping) SAR 79 6 508 F Removal of asbestos-containing plaster (scraping; remote-controlled machinery, very high pressure) SAR 6604 315 78076 G Removal of adhesives for floor coverings and tiles by chipping and sanding PAPR-1 817 24 8106 H Removal of adhesives for floor coverings and tiles by sanding and grinding PAPR-1 20 3 563 I Removal of asbestos-cement ducts by breaking–dismantling PAPR-2 57 6 303 Worksite Process Respirator Co (f.L−1) Geometric mean Minimum value Maximum value A Removal of surface coatings (shot peening), sound- proofing (abrasive blasting), putty (scraping) SAR 1075 25 154941 B Removal of sprayed asbestos (manual scraping) SAR 32 011 7606 84731 C Removal of sprayed asbestos (manual scraping) SAR 8511 1828 36074 D Removal of asbestos-containing plaster (pneumatic scraping) SAR 8448 414 235838 E Removal of insulation (manual scraping) SAR 79 6 508 F Removal of asbestos-containing plaster (scraping; remote-controlled machinery, very high pressure) SAR 6604 315 78076 G Removal of adhesives for floor coverings and tiles by chipping and sanding PAPR-1 817 24 8106 H Removal of adhesives for floor coverings and tiles by sanding and grinding PAPR-1 20 3 563 I Removal of asbestos-cement ducts by breaking–dismantling PAPR-2 57 6 303 View Large Table 2. Results of outside fibre concentration. Worksite Process Respirator Co (f.L−1) Geometric mean Minimum value Maximum value A Removal of surface coatings (shot peening), sound- proofing (abrasive blasting), putty (scraping) SAR 1075 25 154941 B Removal of sprayed asbestos (manual scraping) SAR 32 011 7606 84731 C Removal of sprayed asbestos (manual scraping) SAR 8511 1828 36074 D Removal of asbestos-containing plaster (pneumatic scraping) SAR 8448 414 235838 E Removal of insulation (manual scraping) SAR 79 6 508 F Removal of asbestos-containing plaster (scraping; remote-controlled machinery, very high pressure) SAR 6604 315 78076 G Removal of adhesives for floor coverings and tiles by chipping and sanding PAPR-1 817 24 8106 H Removal of adhesives for floor coverings and tiles by sanding and grinding PAPR-1 20 3 563 I Removal of asbestos-cement ducts by breaking–dismantling PAPR-2 57 6 303 Worksite Process Respirator Co (f.L−1) Geometric mean Minimum value Maximum value A Removal of surface coatings (shot peening), sound- proofing (abrasive blasting), putty (scraping) SAR 1075 25 154941 B Removal of sprayed asbestos (manual scraping) SAR 32 011 7606 84731 C Removal of sprayed asbestos (manual scraping) SAR 8511 1828 36074 D Removal of asbestos-containing plaster (pneumatic scraping) SAR 8448 414 235838 E Removal of insulation (manual scraping) SAR 79 6 508 F Removal of asbestos-containing plaster (scraping; remote-controlled machinery, very high pressure) SAR 6604 315 78076 G Removal of adhesives for floor coverings and tiles by chipping and sanding PAPR-1 817 24 8106 H Removal of adhesives for floor coverings and tiles by sanding and grinding PAPR-1 20 3 563 I Removal of asbestos-cement ducts by breaking–dismantling PAPR-2 57 6 303 View Large The outside concentration values presented in Table 2 illustrate the broad range of exposure encountered on different worksites and during different removal processes (material removed and technique used). Thus, operations to remove asbestos-containing plaster, sprayed asbestos, and removing surface coatings by abrasive blasting tend to generate very high concentrations of asbestos fibres. Maximal values of Co for worksites A, B, C, D, and F are all over 25000 f.L−1 which corresponds to the maximum fibre concentration allowed by French regulations. The outside concentrations measured on worksites G to I where PAPRs were used are much lower with geometric mean values all <1000 f.L−1. Inside samples Limit of detection No fibres were detected on any of the five field blank filters. Taking the analytical parameters of the microscope into account, these results were used to determine the LoD for the method which equals 1 f.L−1. Two classes of outside concentration can then be defined according to Vaughan and Rajan-Sithamparanadarajah (2005): low and high exposures when Co is, respectively, below or above 10 times the product between the LoD and the APF. Results of transmission electronic microscopy analysis Table 3 presents the results of transmission electron microscopy analysis in simplified terms of the presence (number of fibre Nf ≠ 0) or absence (Nf = 0) of asbestos fibres detected on the analyzed fraction of the inside sample. Table 3. Results of analytical transmission electron microscopy analysis of samples taken inside respirators. Worksite ID Respirator Total number of samples where Nf = 0 (corresponding fraction) Total number of samples where Nf = 0 and high exposure (corresponding fraction) Total number of samples where Nf = 0 and low exposure (corresponding fraction) A SAR 19 (70.4%) 3 (11.1%) 16 (59.3%) B SAR 10 (76.9%) 10 (76.9%) 0 (NA) C SAR 21 (80.8%) 19 (73.1%) 2 (7.7%) D SAR 13 (56.5%) 11 (47.8%) 2 (8.7%) E SAR 19 (76%) 0 (NA) 19 (76%) F SAR 10 (55.6%) 6 (33.3%) 4 (22.2%) G PAPR-1 13 (76.5%) 7 (41.2%) 6 (35.3%) H PAPR-1 15 (88.2%) 0 (NA) 16 (88.9%) I PAPR-2 15 (88.2%) 0 (NA) 15 (88.2%) Worksite ID Respirator Total number of samples where Nf = 0 (corresponding fraction) Total number of samples where Nf = 0 and high exposure (corresponding fraction) Total number of samples where Nf = 0 and low exposure (corresponding fraction) A SAR 19 (70.4%) 3 (11.1%) 16 (59.3%) B SAR 10 (76.9%) 10 (76.9%) 0 (NA) C SAR 21 (80.8%) 19 (73.1%) 2 (7.7%) D SAR 13 (56.5%) 11 (47.8%) 2 (8.7%) E SAR 19 (76%) 0 (NA) 19 (76%) F SAR 10 (55.6%) 6 (33.3%) 4 (22.2%) G PAPR-1 13 (76.5%) 7 (41.2%) 6 (35.3%) H PAPR-1 15 (88.2%) 0 (NA) 16 (88.9%) I PAPR-2 15 (88.2%) 0 (NA) 15 (88.2%) View Large Table 3. Results of analytical transmission electron microscopy analysis of samples taken inside respirators. Worksite ID Respirator Total number of samples where Nf = 0 (corresponding fraction) Total number of samples where Nf = 0 and high exposure (corresponding fraction) Total number of samples where Nf = 0 and low exposure (corresponding fraction) A SAR 19 (70.4%) 3 (11.1%) 16 (59.3%) B SAR 10 (76.9%) 10 (76.9%) 0 (NA) C SAR 21 (80.8%) 19 (73.1%) 2 (7.7%) D SAR 13 (56.5%) 11 (47.8%) 2 (8.7%) E SAR 19 (76%) 0 (NA) 19 (76%) F SAR 10 (55.6%) 6 (33.3%) 4 (22.2%) G PAPR-1 13 (76.5%) 7 (41.2%) 6 (35.3%) H PAPR-1 15 (88.2%) 0 (NA) 16 (88.9%) I PAPR-2 15 (88.2%) 0 (NA) 15 (88.2%) Worksite ID Respirator Total number of samples where Nf = 0 (corresponding fraction) Total number of samples where Nf = 0 and high exposure (corresponding fraction) Total number of samples where Nf = 0 and low exposure (corresponding fraction) A SAR 19 (70.4%) 3 (11.1%) 16 (59.3%) B SAR 10 (76.9%) 10 (76.9%) 0 (NA) C SAR 21 (80.8%) 19 (73.1%) 2 (7.7%) D SAR 13 (56.5%) 11 (47.8%) 2 (8.7%) E SAR 19 (76%) 0 (NA) 19 (76%) F SAR 10 (55.6%) 6 (33.3%) 4 (22.2%) G PAPR-1 13 (76.5%) 7 (41.2%) 6 (35.3%) H PAPR-1 15 (88.2%) 0 (NA) 16 (88.9%) I PAPR-2 15 (88.2%) 0 (NA) 15 (88.2%) View Large Results show that in the case of the asbestos-containing plaster removal process (worksites D and F), direct exposures to asbestos fibres inside respirators are the most numerous with a lowest proportion of samples with no fibre detected in the mask. Table 3 shows that even for low exposure, some