Annals of Occupational Hygiene Advance Access originally published online on October 31, 2006
Annals of Occupational Hygiene 2007 51(2):207-218; doi:10.1093/annhyg/mel074
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Occupational Exposure During Application and Removal of Antifouling Paints
TNO Chemistry, Department of Food and Chemical Risk Analysis, Chemical Exposure assessment PO Box 360, 3700 AJ Zeist, The Netherlands
*Author to whom correspondence should be addressed. Tel: +31-30-6944913; fax: +31-30-6944070; e-mail: joop.vanhemmen{at}tno.nl
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Exposure data on biocides are relatively rare in published literature, especially for secondary exposure. This is also the case for antifouling exposure. Therefore, a field study was carried out measuring exposure to antifouling paints. Both primary exposure (rolling and spraying) and secondary exposure (during sand blasting) were studied. Exposure during rolling was measured in boatyards where paints containing dichlofluanid (DCF) were applied. Spraying was measured in dockyards (larger than boatyards) where paints containing copper were applied. Furthermore, during sand blasting the removal of old paint layers containing copper was measured. A total of 54 datasets was collected, both for inhalation and dermal exposure data. For paint and stripped paint bulk analyses were performed. The following values are all arithmetic means of the datasets. Inhalation of copper amounted to 3 mg m–3 during spraying and to 0.8 mg m–3 during sand blasting. Potential body exposure loading amounted to 272 mg h–1 copper during spraying and 33 mg h–1 during sand blasting. For dichlofluanid the inhalation exposure loading was 0.14 mg m–3 during rolling, whereas the potential body exposure loading was 267 mg h–1 and potential hand exposure loading 277 mg h–1. The results for primary exposure compare well to the very few public data available. For the secondary exposure (sand blasting) no comparable data were available. The present study shows that the exposure loading should be considered more extensively, including applicable protective gear. In this light the findings for the potmen during sand blasting suggest that personal protective equipment should be (re)considered carefully.
Keywords: antifouling paint • biocides • copper • dichlofluanid • rolling • spraying • sand blasting • secondary exposure
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
According to the European Biocides Directive (Directive 98/8/EC) it is obligatory to conduct an occupational risk assessment for a biocidal product before it is allowed on the European market. To be able to assess the risks for workers, exposure data are needed. However, good quality worker exposure data for biocides are generally scarce. As Cherrie observed (Cherrie, 2003) new exposure measurements are becoming more and more rare. Post-application exposure data are even scarcer. This makes it very difficult to adequately assess the potential for exposure and thereby health risks for different use scenarios (Van Hemmen et al., 2003). Measured exposure data can be used to derive surrogate exposure values and to develop and improve predictive exposure models, allowing more accurate assessment of health risks associated with the use of biocides.
Biocides are, amongst others, used in antifouling paints to prevent growth of organisms to the underwater part of boats and ships (fouling). On average, navy sea ships enter a dock for major repairs every 3 years (personal communication by Navy dockyard personnel). Antifouling paint is applied to sea-faring ships by air-less spraying because of the large surface area that has to be treated. For sea-faring ships copper (oxide) has long been a commonly used active substance of antifouling paints in The Netherlands. This compound has been used worldwide since the pre-1960s (Omae, 2003). Before applying a new layer, old layers of antifouling paint on sea ships are removed by sand (grit) blasting. Exposure to biocides is thus also possible during removal of biocide-containing paint. In the present study exposure loading of workers to copper was measured during the application of antifouling by air-less spraying and during the removal of antifouling paint by sand blasting. Application of antifouling to the smaller ships and boats that navigate inland waterways is usually performed with a paint roller. Because of its ecotoxicity, the use of copper oxide-based antifouling paints for ships or boats that navigate inland water ways may be banned in the Netherlands. For these vessels other antifouling paints are therefore used more frequently. Exposure to the active substance dichlofluanid is assessed in the present study during paint rolling.
The main reason for doing these studies is that in the public domain there are very few studies available that cover the actual application of paint to ships and none that cover the removal of paint from ships.
The study was carried out with the following objectives in mind.
- To fill data gaps and to gather insight into the distribution of exposure loading over the body of workers under different exposure scenarios, and to study the relationships between potential and actual dermal exposures.
- Furthermore, the study results can be used directly as surrogate exposure values for risk assessment purposes, and for validation of predictive exposure models such as BEAT (TNsG, 2002).