fibres were detected on the inside mask samples. Moreover, it was observed that fibres can penetrate inside respirators whatever the worksite, the task, the subject, or the work period. Distribution of WPFs Plotting the concentration measured inside the mask, Ci, against the concentration measured outside the mask, Co, shows that no obvious relationship links both concentrations (Fig. 2). Thus, the WPF is very variable from one worksite to another and even within the same site. Figure 2. View largeDownload slide Raw data from analytical transmission electron microscopy analyses of concentrations measured outside (Co) and inside (Ci) RPD masks (179 pairs). Figure 2. View largeDownload slide Raw data from analytical transmission electron microscopy analyses of concentrations measured outside (Co) and inside (Ci) RPD masks (179 pairs). Results reveal first that in 135 cases of 179 presented on Fig. 2, no asbestos fibres are detected on the inside sampling. Secondly in 7 cases of 179, the penetration of asbestos fibres inside the mask is such that Ci exceeded the 10 f.L−1 limit value during the work shift. These high values of Ci are measured for outside dust concentrations >2000 f.L−1. Moreover, all of these high asbestos exposures inside RPD involved worksites where operators were equipped with SARs and engaged in removing asbestos-containing plaster. Table 1 shows that very high level of Co values were obtained during this removal process. In the case of PAPRs, which are used on much less dusty sites than SARs, Ci values never exceeded 10 f.L−1. A comparison of WPFs obtained for each model of PAPR under similar outside dust concentration conditions (workplaces H and I) led to the conclusion that no significant difference between the two distributions was found. All the measurements carried out on the two models of PAPR have, therefore, been treated jointly to give only one distribution of WPF for this type of RPD. Statistical analysis of the data described in the ‘Methods’ section provided a distribution of the WPFs for the two types of RPD studied. The fifth percentile of these distributions equal 236 and 104, respectively, for SAR and PAPR (Table 4). Table 4. Characteristics of the distribution of WPF. RPD type SAR PAPR Number of values analysed 132 47 Geometric mean (95% confidence interval) 8982 (5411; 14909) 886 (420; 1871) Geometric standard deviation (95% confidence interval) 9.15 (6.65; 13.3) 3.69 (2.62; 5.86) Fifth percentile for the distribution 236 104 RPD type SAR PAPR Number of values analysed 132 47 Geometric mean (95% confidence interval) 8982 (5411; 14909) 886 (420; 1871) Geometric standard deviation (95% confidence interval) 9.15 (6.65; 13.3) 3.69 (2.62; 5.86) Fifth percentile for the distribution 236 104 View Large Table 4. Characteristics of the distribution of WPF. RPD type SAR PAPR Number of values analysed 132 47 Geometric mean (95% confidence interval) 8982 (5411; 14909) 886 (420; 1871) Geometric standard deviation (95% confidence interval) 9.15 (6.65; 13.3) 3.69 (2.62; 5.86) Fifth percentile for the distribution 236 104 RPD type SAR PAPR Number of values analysed 132 47 Geometric mean (95% confidence interval) 8982 (5411; 14909) 886 (420; 1871) Geometric standard deviation (95% confidence interval) 9.15 (6.65; 13.3) 3.69 (2.62; 5.86) Fifth percentile for the distribution 236 104 View Large Discussion The objective of this study was to determine the WPFs for two types of RPD used on asbestos removal worksites: pressure demand SAR and PAPR. Measurements were performed during asbestos removal, examining the processes likely to lead to the highest emissions of asbestos fibres. Worksites where sprayed asbestos removal and asbestos-containing plaster removal were performed were then selected to evaluate the WPF of the SAR. Worksites of removal of tile and floor tile adhesives were selected to evaluate WPFs of the PAPRs. Inside and outside respirator samples were carried out on 25 operators on 9 worksites and were analysed by analytical transmission electron microscopy to obtain WPFs. Inside respirator samples do not contain asbestos fibres in 75% of cases on all sites. But in 4% of cases, the concentration of asbestos fibres inside the masks is higher than the OEL value of 10 f.L−1. These high values of Ci were all measured inside air-supplied respirators worn by operators performing asbestos-containing plaster removal tasks and at outside asbestos concentrations >2500 f.L−1. For the PAPRs, the two models studied have similar WPF distributions. A statistical treatment of the Co and Ci data allowed the determination of two WPF distributions for each type of respirator. The fifth percentiles of these distributions were equal to 104 for PAPR and to 236 for SAR. Few studies exist in the literature to report WPF of SARs or PAPRs. The work of Villa et al. (1994) and Howie et al. (1996) are the only ones that provide WPF values for different models of PAPRs measured during asbestos removal operations at different sites. Villa et al. (1994) describes highly variable WPFs between 140 and >10000 for different operators performing the same removal task. The work of Howie et al. (1996), which has many more measurements than those of Villa et al. (1994), shows that out of 179 Co/Ci ratios, 57 are unquantifiable due to a Ci below the LoD. They also show that in 90% of the cases, the mask internal concentrations are below the limit value. In addition, this article concludes that there is no significant difference between the WPF distributions obtained on the different PAPR models. Our findings are consistent with this work, which at that time led to a proposal for an APF for PAPRs of 40. Campaigns to measure protective factors on PAPRs against other pollutants (Myers, 1983; Lenhart, 1984; Myers et al., 1984) came to the same conclusions from a WPF distribution of log-type normal and with a fifth percentile value <50. To explain the WPF differences measured during the Myers campaign, the differences in carrier behaviour and facial leakage were discussed. Two measurement campaigns focused on the performance of hood style SARs (Nelson et al., 2001; Janssen et al., 2008), both of whom concluded high WPF for this type of APR during sanding and painting operations. Pollutant detection was done by detection of chemical species including strontium in both articles. Nelson et al. (2001), who reports the most data, indicates that in sanding operations 50% of intakes are below the detection limit of strontium compared with 100% in paint operations. Their authors offer APFs over 1000 for this APR. There is no work on the WPF’s pressure demand SAR with full facemask. The low values of WPF obtained, in comparison with those of Nelson and Janssen, can be explained by the use of this SAR model in more restrictive conditions: site configuration (multi-storey work on sites B, D, E, and F), use of scaffolding (sites B, C, D, F). These posture constraints and these efforts can lead to a penetration of fibres in the mask, as evoked by Howie (1996). When Ci is less than the detection limit of the analytical method adapted to the pollutant, the works cited above all propose a calculation of WPF equal to Co divided by the LoD, except Howie which considers that, in this case, WPF > nominal protection factor (NPF). One major feature of the data of the present paper is that a majority of the within mask fibre counts were zero. Thus, no value for a WPF can be obtained from any such data point (pair of counts). However, excluding these data was not an option, as this would have led to a serious bias by selecting only the worst situations in terms of presence of within-mask fibres. In order to cope with these data, we relied on a statistical model fitted using maximum likelihood that allows incorporating all the data including those with zero within-mask fibre counts. This model is based on three hypotheses which are given in Supplementary Material (available at Annals of Occupational Hygiene online) and are standard in the field of industrial hygiene. The numerical values of the fifth percentile of each WPF distributions obtained (236 for SAR and 104 for PAPR) are of the same order of magnitude as those used up to now in the French regulations (250 for SAR and 60 for PAPR) (Villa et al., 1994). However, the level of operator training has increased considerably since the 1990s, as well as the level of collective protective measures implemented on worksites. Moreover, the RPDs assessed in this study present more protective characteristics than the RPDs assessed in the 1990s. As a matter of fact the supplied-air RPD used was equipped with a pressure demand valve which did not exist at that time, and the PAPRs delivered flow rates >160 l min−1. These factors should have led to higher WPFs than those measured in the 1990s. But, today, new asbestos-containing materials, such as asbestos-containing plaster, are being removed, in some cases through the application of new, more aggressive procedures such as very high pressure. These new asbestos removal situations were linked to the highest risks of exposure to asbestos fibres. This result confirms those presented by Radauceanu et al. (2008) in the form of a task-exposure matrix. These results remind us of the limitations of using personal protective equipment as these devices alone may not guarantee optimal worker protection. Development of protective measures should today focus on the reduction of dustiness in the work area as far as possible, though the development of aspiration at source, work on humidified materials, or even mechanization when these measures prove insufficient. Supplementary Data Supplementary data are available at Annals of Work Exposures and Health online. Funding Funding for this project was entirely based on the core funding of the Institut National de Recherche et Sécurité (INRS) which is the sole employer of the authors. Acknowledgements The authors would like to thank the INRS teams who participated in the measurement campaigns and the one who carried out all the microscopic analyzes. The authors declare no conflict of interest relating to the material presented in this article. Its contents, including any opinions and/or conclusions expressed, are solely those of the authors. References Association Française de NORmalisation (AFNOR) . ( 1998 ) NF EN 136:1998. Respiratory protective devices—full face masks—requirements, testing, marking . La Plaine Saint-Denis, France : AFNOR . Association Française de NORmalisation (AFNOR) . ( 2005 ) NF EN ISO 8996:2005. Ergonomics of the thermal environment—determination of metabolic rate . La Plaine Saint-Denis, France : AFNOR . Association Française de NORmalisation (AFNOR) . ( 2012 ) XP X 43-269:2012. Air quality—workplace atmospheres—sampling on membrane filters for the determination of the fibre number concentration by microscopic techniques: phase contrast optical microscopy, scanning electron microscopy analysis and transmission electron microscopy analysis—counting by phase contrast optical microscopy . La Plaine Saint-Denis, France : AFNOR . Chazelet S , Silvente E , Eypert-Blaison C et al. ( 2016 ) Asbestos fibre sampling through a facepiece visor to measure Workplace Protection Factor . J Int Soc Respir Prot ; 33 : 21 – 35 . Health and Safety Executive (HSE) . ( 2012 ) Fit testing of respiratory protective equipment facepieces. HSE OC 282/28 . Internal report to Health and Safety Executive (HSE) . Howie RM , Johnstone JBG , Weston P et al. ( 1996 ) Workplace effectiveness of respiratory protective equipment for asbestos removal work. HSE Contract Research Report n° 112/1996 . Sheffield: Health and Safety Executive . International Organization for Standardization (ISO) . ( 1999 ) ISO 13794:1999. Ambient air. Determination of asbestos fibres. Indirect-transfer transmission electron microscopy procedure . Geneva, Switzerland : International Standardization Organisation . Janssen L , Bidwell J , Cuta K et al. ( 2008 ) Workplace performance of a hood-style supplied-air respirator . J Occup Environ Hyg ; 5 : 438 – 43 . Google Scholar CrossRef Search ADS PubMed Kauffer E , Billon-Galland MA , Vigneron JC et al. ( 1996 ) Effects of preparation methods of airborne concentrations of asbestos fibres by transmission electron microscopy . Ann Occup Hyg ; 40 : 321 – 30 . Google Scholar CrossRef Search ADS Lenhart SW , Campbell DL . ( 1984 ). Assigned protection factors for two respirator types based upon workplace performance testing . Ann Occup Hyg ; 28 : 173 – 182 . doi: 10.1093/annhyg/28.2.173 Google Scholar PubMed Myers W , Peach MJ . ( 1983 ). Performance measurements on a powered air purifying respirator made during field use in silica ragging operation . Ann Occup Hyg ; 27 : 251 – 259 . doi: 10.1093/annhyg/27.3.251 Google Scholar PubMed Myers WR , Peach MJ 3rd , Allender J . ( 1984 ) Workplace Protection Factor measurements on powered air-purifying respirators at a secondary lead smelter–test protocol . Am Ind Hyg Assoc J ; 45 : 236 – 41 . Google Scholar CrossRef Search ADS PubMed Nelson TJ , Wheeler TH , Mustard TS . ( 2001 ) Workplace Protection Factors—supplied air hood . Am Ind Hyg Assoc ; 62 : 96 – 9 . Occupational Safety and Health Administration (OSHA), US Department of Labor . ( 2009 ) Assigned Protection Factors for the revised respiratory protection standard, OSHA 3352-02 . Washington, DC : OSHA Publications Office, U.S. Department of Labor . Rabe-Hesketh S , Skrondal A . ( 2012 ) Multilevel and longitudinal modeling using stata – volume ii: categorical responses, counts and survival . 3rd edn . College Station , TX : Stata Press Publication . Radauceanu A , Chouanière D , Wild P et al. ( 2008 ) Assessment of asbestos exposure in asbestos removal operations: proposal of a task exposure matrix . Archives des maladies professionnelles et de l’Environnement ; 69 : 586 – 92 . Google Scholar CrossRef Search ADS Vaughan N , Rajan-Sithamparanadarajah B . ( 2005 ) Meaningful Workplace Protection Factor measurement: experimental protocols and data treatment . Ann Occup Hyg ; 49 : 549 – 61 . Google Scholar PubMed Villa M , Hubert G , Lima S et al. ( 1994 ) Occupational exposure during asbestos removal operations . J Int Soc Respir Prot ; 12 : 7 – 14 . © The Author(s) 2018. Published by Oxford University Press on behalf of the British Occupational Hygiene Society. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Annals of Work Exposures and Health (formerly Annals Of Occupational Hygiene) Oxford University Press

Workplace Respiratory Protection Factors during Asbestos Removal Operations

Loading next page...