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Occupational settings
All participating workers were experienced professionals who performed their work as normal, and volunteered to participate in the study. Exposure measurements during rolling dichlofluanid (DCF) containing antifouling paints were carried out from September through December 2002 in seven boatyards in The Netherlands. Exposure measurements of the active substance copper, during spray painting and sand blasting were carried out from February through July 2003 in three dockyards in The Netherlands.
Dichlofluanid
The boatyards were large halls (estimated average length of 25 m), where one or more boats were being built or maintained. Boatyards are in general intended for boats of the size of pleasure yachts, whereas dockyards are intended for professional (seafaring) boats (larger in size, professional use only). The boatyards were selected by obtaining phone numbers from the Dutch yellow pages. A short phone questionnaire was carried out to establish whether DCF containing paint was used for antifouling treatment.
For the rolling scenario both tasks—mixing/loading and application—were included in the same measurement (continuous monitoring on unchanged dosimeters). Potential and actual dermal exposure loading, hand exposure loading and inhalation exposure loading were measured for thirteen workers. One worker was sampled three times on separate days, so a total of 15 data points (replicates) were collected. The different boatyards were in general enclosed, or had partial walls (open or half-open setting).
Copper
Exposure loading during spray painting and sand blasting (exposure to copper) was measured in dockyards that varied in size from 75 to 200 m in length. To ensure that exposure to copper-containing antifouling paint during the sand blasting scenario was monitored, only dockyards that kept a log of the paints used on the ships being sand blasted were selected.
During the study usually one large boat or two small vessels were present in the dockyard. All entrances to the dockyard were closed, fresh air was supplied and contaminated air was simultaneously removed with mechanical ventilation (enclosed situations with restricted access). The person who was paint spraying (paint sprayer) and sand blasting (sand blaster) were assisted by a worker on the floor who mainly took care of the paint or grit supply (the potman). Potential and actual dermal exposure loading, including hand exposure, and inhalation exposure were measured 12 times for paint spraying and 12 times for sand blasting. Potential dermal exposure loading, hand exposure and inhalation exposure were measured 11 times for paint filling and three times for grit filling. The 14 workers who participated in this part of the study (exposure to copper) were sampled repeatedly. The workers were sampled for different tasks, e.g. 2 times for paint filling or 4 times for sand blasting. Some were sampled for both tasks (carried out at separate occasions). The paint sprayers often worked in pairs, e.g. one worked the upper part while the other worked on the bottom side, or one on port side and the other on starboard side. This was also the case for the sand blast scenario.
Exposure scenarios
Rolling
For rolling, the application, including mixing/loading, was as follows: the worker opened a paint can, stirred the paint with a wooden or metal ‘stick’ of about 30 cm, and poured (a part of) the content into a paint tray or square bucket with a grate. The paint roller was moved back and forward several times until the roller was saturated with paint. Next, the paint was applied to the hull of the boat by moving the roller over the surface of the boat. Depending on the type of boat (and scaffold) the worker worked partly underneath the boat, often kneeling down. Usually, workers wear safety shoes, an overall (Tyvek® or cotton) or sometimes only normal clothing (e.g. trousers and a jumper), nitril rubber gloves and a variable degree of respiratory protection during mixing/loading and application.
Spraying
The worker applied the antifouling paint by means of air less spraying. To be able to reach the higher parts of the ship, the worker worked from an automated tower wagon that he manoeuvred himself. A spray-gun or lance was used with a type 23.40 or 26.40 (26 µm, 40 cm) nozzle, respectively. One or two layers of the antifouling paint were applied.
During spray painting especially during application to the stern of the ship (e.g. blades of the propeller) the worker worked in uncomfortable positions. Overhead spraying occurred regularly when spraying the lower sides of the boat from a standing position on the floor (spraying overhead). The following protective equipment was normally used by the paint sprayer: half (n = 3) or full face respirator (n = 9), nitrile gloves, Tyvek® coverall or cotton overall with Tyvek® hood, safety shoes and a safety helmet.
The paint sprayer (on the tower wagon) was assisted by a worker on the floor. These assistants may have also been exposed to the paint (mist), because they stayed in the vicinity of the paint sprayer and were positioned between source and extraction ventilation. Furthermore, exposure was possible during mixing, loading of paint, and during cleaning of (spray) equipment. The paint fillers/assistants usually wore a half face respirator, gloves, a Tyvek® coverall or cotton overall with Tyvek® hood, safety shoes (except one person) and a safety helmet.
Sand blasting
To remove old paint layers (containing copper oxide) from the underwater part of the hull, the sand blasters used a mixture of water and grit under high pressure. The water and grit mixture was supplied by a ‘mixing kettle’ that was regularly filled with grit by the assistant of the sand blaster by emptying paper bags containing grit (25 kg) into this kettle. The water: grit ratio was adjusted by the assistant/filler, based on the experience of and directions received from the sand blaster. The length of the spray lance was generally 1 m and the diameter of the spray nozzle 1 cm.
Sand blasting generates a lot of dust and noise. The workers who sand blasted wore protective, waterproof overalls, an Airstream helmet with rubber flaps that covered a large part of their upper body and strong protective gloves. Especially at the back of the ship (e.g. sand blasting blades of screw) the worker worked in uncomfortable (e.g. kneeling, crouching) positions and sometimes sprayed over head.
During sand blasting the grit filler/assistant stays close (within 10 m) to the sand blaster and may therefore also be exposed. The assistant was responsible for monitoring the grit and water supply to the kettle and air to the tower wagon. The fillers usually wore a Tyvek® coverall, a plastic helmet, a white mouth mask and rubber coated cotton protective gloves.
Table 1 gives an overview of the study details.
|
Sampling methods
Inhalation exposure
Inhalation exposure loading during rolling, spray painting and paint filling of DCF containing paints was evaluated using a personal air sampling pump (Gillian) equipped with a sampling head, containing two (front and back) 25 mm glass fibre filters (Millipore). During paint rolling splashing was considered a possible process for contaminating the inhalation sampling (intended for measuring aerosols and not splashes). In the literature, it was observed that the IOM (Institute of Occupational Medicine) sampling head tends to collect large projectile particles, which would normally be too large to be aspirated, compared to samplers with smaller orifices, such as the 7-hole sampler (Vaughan et al., 1990). Because splashing was expected during rolling, the rollers were equipped with a 7-hole sampler (SHS) next to an IOM sampling head. Because the generation of splashes was estimated to be less likely during spraying than during rolling (as confirmed in actual practice), only an IOM sampling head was used during spraying and filling of paint. The sampling heads were attached to the outside of the Tyvek® coverall, in the breathing zone of the worker. The recovery was in all cases better than 70%.
Inhalation exposure loading during sand blasting and grit filling was evaluated using a personal air sampling pump (Gillian) equipped with a PAS6 sampling head for weighing total inhalable dust, containing two (front and back) 25 mm glass fibre filter (Millipore).
The filters were weighed before and after sampling to determine total inhalable dust. Weighing was performed with a calibrated balance with a reproducibility of ±0.1 mg (Mettler AT 200) according to a standard procedure, i.e. after at least 24 h of acclimatization. Subsequently the filters were analysed for copper or dichlofluanid.
The flow of all pumps was controlled at 2 l min–1 with a calibrated Rotameter. The average of the flow at the start and the end of the sampling in combination with the sample monitoring time was used for calculations of the amount of air sampled. If applicable, pumps were disconnected during breaks.
Dermal exposure
Potential exposure was defined as contamination from all dosimeters (inner and outer).
A whole body method (OECD, 1997) was used to determine potential dermal exposure. The workers wore a Tyvek® coverall with a hood and cotton monitoring socks as outer dosimeter during mixing, loading and rolling, spray painting, sand blasting and filling paint or grit. The paint rollers wore the Tyvek® coverall instead of their normal work clothing (i.e. directly over cotton monitoring dosimeters for collecting actual dermal exposure data). Sand blasters, sprayers, paint- and grit-fillers wore a cotton overall under their Tyvek® coverall, the cotton overall was not analysed. The clean cotton overall was worn instead of the normal clothing, which the worker usually wore under their protective clothing. Sand blasters' protective gear included a special protective self-contained hood. Consequently, the patch method as described by Soutar et al. (2000) was employed to obtain outer dosimeter exposure loading for the ‘head’ body part during sand blasting. The hood of the paint fillers was excluded from the analysis, because the hood was not always worn. It was put off now and then for reasons of convenience for the paint fillers. The OECD (1997) surface areas were used for extrapolation.
For all exposure scenarios, monitoring socks were worn instead of the applicator's own socks. After application the coverall was cut into the following parts: lower legs (left and right), upper legs (left and right), forearms (left and right), upper arms (left and right), torso front, torso back and hood. The parts were analysed separately, combining left and right parts to one sample.
Actual exposure loading was defined as contamination from inner dosimeters only.
A whole body method was also used to determine actual dermal exposure of the rollers, spray painters and sand blasters. The fillers often worked a certain distance from the painters/blasters. The filling-point was stationary and the painters/blasters worked their way along the ship. Due to the assumed effectiveness of the ventilation system along the ship, exposure was assumed to be relatively low. Therefore, no actual dermal exposure loading was measured for the grit and paint fillers, except for the hands, since no significant exposure was expected on the body.