 
/lp/ou_press/workplace-respiratory-protection-factors-during-asbestos-removal-xZgAcVs7rp
Publisher
British Occupational Hygiene Society
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the British Occupational Hygiene Society.
ISSN
2398-7308
eISSN
2398-7316
D.O.I.
10.1093/annweh/wxy013
Publisher site
See Article on Publisher Site

Abstract

Abstract Numerous changes have been made to the French labour regulations in recent years relating to the prevention of risks of exposure to asbestos fibres for operators removing asbestos-containing materials. These changes refer to the method used to count fibres, the collective and personal protective devices to be used on these worksites, and the occupational exposure limit value, which was reduced to 10 f.L−1 on 2 July 2015. In this context, this study assessed the level of respiratory protection afforded by supplied-air respirators and powered air-purifying respirators by monitoring exposure for several operators on nine worksites. The levels of dustiness measured in personal samples taken outside masks showed significant evidence of potential exposure during removal of asbestos-containing plaster or sprayed asbestos, and when using abrasive blasting to treat asbestos-containing materials. For these tasks outside concentration regularly exceeds 25000 f.L−1. Measurements inside masks were generally low, under 10 f.L−1, except in some situations involving the removal of asbestos-containing plaster. This partial penetration of fibres inside masks could be due to the high loading linked to this material. The distributions of Workplace Protection Factors obtained for the two types of respiratory protective devices studied were broad, and the fifth percentile values equal to 236 and 104, respectively, for supplied-air respirators and powered air-purifying respirators. This work highlights once again the need to prioritize collective protection when seeking to prevent asbestos-related risks. asbestos exposure, powered air-purifying respirator, respiratory protection, respirator testing, supplied-air respirator, Workplace Protection Factor Introduction The French regulations define the collective and personal protective measures that should be implemented during operations to remove asbestos-containing materials. These measures are adapted depending on the expected level of dustiness. In addition to collective protective approaches (confinement of the work area, maintaining the area in a depressurized state, aspiration at source, humidification of the material to be removed, etc.), respiratory protective devices (RPDs) may be used in situations resulting in most exposure. RPDs include powered air-purifying respirators (PAPRs), which ensure a minimum flow rate of 160 l min−1 and supplied-air respirators (SARs), which can meet requirements up to 300 l min−1, both with full facepiece mask. The recommendations made in the regulations are based on the Assigned Protection Factor (APF) for these two types of RPD. According to EN 529:2005 (AFNOR, 2005), the APF is the level of protection that can be expected for 95% of well-trained wearers using properly fitted and functioning respirators. In such conditions, measure of the protection is defined by the Workplace Protection Factor (WPF). The WPF corresponds to the level of protection measured on an operator while wearing the RPD in an occupational setting. APF is determined from a distribution of WPF values measured on different operators and workplaces. This distribution generally follows a log-normal law and APF equals the fifth percentile of this distribution. The statistical analysis of the WPF distribution will allow to get an updated APF value for both types of RPD tested. The APF values currently used were determined following several INRS campaigns performed in the 1990s (Villa et al., 1994) and are equal to 60 for PAPRs and 250 for SARs. Since these values were determined, the changes to regulatory constraints and technological advances have led to the development of new RPD models, in particular with the development of pressure demand SARs and an increase in the air flow rates for both PAPRs and SARs. In parallel, new asbestos-containing materials have emerged on worksites which require new techniques for their removal. This is the case, for example, with asbestos-containing plaster, the removal of which started in the early years of this millennium. In addition, since 2012, the French regulations require personal exposure to asbestos fibres to be measured using analytical transmission electron microscopy rather than optical phase-contrast microscopy, the technique used previously (Villa et al., 1994). Finally, since 2 July 2015, the occupational exposure limit (OEL) for asbestos fibres has been reduced from 100 to 10 f.L−1. The objective of this study was, therefore, to determine WPFs for the most protective and most frequently used RPD in the French asbestos removal sector. Methods Modified masks First, a prototype modified mask had to be developed to allow sampling of asbestos fibres inside the mask. Several RPD models among those used for asbestos removal tasks were selected for the study based on their frequency of use and an expected high level of protection on the basis of their APF. Thus, one positive pressure demand SAR and two models of PAPRs (PAPR-1 and PAPR-2) were selected. PAPR-1 is composed of a full face mask and a fan worn at the belt and equipped with two P3 filters. PAPR-2 is composed of a full face mask and the fan is connected to the mask and is equipped with one P3 filter. The sampling system used inside masks was developed in the laboratory in collaboration with manufacturers. The location of the sample inside mask was imposed by two constraints: • the introduction into the mask of a three-pieces open-face cassette for the sampling of asbestos fibres complying with the recommendations of the standard XP X 43-269 (AFNOR, 2012) to guarantee identical sampling conditions between the outside and the inside of the mask and • the development of a system common to all mask models. In a previous study, Chazelet (2016) demonstrated that aerosol sampling through the visor of these different models of SAR and PAPR gave a measurement result identical to that measured at the level of the mouth of the wearer. In the same paper (Chazelet, 2016), measurements of Co/Ci on wearers in laboratory showed that respiratory protection was not affected by the introduction of this sampling in the visor provided that the RPDs were used in accordance with the manufacturer’s recommendations: fit of the mask checked and supply of sufficient breathing air. In order to guarantee reliable measures, a quantitative fit test on all the wearers included in the study must be carried out. As the RPDs have undergone significant changes, its compliance with the European directive 89/686/EEC has had to be verified. Certification tests have ensured this compliance AFNOR, 1998. Equipping operators Full equipment of an operator consisted of two sampling lines for asbestos fibres, one drawing a sample from inside the respirator and one taking a sample from the person’s breathing zone outside the respirator (Fig. 1). Figure 1. View largeDownload slide Operator’s equipment. Figure 1. View largeDownload slide Operator’s equipment. The outside respirator sampling line included a personal sampling pump (Gilair Plus, Sensidyne), made decontaminable, a PVC tubing, and an antistatic open cartridge (37 mm, electrically conducting three-pieces open-face cassette, SKC Omega Specialty Division) placed in the personal breathing zone of the operator. This setup is in accordance with the recommendations set out in standard XP X 43-269 (AFNOR, 2012) and ISO 13794 (1999) for sampling of asbestos fibres, which require a sampling flow rate of 3 l min−1 and the use of an electrically conducting three-pieces open-face cassette [37 mm in diameter and fitted with a buffer and cellulose ester membrane filter (pore