A cotton undergarment (long-sleeved cotton T-shirt and cotton long johns) was worn under the Tyvek® coverall (rollers) or cotton overall (sprayers, sand blasters and paint or grit-fillers). After application, the cotton underwear was cut into the following parts: lower legs (left and right), upper legs (left and right), forearms (left and right), upper arms (left and right), torso front and torso back. As was done for potential dermal exposure loading, the 10 parts were combined to 6 samples that were analysed separately.
To assess potential and actual hand exposure loading, cotton monitoring gloves were used. Rollers, paint sprayers, paint fillers and grit fillers wore one pair under (actual), and one pair over (potential) nitrile rubber gloves. The sand blasters only wore one pair of cotton monitoring gloves, under newly provided strong protective gloves. As a measure for potential hand exposure loading the protective gloves were analysed.
As described above, the general definition of potential dermal body exposure loading is the sum of the exposure values obtained on the outer and inner clothing. Actual body exposure loading is the combined exposure found on the inner clothing only (similarly for hands with inner and outer gloves). However, due to the clothing configuration for the different scenarios the calculation of the potential and actual dermal exposure varied slightly per scenario depending on whether workers always wore the hood of the Tyvek® coverall, if a helmet was worn over the hood or if other protective clothing—e.g. the special protective hood for the sand blasters—was used over the coveralls. Table 2 gives the specific definitions of potential and actual exposure loading for the different scenarios.
|
Product samples and quality control samples
Dichlofluanid (DCF)
A sample of paint product (20–50 ml) was taken on each sampling day.
For each sampling day blank samples were taken: a blank filter, a part of the Tyvek® coverall, hood, socks and a part of the cotton underwear.
Field control samples were analysed for determining the stability of the samples in the field. This was done for each sampling day and for each type of monitoring matrix (GF-filter, Tyvek®, gloves, socks). A known amount of paint was applied to the matrices at the laboratory and these field control samples were then taken to the site and were treated the same way as the study samples. Exposure loading data were expressed as amount of active substance and in amount of paint, using the amount of DCF in paint as observed in bulk analysis.
Copper
A sample of paint product (20–50 ml) was taken on each sampling day.
For each sampling day blank samples were taken: a blank filter, a part of the Tyvek® coverall, hood, socks and a part of the cotton underwear. As copper is known to be very stable, no field control samples were taken.
On the days that sand blasting was being examined, samples of clean grit, dirty grit and water used for the grit/water mixture were taken. Exposure loading data were expressed as amount of active substance and in amount of paint, using the amount of copper in paint as observed in bulk analysis.
Sample extraction and analysis
Dichlofluanid
The chemical analytical method for the determination of dichlofluanid on the different matrices was validated with respect to linearity, specificity, within day and between day repeatability, extraction efficiency, recovery, air aspiration tests, stability and limit of detection and limit of quantification. For the validation of the air sampling method GF filters, spiked with DCF-paint, were sucked through with air (2 l min–1, temp. 19–23°C, RH 30–50%) for up to 3 h. The front and back section filter were extracted separately. The average recovery found was 104% (after 3 h; n = 3). No breakthrough was observed.
Dichlofluanid was extracted from the different matrices (dosimeters) into acidified methanol by means of treatment in an ultrasonic bath (10 min), and 30 min of shaking at 300 strikes per minute in a shaking table. The matrix extraction solution was then purified by n-hexane extraction prior to analysis. The hexane phase was subsequently dried under nitrogen. The dry residue in the eluents was then analysed by C18-HPLC with UV detection at 200 nm.
Recovery tests were performed for all types of matrix body parts at different concentration levels by means of spikes with standard solution and paint. Linearity was based on R-squared >0.9995 (till 45 000 µg l–1). Recovery from spiked samples ranged from 79 to 116%, no adjustments were made to the analytical results. Within-day and between-days repeatability and stability were tested. The variation was <10% for all matrices at all concentrations. Stability was tested at lab temperature for the dried matrix, stable for 7 days and in extraction liquid, stable for 56 days.
All samples were extracted from the dried matrix within a week and purified just before analysis. The limits of detection and limits of quantification for the different matrices are provided in Table 3.
|
Copper
The chemical analytical method for the determination of copper on the different matrices was validated with respect to linearity, repeatability, limit of detection and limit of quantification. No air aspiration tests were performed.
An amount of diluted nitric acid (1 mol l–1) was added to the samples, so that the total sample was covered with nitric acid. The bottles were heated at 60°C during one hour and after shaking for 10 min, copper was determined using inductively coupled plasma atomic emission spectrometry (ICP-AES) at a wavelength of 324 nm. The calibration samples were prepared by adding different concentrations of a commercial copper standard solution of 1000 mg l–1 to 100 ml with diluted nitric acid (1 mol l–1). To analyse the samples, an ICP-AES type IRIS Intrepid with auto sampler was used.
The method was validated with respect to the linearity (R-squared 0.999), recovery from spiked samples (range 80–120%), within-day and between-days repeatability, limit of detection and limit of quantification.
Statistical analysis
Descriptive statistics of inhalation and dermal exposure, i.e. arithmetic mean (AM) and geometric mean (GM), geometric standard deviations (GSD), range and 90th percentile were determined.
Fifty-three complete sets of measurements were collected for five different activities (Table 1). Observations below the LOQ were assigned values of 1/2 LOQ and values below LOD were assigned the value of the LOD. All statistical analyses were performed using SAS Statistical Software (SAS Institute, Cary, NC, USA). PROC MEANS was used to generate simple descriptive statistics for each scenario (AM, standard deviation and range). The Univariate procedure was used to investigate the normality of the log transformed data and descriptive statistics of the log transformed data were calculated. Pearson's correlation coefficients generated using the CORR procedure were used to investigate the following:
- The relationships between exposure on the outer clothing per body part and the exposure on the corresponding inner clothing.
- The relationships between potential exposure per body part and the inhalation exposure.
- The relationships between potential exposure per body part and the potential hand exposure.
- The relationship between the inhalation exposure estimated using the IOM inhalable dust sampler and that estimated using the 7-hole sampler.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Potential and actual dermal exposure loading was expressed per hour to enable cross scenario comparison. For inhalation the mg m–3 metric was used, taking 1.25 m3 h–1 as a default breathing rate. The results are presented in Tables 4–7. No corrections for percentage recovery were made.
|
Dichlofluanid
In 2 (12.5%) backup filter samples for the IOM and in 8 (50%) back filter samples for the 7-hole sampler a concentration above the LOD was found (Table 4). For dermal body exposure loading, 44% of the actual body exposure loading samples were below LOQ (Tables 5 and 6), while for actual hand exposure loading 50% of the samples was below LOQ (Tables 7 and 8).
|
|
|
|
For rolling, the paint used contained an average (AM) of 9.3% (w/w) dichlofluanid (SD 0.2 %) (average of 12 product samples). Duration of painting varied between 18 to 124 minutes (AM 55, SD 26). Mixing and loading only took a few minutes. The average amount of paint used was 4.9 kg (AM, SD 1.9).
Copper
The back filters of the IOM sampler were above the LOD in 12, 6 and 8 cases for sand blasting, spraying and paint filling, respectively. For sand blasting, 6% of the actual body exposure loading samples was below LOQ. For spraying, 2% of the actual body exposure loading samples was below LOQ and one (4%) potential hand exposure loading sample was below LOQ. For paint filling, all samples were above LOQ. Contamination beneath the coverall (actual body exposure loading) was not sampled for this scenario. Contamination beneath protective gloves was sampled, with antifouling observed beneath the protective gloves in almost 100% of the surveys.
The paint used for spraying contained an average of 364 g copper kg–1 (36% w/w) (range 311–393 g kg–1). Duration of painting varied between 66 and 151 min (AM 102 min). The average amount of paint used was 218 kg (range 108–347 kg, n = 11). This is about 120 l paint.
The duration for sand blasting varied from 71 to 196 min (AM 117, SD 44), the area treated varied from 8 to 90 m2 (AM 47, SD 34). The percentage copper present in the used grit, contaminated with old paint layers, varied from 0.46 to 5.3 with an average of 2.28 g kg–1 grit. The old paint layers collected contained 0.08–221 with an average of 108 g kg–1 of copper. (Clean grit contained 0.012–0.028 g copper kg–1 product. Water contained 0.1–0.8 mg copper L–1.)
Actual versus potential exposure loading
In Tables 9 and 10 the contribution of the actual exposure loading to potential exposure loading is presented. The data cover average values, since it was considered that for a more in depth analysis of body parts, the number of replicates was relatively small.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
In the discussion the following items will be covered: the routes of exposure, the patterns of exposure, the penetration through protective clothing (comparison of actual and potential dermal exposure loading). Additional attention will be given to the three main scenarios: rolling, spraying and sand blasting.