size: 45 µm)]. The inside respirator sampling line consisted of a second personal sampling pump (Gilair Plus, Sensidyne), made decontaminable, connected to an antistatic sampling cartridge connected in an air-tight manner to the visor of the modified respiratory mask. The same type of membrane as the first sampling line is used. Measurement campaigns in work situations Description of the worksites Asbestos removal operations in France are subjected to strict regulations, in particular in relation to the operators’ work organization. An asbestos removal operator’s working day is typically broken down into three phases of work in a confined area, with each phase lasting 90 min. These work phases are preceded by suiting-up and followed by decontamination. Between shifts, recovery periods are scheduled. These worksites were selected to allow analysis of a variety of procedures and expected dustiness levels. On each site, two to three operators fitted with the sampling systems illustrated in Fig. 1 were monitored over 3 days. In total on all worksites 25 operators participated in the study. On the first industrial site (A), the operators removed various asbestos-containing materials (surface coatings, sound-proofing, putty) from railway wagons, mainly by pneumatic scraping and sandblasting. On the industrial sites B and C, the operators carried out the removal of sprayed asbestos after watering of the latter. This removal was carried out by scraping with a spatula. The difference between these two workplaces stems from the configuration of the premises, a lot of small areas for the first and a wide area for the second. The organization of work was also different with several operators performing removal and maintenance tasks simultaneously on the first worksite and on the second worksite one operator successively performing the removal and maintenance tasks. Operations of removal of asbestos-containing plaster were followed at worksites D and F in identical configuration but with different removal technique: wet scraping for the first one and scraping; remote-controlled machinery and very high pressure for the second. At the worksite E, the operators carried out thermal insulation removal operations. In the first two workplaces where the PAPRs were evaluated (G and H), operators removed adhesive of floor coverings and tiles by sanding and grinding. On the worksite I, the asbestos-cement pipes were disassembled. Respirators tested The positive pressure demand SAR is tested on worksites A to F. The model of PAPR-1 is evaluated at worksites G and H and the model PAPR-2 at worksite I. Fit test protocol During the first day of the campaign, quantitative fit tests were performed on the operators selected for participation in the study. Operators donned their modified masks without air supply and performed a protocol consisting of eight exercises [seven exercises recommended by OSHA (2009) plus one step exercise]. Fit factors were monitored with a Portacount (TSI). This test was considered successful, and the operator was included in the measurement campaign, only if the fit factor for each of the test exercises exceeded a threshold of 2000 as recommended by HSE (2012). Description of the samplings The sampling campaign per se was performed on the 3 days following these tests. Two or three operators were equipped with the sampling lines according to the site. Samples were only collected while the operator worked in the contaminated area. The sampling duration for air inside the mask corresponded to the full time spent in the contaminated area, i.e. around 90 min per work shift. Several personal samples were collected outside the RPD in the personal breathing zone of the operator to cover the duration of the period. The sampling duration was adjusted depending on the expected dustiness level for the process being studied; it was generally between 30 min and 1 h, and thus two samples collected during each work shift. The flow rate for the two personal sampling pumps was set to 3 l min−1, as calibrated using a bubbling flow meter (Gilibrator) before use. The flow rate was recorded throughout the sampling period (Gilair + pump, Sensidyne). To ensure that sampling was performed as defined in the protocol, an INRS agent was present in the work area throughout. The numbers of inside and outside samples collected on each worksite are given in Table 1. Table 1. Numbers of asbestos samples realized during the study. Worksite ID Number of outside respirator samples Number of inside respirator samples A 60 27 B 42 21 C 36 27 D 52 27 E 47 25 F 41 18 G 36 18 H 32 18 I 17 18 Worksite ID Number of outside respirator samples Number of inside respirator samples A 60 27 B 42 21 C 36 27 D 52 27 E 47 25 F 41 18 G 36 18 H 32 18 I 17 18 View Large Table 1. Numbers of asbestos samples realized during the study. Worksite ID Number of outside respirator samples Number of inside respirator samples A 60 27 B 42 21 C 36 27 D 52 27 E 47 25 F 41 18 G 36 18 H 32 18 I 17 18 Worksite ID Number of outside respirator samples Number of inside respirator samples A 60 27 B 42 21 C 36 27 D 52 27 E 47 25 F 41 18 G 36 18 H 32 18 I 17 18 View Large End-of-sampling protocol A protocol was developed to close the cartridges as soon as the sampling pump was switched off, followed by cleaning of the casing with a wipe and placing them in a closed transport case. The transport case was double-bagged before removing it from the worksite (at the end of the campaign) and subsequently opened under a fume hood in the analysis laboratory. This protocol aimed to limit the risk of contamination of samples as far as possible once sampling cartridges were removed from the work area. Blank measurements Five measurements were performed inside respirators on worksites B, C, D, G, and I to serve as field blanks. These measurements were used to determine the limit of detection (LoD) for the analytical method. During these tests, the pumps associated with these samples were not switched on. Sample analysis For samples taken inside the mask, the analytical sensitivity was set to 1 f.L−1, which is 1/10th of the new OEL applicable in France. In order to reach this analytical sensitivity, whole sampling filters for in-mask samples were treated. For the outside sampling filters, fractions of each filter to cover the whole shift were treated together to return only a single concentration value. The filter fractions treated varied between 1 and 1/8 depending on the overall dustiness level. Grids were prepared by an indirect method (ISO 13794, 1999). As usual in France (Kauffer et al., 1996), for this preparation all or part of the sampling filter (or filters) was totally calcinated in an oxygen plasma oven so as to recover only fibres and mineral particles. Hydrochloric acid (1 N) treatment was used to eliminate any acid-soluble residues. The calcination residues were then re-dispersed and filtered on a pre-carbonated polycarbonate filter (with a pore diameter of 0.2 µm, according to the French standard NF X43-050). A second carbon deposit was then applied to trap the matter between the two layers. The filter was selectively dissolved to allow sample transfer onto microscopy grids. Fibres were then observed and analysed by analytical transmission electron microscopy, based on their morphological, chemical, and structural criteria. To comply with the regulatory exposure controls, both WHO-type asbestos fibres (length L ≥ 5 µm, diameter 0.2 µm < D < 3 µm, L/D ≥ 3) and thin fibres (L ≥5 µm, D ≤ 0.2, L/D ≥ 3) were counted. Counting was stopped when 100 asbestos fibres had been counted or when the analytical sensitivity of the technique was reached. Statistical analysis All statistical analyses were based on validated pairs of external and internal fibre counts. Fifteen couples of inside/outside measurements were excluded from the analysis for the following reasons: one measurement missing, incorrect use of the RPD (operator wearing a PAPR for which