Routes of exposure
Exposure via the skin and inhalation are important routes of exposure for the different scenarios considered in the present study. The relationships (Pearson's correlation coefficient) between potential dermal hand exposure and potential dermal exposure at the individual body sites and the relationship between inhalation exposure and the potential dermal exposure at the individual body sites have been explored. Based upon the routes of exposure as defined by Schneider et al. (1999) the results of the present study were evaluated. The results indicate that inhalation exposure loading correlates reasonably well with dermal exposure loading for the scenarios sand blasting and spraying, but not so much for rolling and hardly for paint filling (Table 11). Furthermore, from observation of the study participants during sand blasting, it appears that transfer from contaminated surfaces is very likely to occur. Crouching, sitting on the knees and handling of (contaminated) hoses were observed often for sand blasting, to some extent for spraying and partly for rolling as well. The data in Table 11 also seem to support the observation that appreciable transfer from contaminated surfaces has occurred in some of the scenarios.
|
Exposure patterns
A comparison of potential exposure loading by body part indicates a different exposure pattern for the scenarios sand blasting and spraying compared to rolling. For the sand blasting and spraying scenarios an additional cotton layer (coverall) was used, as the workers normally would wear an additional layer of clothing underneath their protective gear. The results show that the contamination for these two scenarios to the inner clothing is still higher than that for rolling. Furthermore, whereas the AM and GM for rolling and spraying seem to be comparable (for both dermal body and hand exposure loading), this is not the case for paint filling, suggesting a different distribution for the exposure loading data and different loading mechanisms. Splashing and spilling is likely to occur during mixing/loading, which could explain some of the results.
As can be seen from Tables 5 to 8, in all scenarios the loading of the hands is very important, when compared to other body parts, especially for spraying and paint filling, even when it is taken into consideration that the potential loading of the hands (as measured on protective gloves) may give an overestimation of the real potential hand exposure loading (Garrod et al., 2001).
Clothing penetration
The protective clothing worn during rolling seems to be quite effective as is shown in Table 9. For sand blasting the penetration is on average about 5%, and for spraying 3%, although an additional layer of clothing was worn (Table 10). It should be noted that there is quite some variation related to body part and between workers. For sand blasting the spray material was relatively wet (mixture of grit and water), which could have played an important role in the present data. Both the sand blasting and the spraying occurred under high pressure, which might have caused some permeation on the seams of the Tyvek®, as penetration is more likely to occur under wet conditions (a high degree of perspiration was observed). Machera et al. (2003) found that the penetration of the coverall varied with the amount of friction created by the mobility of the worker. It is therefore interesting to note that both the sand blast and the spray scenario required a lot of very active body work, e.g. kneeling and crouching under the boats, which might explain the higher penetration too. Another possible issue that may affect the observed clothing penetration is due to the ‘bellows effect’ for impermeable clothing, where due to movements air is drawn in and out at the openings (neck, wrist, ankle, closures).
Rolling scenario
Garrod et al. (2000) reported exposures while applying antifouling paint. Amateur applicators used a brush or roller to apply paint containing copper to boat hulls. Application took place outdoors. Potential dermal exposure loading was determined using gauze patches and cotton sampling gloves. Antifouling total inhalable particulate was sampled by drawing air at 2 l min–1 through a cellulose acetate filter head in a seven hole head. Amount used ranged from 1.75 to 5 l and job duration varied from 35 to 112 min (n = 10). Potential body exposure loading (not including socks) observed was 269–9960 mg paint (n = 10). Hand exposure loading ranged from 31.5 to 7890 mg paint. The wide range for hand exposure loading may be explained by the fact that the cotton monitoring gloves were worn under protective gloves. However, the worker did not always wear gloves and it was not always clear to what extent workers were using protective gloves. Statistical indicators were not quoted, because the data were sparse and did not fit log-normal distributions. The data are largely comparable with respect to exposure scenario (rolling) and product type (antifouling paint). However, there are some important differences: a different active substance (copper) was used, non-professional application and also a brush was sometimes used for application.
The professional rolling scenario of the present study resulted in a potential exposure loading (excluding hands) of 2359 mg paint per hour (GM, range 338–6296) and inhalation exposure of 0.12 mg m–3 paint. The results seem reasonably comparable, and therefore fit for regulatory purposes, even though the professional applicator shows lower body exposure loading. This seems not unreasonable and might be due to experience.