the battery indicated an inadequate charge level), or failure to respect the rules for wearing RPD (poorly shaved operator). No data were excluded based on the fibre counts recorded. In total, 132 pairs of external and internal samples were retained to determine the WPFs for SARs, and 47 pairs were retained for PAPRs. The analyses were performed in two steps: first, descriptive analyses are presented for both outside and inside fibre concentrations for the different RPDs and worksites. Secondly, the internal and external fibre counts were modelled using the standard statistical model for count data that is random effect Poisson regression fitted using maximum likelihood (Rabe-Hesketh and Skrondal, 2012). The hypotheses on which this model is based are: the Poisson distribution of asbestos counts, the log-normal distribution of exposure concentrations in workplaces, and an assumed log-normal distribution of the WPFs (see Supplementary Material, available at Annals of Occupational Hygiene online). Note that for this model zero counts are data like any other data. Results Outside samples The results from analysis of the personal samples taken outside the RPD are expressed as concentrations, Co, for the 25 different operators monitored in the nine companies. For the worksites A to F where SAR was used, the mean geometric outside concentration varies between 79 and 32011 f.L−1 (Table 2). For the G to I worksites where PAPRs were used, mean geometric outside concentration varies between 20 and 817 f.L−1 (Table 2). Table 2. Results of outside fibre concentration. Worksite Process Respirator Co (f.L−1) Geometric mean Minimum value Maximum value A Removal of surface coatings (shot peening), sound- proofing (abrasive blasting), putty (scraping) SAR 1075 25 154941 B Removal of sprayed asbestos (manual scraping) SAR 32 011 7606 84731 C Removal of sprayed asbestos (manual scraping) SAR 8511 1828 36074 D Removal of asbestos-containing plaster (pneumatic scraping) SAR 8448 414 235838 E Removal of insulation (manual scraping) SAR 79 6 508 F Removal of asbestos-containing plaster (scraping; remote-controlled machinery, very high pressure) SAR 6604 315 78076 G Removal of adhesives for floor coverings and tiles by chipping and sanding PAPR-1 817 24 8106 H Removal of adhesives for floor coverings and tiles by sanding and grinding PAPR-1 20 3 563 I Removal of asbestos-cement ducts by breaking–dismantling PAPR-2 57 6 303 Worksite Process Respirator Co (f.L−1) Geometric mean Minimum value Maximum value A Removal of surface coatings (shot peening), sound- proofing (abrasive blasting), putty (scraping) SAR 1075 25 154941 B Removal of sprayed asbestos (manual scraping) SAR 32 011 7606 84731 C Removal of sprayed asbestos (manual scraping) SAR 8511 1828 36074 D Removal of asbestos-containing plaster (pneumatic scraping) SAR 8448 414 235838 E Removal of insulation (manual scraping) SAR 79 6 508 F Removal of asbestos-containing plaster (scraping; remote-controlled machinery, very high pressure) SAR 6604 315 78076 G Removal of adhesives for floor coverings and tiles by chipping and sanding PAPR-1 817 24 8106 H Removal of adhesives for floor coverings and tiles by sanding and grinding PAPR-1 20 3 563 I Removal of asbestos-cement ducts by breaking–dismantling PAPR-2 57 6 303 View Large Table 2. Results of outside fibre concentration. Worksite Process Respirator Co (f.L−1) Geometric mean Minimum value Maximum value A Removal of surface coatings (shot peening), sound- proofing (abrasive blasting), putty (scraping) SAR 1075 25 154941 B Removal of sprayed asbestos (manual scraping) SAR 32 011 7606 84731 C Removal of sprayed asbestos (manual scraping) SAR 8511 1828 36074 D Removal of asbestos-containing plaster (pneumatic scraping) SAR 8448 414 235838 E Removal of insulation (manual scraping) SAR 79 6 508 F Removal of asbestos-containing plaster (scraping; remote-controlled machinery, very high pressure) SAR 6604 315 78076 G Removal of adhesives for floor coverings and tiles by chipping and sanding PAPR-1 817 24 8106 H Removal of adhesives for floor coverings and tiles by sanding and grinding PAPR-1 20 3 563 I Removal of asbestos-cement ducts by breaking–dismantling PAPR-2 57 6 303 Worksite Process Respirator Co (f.L−1) Geometric mean Minimum value Maximum value A Removal of surface coatings (shot peening), sound- proofing (abrasive blasting), putty (scraping) SAR 1075 25 154941 B Removal of sprayed asbestos (manual scraping) SAR 32 011 7606 84731 C Removal of sprayed asbestos (manual scraping) SAR 8511 1828 36074 D Removal of asbestos-containing plaster (pneumatic scraping) SAR 8448 414 235838 E Removal of insulation (manual scraping) SAR 79 6 508 F Removal of asbestos-containing plaster (scraping; remote-controlled machinery, very high pressure) SAR 6604 315 78076 G Removal of adhesives for floor coverings and tiles by chipping and sanding PAPR-1 817 24 8106 H Removal of adhesives for floor coverings and tiles by sanding and grinding PAPR-1 20 3 563 I Removal of asbestos-cement ducts by breaking–dismantling PAPR-2 57 6 303 View Large The outside concentration values presented in Table 2 illustrate the broad range of exposure encountered on different worksites and during different removal processes (material removed and technique used). Thus, operations to remove asbestos-containing plaster, sprayed asbestos, and removing surface coatings by abrasive blasting tend to generate very high concentrations of asbestos fibres. Maximal values of Co for worksites A, B, C, D, and F are all over 25000 f.L−1 which corresponds to the maximum fibre concentration allowed by French regulations. The outside concentrations measured on worksites G to I where PAPRs were used are much lower with geometric mean values all <1000 f.L−1. Inside samples Limit of detection No fibres were detected on any of the five field blank filters. Taking the analytical parameters of the microscope into account, these results were used to determine the LoD for the method which equals 1 f.L−1. Two classes of outside concentration can then be defined according to Vaughan and Rajan-Sithamparanadarajah (2005): low and high exposures when Co is, respectively, below or above 10 times the product between the LoD and the APF. Results of transmission electronic microscopy analysis Table 3 presents the results of transmission electron microscopy analysis in simplified terms of the presence (number of fibre Nf ≠ 0) or absence (Nf = 0) of asbestos fibres detected on the analyzed fraction of the inside sample. Table 3. Results of analytical transmission electron microscopy analysis of samples taken inside respirators. Worksite ID Respirator Total number of samples where Nf = 0 (corresponding fraction) Total number of samples where Nf = 0 and high exposure (corresponding fraction) Total number of samples where Nf = 0 and low exposure (corresponding fraction) A SAR 19 (70.4%) 3 (11.1%) 16 (59.3%) B SAR 10 (76.9%) 10 (76.9%) 0 (NA) C SAR 21 (80.8%) 19 (73.1%) 2 (7.7%) D SAR 13 (56.5%) 11 (47.8%) 2 (8.7%) E SAR 19 (76%) 0 (NA) 19 (76%) F SAR 10 (55.6%) 6 (33.3%) 4 (22.2%) G PAPR-1 13 (76.5%) 7 (41.2%) 6 (35.3%) H PAPR-1 15 (88.2%) 0 (NA) 16 (88.9%) I PAPR-2 15 (88.2%) 0 (NA) 15 (88.2%) Worksite ID Respirator Total number of samples where Nf = 0 (corresponding fraction) Total number of samples where Nf = 0 and high exposure (corresponding fraction) Total number of samples where Nf = 0 and low exposure (corresponding fraction) A SAR 19 (70.4%) 3 (11.1%) 16 (59.3%) B SAR 10 (76.9%) 10 (76.9%) 0 (NA) C SAR 21 (80.8%) 19 (73.1%) 2 (7.7%) D SAR 13 (56.5%) 11 (47.8%) 2 (8.7%) E SAR 19 (76%) 0 (NA) 19 (76%) F SAR 10 (55.6%) 6 (33.3%) 4 (22.2%) G PAPR-1 13 (76.5%) 7 (41.2%) 6 (35.3%) H PAPR-1 15 (88.2%) 0 (NA) 16 (88.9%) I PAPR-2 15 (88.2%) 0 (NA) 15 (88.2%) View Large Table 3. Results of analytical transmission electron microscopy analysis of samples taken inside respirators. Worksite ID Respirator Total number of samples where Nf = 0 (corresponding fraction) Total number of samples where Nf = 0 and high exposure (corresponding fraction) Total number of samples where Nf = 0 and low exposure (corresponding fraction) A SAR 19 (70.4%) 3 (11.1%) 16 (59.3%) B SAR 10 (76.9%) 10 (76.9%) 0 (NA) C SAR 21 (80.8%) 19 (73.1%) 2 (7.7%) D SAR 13 (56.5%) 11 (47.8%) 2 (8.7%) E SAR 19 (76%) 0 (NA) 19 (76%) F SAR 10 (55.6%) 6 (33.3%) 4 (22.2%) G PAPR-1 13 (76.5%) 7 (41.2%) 6 (35.3%) H PAPR-1 15 (88.2%) 0 (NA) 16 (88.9%) I PAPR-2 15 (88.2%) 0 (NA) 15 (88.2%) Worksite ID Respirator Total number of samples where Nf = 0 (corresponding fraction) Total number of samples where Nf = 0 and high exposure (corresponding fraction) Total number of samples where Nf = 0 and low exposure (corresponding fraction) A SAR 19 (70.4%) 3 (11.1%) 16 (59.3%) B SAR 10 (76.9%) 10 (76.9%) 0 (NA) C SAR 21 (80.8%) 19 (73.1%) 2 (7.7%) D SAR 13 (56.5%) 11 (47.8%) 2 (8.7%) E SAR 19 (76%) 0 (NA) 19 (76%) F SAR 10 (55.6%) 6 (33.3%) 4 (22.2%) G PAPR-1 13 (76.5%) 7 (41.2%) 6 (35.3%) H PAPR-1 15 (88.2%) 0 (NA) 16 (88.9%) I PAPR-2 15 (88.2%) 0 (NA) 15 (88.2%) View Large Results show that in the case of the asbestos-containing plaster removal process (worksites D and F), direct exposures to asbestos fibres inside respirators are the most numerous with a lowest proportion of samples with no fibre detected in the mask. Table 3 shows that even for low exposure, some fibres were detected on the inside mask samples. Moreover, it was observed that fibres can penetrate inside respirators whatever the worksite, the task, the subject, or the work period. Distribution of WPFs Plotting the concentration measured inside the mask, Ci, against the concentration measured outside the mask, Co, shows that no obvious relationship links both concentrations (Fig. 2). Thus, the WPF is very variable from one worksite to another and even within the same site. Figure 2. View largeDownload slide Raw data from analytical transmission electron microscopy analyses of concentrations measured outside (Co) and inside (Ci) RPD masks (179 pairs). Figure 2. View largeDownload slide Raw data from analytical transmission electron microscopy analyses of concentrations measured outside (Co) and inside (Ci) RPD masks (179 pairs). Results reveal first that in 135 cases of 179 presented on Fig. 2, no asbestos fibres are detected on the inside sampling. Secondly in 7 cases of 179, the penetration of asbestos fibres inside the mask is such that Ci exceeded the 10 f.L−1 limit value during the work shift. These high values of Ci are measured for outside dust concentrations >2000 f.L−1. Moreover, all of these high asbestos exposures inside RPD involved worksites where operators were equipped with SARs and engaged in removing asbestos-containing plaster. Table 1 shows that very high level of Co values were obtained during this removal process. In the case of PAPRs, which are used on much less dusty sites than SARs, Ci values never exceeded 10 f.L−1. A comparison of WPFs obtained for each model of PAPR under similar outside dust concentration conditions (workplaces H and I) led to the conclusion that no significant difference between the two distributions was found. All the measurements carried out on the two models of PAPR have, therefore, been treated jointly to give only one distribution of WPF for this type of RPD. Statistical analysis of the data described in the ‘Methods’ section provided a distribution of the WPFs for the two types of RPD studied. The fifth percentile of these distributions equal 236 and 104, respectively, for SAR and PAPR (Table 4). Table 4. Characteristics of the distribution of WPF. RPD type SAR PAPR Number of values analysed 132 47 Geometric mean (95% confidence interval) 8982 (5411; 14909) 886 (420; 1871) Geometric standard deviation (95% confidence interval) 9.15 (6.65; 13.3) 3.69 (2.62; 5.86) Fifth percentile for the distribution 236 104 RPD type SAR PAPR Number of values analysed 132 47 Geometric mean (95% confidence interval) 8982 (5411; 14909) 886 (420; 1871) Geometric standard deviation (95% confidence interval) 9.15 (6.65; 13.3) 3.69 (2.62; 5.86) Fifth percentile for the distribution 236 104 View Large Table 4. Characteristics of the distribution of WPF. RPD type SAR PAPR Number of values analysed 132 47 Geometric mean (95% confidence interval) 8982 (5411; 14909) 886 (420; 1871) Geometric standard deviation (95% confidence interval) 9.15 (6.65; 13.3) 3.69 (2.62; 5.86) Fifth percentile for the distribution 236 104 RPD type SAR PAPR Number of values analysed 132 47 Geometric mean (95% confidence interval) 8982 (5411; 14909) 886 (420; 1871) Geometric standard deviation (95% confidence interval) 9.15 (6.65; 13.3) 3.69 (2.62; 5.86) Fifth percentile for the distribution 236 104 View Large Discussion The objective of this study was to determine the WPFs for two types of RPD used on asbestos removal worksites: pressure demand SAR and PAPR. Measurements were performed during asbestos removal, examining the processes likely to lead to the highest emissions of asbestos fibres. Worksites where sprayed asbestos removal and asbestos-containing plaster removal were performed were then selected to evaluate the WPF of the SAR. Worksites of removal of tile and floor tile adhesives were selected to evaluate WPFs of the PAPRs. Inside and outside respirator samples were carried out on 25 operators on 9 worksites and were analysed by analytical transmission electron microscopy to obtain WPFs. Inside respirator samples do not contain asbestos fibres in 75% of cases on all sites. But in 4% of cases, the concentration of asbestos fibres inside the masks is higher than the OEL value of 10 f.L−1. These high values of Ci were all measured inside air-supplied respirators worn by operators performing asbestos-containing plaster removal tasks and at outside asbestos concentrations >2500 f.L−1. For the PAPRs, the two models studied have similar WPF distributions. A statistical treatment of the Co and Ci data allowed the determination of two WPF distributions for each type of respirator. The fifth percentiles of these distributions were equal to 104 for PAPR and to 236 for SAR. Few studies exist in the literature to report WPF of SARs or PAPRs. The work of Villa et al. (1994) and Howie et al. (1996) are the only ones that provide WPF values for different models of PAPRs measured during asbestos removal operations at different sites. Villa et al. (1994) describes highly variable WPFs between 140 and >10000 for different operators performing the same removal task. The work of Howie et al. (1996), which has many more measurements than those of Villa et al. (1994), shows that out of 179 Co/Ci ratios, 57 are unquantifiable due to a Ci below the LoD. They also show that in 90% of the cases, the mask internal concentrations are below the limit value. In addition, this article concludes that there is no significant difference between the WPF distributions obtained on the different PAPR models. Our findings are consistent with this work, which at that time led to a proposal for an APF for PAPRs of 40. Campaigns to measure protective factors on PAPRs against other pollutants (Myers, 1983; Lenhart, 1984; Myers et al., 1984) came to the same conclusions from a WPF distribution of log-type normal and with a fifth percentile value <50. To explain the WPF differences measured during the Myers campaign, the differences in carrier behaviour and facial leakage were discussed. Two measurement campaigns focused on the performance of hood style SARs (Nelson et al., 2001; Janssen et al., 2008), both of whom concluded high WPF for this type of APR during sanding and painting operations. Pollutant detection was done by detection of chemical species including strontium in both articles. Nelson et al. (2001), who reports the most data, indicates that in sanding operations 50% of intakes are below the detection limit of strontium compared with 100% in paint operations. Their authors offer APFs over 1000 for this APR. There is no work on the