For the rolling scenario both the 7-hole sampler (SHS) and the IOM sampler were used. The SHS was used as contamination due to splashing was expected to result from the scenario (rolling), and the SHS was expected to perform better due to the smaller sample holes. Vaughan et al. (1990) measured exposure to inhalable dust in different industrial sites and found the IOM sampler to oversample in comparison with the 7-hole sampler and that the degree of oversampling was related to the tendency of the IOM sampler to collect large projectile particles. The results for the present study show a ratio IOM/SHS of about 8. It is known that the IOM samples higher than the 7-hole sampler (Vaughan et al., 1990). Kenny et al. (1997) found a median IOM/SHS ratio of 1.17 in an experimental setting (wind tunnel), but a good explanation for the large difference in sampling yield is lacking. The correlation between the pair-wise measurements of the SHS and IOM sampler was relatively high [Pearson's correlation coefficient = 0.645 (P = 0.005) (N = 17)]. If a significant amount of splashing had indeed occurred, the high correlation between the sampling techniques would not be expected since splashing is not a systematic mechanism. However, the variability of the inhalation exposure data collected using the IOM sampler (GSD = 3.2) was greater than that collected using the SHS sampler (GSD = 1.9) suggesting that there was some mechanism of the sampling method which introduced variability in the exposure data. The average inhalation exposure data collected using the IOM sampler was greater than that collected using the SHS sampler so it is possible that some degree of splashing did indeed occur. Tatum et al. (2001) found that measurements made with both IOM and 7-hole were quite variable, however could not explain this high variability. The position of the sampling head on the worker may explain differences in sampling efficiency (Kenny et al., 1999), but the position of the sampling head was not recorded in such a way that this could be used for explanation. The inside surfaces of the IOM cassette were also analysed and no inside loss in the IOM sampling head was observed. In experiments of Li et al. (2000) it was found that, depending on the orientation of the sampling head in the wind, inside losses of the 7-hole sampler increase as the particle diameter increased.
Spraying scenario
Comparable published antifouling data proved scarce; therefore, an internal report of the Health and Safety Executive was used for comparison. Some proprietary antifouling paint exposure loading studies are available in files of regulatory authorities. However, the study results can not be compared with these in the public domain.
Llewellyn (1995) studied the spray application of a copper containing antifouling paint. Potential dermal exposure loading was measured through patches (modified version of WHO, 1982) and hand exposure through thin cotton gloves under protective gloves. The patch data were extrapolated to represent the whole body. The data are given for the amount of active substance. The dermal exposure loading for this study for the spray applicators (n = 12) resulted in 1100–80 000 mg (GM 11 000) exposure (including hands) for a full days work. The hand exposure loading ranged from <0.001 to 6200 mg. Inhalation exposure loading ranged from <0.01 to 5.5 mg m–3 (GM 0.59). The present study found for inhalation exposure loading 0.26–9.0 mg m–3 (GM 2.1) (n = 12), for potential dermal body exposure loading 25–1046 mg h–1 (GM 182) and for dermal actual hand exposure loading 0.40–11 mg h–1 (GM 1.8). The inhalation exposure loading seems to compare relatively well. Assuming a 6 h day of work, the dermal exposure loading estimates are about a 10-fold apart. From observations it seems that the work carried out in the Llewellyn study occurred more often in a ‘confined space’ situation, where transfer from contaminated surfaces are more likely to occur. Furthermore, no information was provided for the Llewellyn study on the total amount mixed. This makes the results relatively hard to compare.
Also the potmen (responsible for paint filling) were studied by Llewellyn, but in this study the potmen were not present in the vicinity of the applicators. The study did monitor the ‘tender on’, a person who assist the applicator by moving the hoses, moving the applicator when necessary (while standing on a movable platform) and other tasks. The potmen in the present study carried out both tasks concurrently. Llewellyn found that the potential dermal exposure loading for the ‘tender on’ ranged from 490 to 3700 mg (GM 1400) and the inhalation exposure from 0.02 to 0.65 mgm–3 (GM 0.1) (both for 4 men). The exposure loading for the potmen ranged from 110 to 17 000 mg (GM 2500) and inhalation exposure loading from <0.01 to 0.29 mg m–3 (GM 0.07). The present study found for inhalation exposure during paint filling an exposure ranging from 0.12 to 2.5 mg m–3 (GM 0.65) (n = 10). The body exposure loading ranged from 73 to 1860 mg h–1 (GM 371) and for actual hand exposure loading from 0.2 to 3.4 mg h–1 (GM 0.9). The difference in inhalation exposure loading might be very well explained by the fact that the potmen in the Llewellyn study were not in the vicinity of the applicator during spraying. The results for the ‘tender on’ (in vicinity of the sprayer) are given above and these results compare much better with the results of the present study. The dermal results for the ‘tender on’ ranged from 490 to 3700 mg (GM 1400) for a full day's work. Again the results for the potmen are much higher in the Llewellyn study.