WPF’s pressure demand SAR with full facemask. The low values of WPF obtained, in comparison with those of Nelson and Janssen, can be explained by the use of this SAR model in more restrictive conditions: site configuration (multi-storey work on sites B, D, E, and F), use of scaffolding (sites B, C, D, F). These posture constraints and these efforts can lead to a penetration of fibres in the mask, as evoked by Howie (1996). When Ci is less than the detection limit of the analytical method adapted to the pollutant, the works cited above all propose a calculation of WPF equal to Co divided by the LoD, except Howie which considers that, in this case, WPF > nominal protection factor (NPF). One major feature of the data of the present paper is that a majority of the within mask fibre counts were zero. Thus, no value for a WPF can be obtained from any such data point (pair of counts). However, excluding these data was not an option, as this would have led to a serious bias by selecting only the worst situations in terms of presence of within-mask fibres. In order to cope with these data, we relied on a statistical model fitted using maximum likelihood that allows incorporating all the data including those with zero within-mask fibre counts. This model is based on three hypotheses which are given in Supplementary Material (available at Annals of Occupational Hygiene online) and are standard in the field of industrial hygiene. The numerical values of the fifth percentile of each WPF distributions obtained (236 for SAR and 104 for PAPR) are of the same order of magnitude as those used up to now in the French regulations (250 for SAR and 60 for PAPR) (Villa et al., 1994). However, the level of operator training has increased considerably since the 1990s, as well as the level of collective protective measures implemented on worksites. Moreover, the RPDs assessed in this study present more protective characteristics than the RPDs assessed in the 1990s. As a matter of fact the supplied-air RPD used was equipped with a pressure demand valve which did not exist at that time, and the PAPRs delivered flow rates >160 l min−1. These factors should have led to higher WPFs than those measured in the 1990s. But, today, new asbestos-containing materials, such as asbestos-containing plaster, are being removed, in some cases through the application of new, more aggressive procedures such as very high pressure. These new asbestos removal situations were linked to the highest risks of exposure to asbestos fibres. This result confirms those presented by Radauceanu et al. (2008) in the form of a task-exposure matrix. These results remind us of the limitations of using personal protective equipment as these devices alone may not guarantee optimal worker protection. Development of protective measures should today focus on the reduction of dustiness in the work area as far as possible, though the development of aspiration at source, work on humidified materials, or even mechanization when these measures prove insufficient. Supplementary Data Supplementary data are available at Annals of Work Exposures and Health online. Funding Funding for this project was entirely based on the core funding of the Institut National de Recherche et Sécurité (INRS) which is the sole employer of the authors. Acknowledgements The authors would like to thank the INRS teams who participated in the measurement campaigns and the one who carried out all the microscopic analyzes. The authors declare no conflict of interest relating to the material presented in this article. Its contents, including any opinions and/or conclusions expressed, are solely those of the authors. References Association Française de NORmalisation (AFNOR) . ( 1998 ) NF EN 136:1998. Respiratory protective devices—full face masks—requirements, testing, marking . La Plaine Saint-Denis, France : AFNOR . Association Française de NORmalisation (AFNOR) . ( 2005 ) NF EN ISO 8996:2005. Ergonomics of the thermal environment—determination of metabolic rate . La Plaine Saint-Denis, France : AFNOR . Association Française de NORmalisation (AFNOR) . ( 2012 ) XP X 43-269:2012. Air quality—workplace atmospheres—sampling on membrane filters for the determination of the fibre number concentration by microscopic techniques: phase contrast optical microscopy, scanning electron microscopy analysis and transmission electron microscopy analysis—counting by phase contrast optical microscopy . La Plaine Saint-Denis, France : AFNOR . Chazelet S , Silvente E , Eypert-Blaison C et al. ( 2016 ) Asbestos fibre sampling through a facepiece visor to measure Workplace Protection Factor . J Int Soc Respir Prot ; 33 : 21 – 35 . Health and Safety Executive (HSE) . ( 2012 ) Fit testing of respiratory protective equipment facepieces. HSE OC 282/28 . Internal report to Health and Safety Executive (HSE) . Howie RM , Johnstone JBG , Weston P et al. ( 1996 ) Workplace effectiveness of respiratory protective equipment for asbestos removal work. HSE Contract Research Report n° 112/1996 . Sheffield: Health and Safety Executive . International Organization for Standardization (ISO) . ( 1999 ) ISO 13794:1999. Ambient air. Determination of asbestos fibres. Indirect-transfer transmission electron microscopy procedure . Geneva, Switzerland : International Standardization Organisation . Janssen L , Bidwell J , Cuta K et al. ( 2008 ) Workplace performance of a hood-style supplied-air respirator . J Occup Environ Hyg ; 5 : 438 – 43 . Google Scholar CrossRef Search ADS PubMed Kauffer E , Billon-Galland MA , Vigneron JC et al. ( 1996 ) Effects of preparation methods of airborne concentrations of asbestos fibres by transmission electron microscopy . Ann Occup Hyg ; 40 : 321 – 30 . Google Scholar CrossRef Search ADS Lenhart SW , Campbell DL . ( 1984 ). Assigned protection factors for two respirator types based upon workplace performance testing . Ann Occup Hyg ; 28 : 173 – 182 . doi: 10.1093/annhyg/28.2.173 Google Scholar PubMed Myers W , Peach MJ . ( 1983 ). Performance measurements on a powered air purifying respirator made during field use in silica ragging operation . Ann Occup Hyg ; 27 : 251 – 259 . doi: 10.1093/annhyg/27.3.251 Google Scholar PubMed Myers WR , Peach MJ 3rd , Allender J . ( 1984 ) Workplace Protection Factor measurements on powered air-purifying respirators at a secondary lead smelter–test protocol . Am Ind Hyg Assoc J ; 45 : 236 – 41 . Google Scholar CrossRef Search ADS PubMed Nelson TJ , Wheeler TH , Mustard TS . ( 2001 ) Workplace Protection Factors—supplied air hood . Am Ind Hyg Assoc ; 62 : 96 – 9 . Occupational Safety and Health Administration (OSHA), US Department of Labor . ( 2009 ) Assigned Protection Factors for the revised respiratory protection standard, OSHA 3352-02 . Washington, DC : OSHA Publications Office, U.S. Department of Labor . Rabe-Hesketh S , Skrondal A . ( 2012 ) Multilevel and longitudinal modeling using stata – volume ii: categorical responses, counts and survival . 3rd edn . College Station , TX : Stata Press Publication . Radauceanu A , Chouanière D , Wild P et al. ( 2008 ) Assessment of asbestos exposure in asbestos removal operations: proposal of a task exposure matrix . Archives des maladies professionnelles et de l’Environnement ; 69 : 586 – 92 . Google Scholar CrossRef Search ADS Vaughan N , Rajan-Sithamparanadarajah B . ( 2005 ) Meaningful Workplace Protection Factor measurement: experimental protocols and data treatment . Ann Occup Hyg ; 49 : 549 – 61 . Google Scholar PubMed Villa M , Hubert G , Lima S et al. ( 1994 ) Occupational exposure during asbestos removal operations . J Int Soc Respir Prot ; 12 : 7 – 14 . © The Author(s) 2018. Published by Oxford University Press on behalf of the British Occupational Hygiene Society. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Annals of Work Exposures and Health (formerly Annals Of Occupational Hygiene)Oxford University Press

Published: Mar 27, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off