Garrod et al. (1998) studied spraying of remedial pesticides and found that in 78% of surveys contamination of the socks was observed. It is interesting to note that for the spraying scenario a very strong correlation with the total dermal exposure loading was observed.
Sand blasting scenario
No comparable study data were found for the secondary exposure scenario of sand blasting. However, it is of interest that exposure by inhalation to spray liquid (mixture of old paint/water/grit) occurred in 100% of the surveys. This could be of special interest as the potmen helping the operator were observed wearing surgical masks during the sand blasting, which might not provide sufficient protection. The inhalation exposure for the potmen (n = 3) varied from 0.10 to 3.9 mg m–3, which is in the same range as the inhalation observed by the sand blasting operator (0.04–1.9 mg m–3). Potential exposure for sand blasting varied from 6.1 to 106 mg h–1 and for the potmen from 4 to 433 mg h–1. This is again comparable, with the potmen wearing significantly less protective clothing than the sand blasting operators. The data refer to amount of active substance.
It should be noted that foot exposure may be relatively high due to the fact that workers wore their own safety shoes which may have been previously contaminated. Also, for one participant it was noted that he worked on his knees a lot. His exposure loading data showed an unusual high contamination on the inner dosimeter of the lower leg (compared to the other data). The data suggests that this occurred due the crawling up of clothing while sitting on the knees. This might indicate a poor fit of clothing.
Furthermore, the study data suggest that the exposure loading of the sand blasters and fillers of grit depends on the (concentration present in) antifouling paint that is being removed and the number of antifouling layers being removed. As the percentage of (still) active biocide present in the old paint varied considerably it was interesting to note that on the days that a high percentage of active biocide was present, the exposures were considerably higher (data not shown).
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
In the present study the exposure loading data are given in amount of active substance (as measured), but also as amount of paint, to make the results usable for registration purposes of paints with other active substances and/or other concentrations of the active substances.
The exposure data from the present study compare relatively well to other publicly available data for the rolling scenario, not so much for the spraying scenario. The nominal inhalation exposure levels of copper for the sprayers are quite high, and to a lesser extent for the sand blaster and the fillers, in comparison with the Occupational Exposure Level of 1 mg m–3 for an 8 h reference period for copper. For dichlofluanid such a comparison is not possible. Little data are available on secondary exposure loading.
The present study shows that the exposure loading should be considered more extensively, including applicable protective gear. In this light the findings for the potmen during sand blasting suggest that personal protective equipment should be (re)considered carefully. The results of the study seem fairly consistent and can be used for risk assessment purposes. The data will be used for validation of the BEAT dermal model (TNsG, 2002) and added to the BEAT database. Results will be published elsewhere.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
The authors thank all participating ship- and dockyards and all study participants. The authors also thank Sjaak de Vreede, Marie-José van Leeuwen, Henk Goede and Rianda Gerritsen who were instrumental to the field work. This study was financially supported by the Dutch Ministry of Social Affairs and Employment. Edith de Haan, Jeannette de Wolf, Roel Engel and Luco Ravensberg are acknowledged for chemical analysis.
|
FOOTNOTES |
|---|
Present address: DG ENTR, European Commission, Brussels, Belgium. The views expressed in this article are those of the author and do not necessarily reflect the official European Commission's view on the subject. 
Present address: Institute of Occupational Medicine, Edinburgh, UK
Received May 6, 2006; in final form August 29, 2006
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Cherrie JW. (2003) Commentary: the beginning of the science underpinning occupational hygiene. Ann Occup Hyg 47:179–85.
Garrod ANI, Rimmer DA, Robertshaw L, et al. (1998) Occupational exposure through spraying remedial pesticides. Ann Occup Hyg 42:159–65.
Garrod ANI, Guiver R, Rimmer DA. (2000) Potential exposure of amateurs (consumers) through painting wood preservative and antifouling preparations. Ann Occup Hyg 44:421–6.
Garrod ANI, Phillips AM, Pemberton JA. (2001) Potential exposure of hands inside protective gloves- a summary of data from non-agricultural pesticide surveys. Ann Occup Hyg 45:55–60.
Kenny LC, Aitken R, Chalmers C, et al. (1997) A collaborative European study of personal inhalable aerosol sampler performance. Ann Occup Hyg 41:135–53.[Medline]
Kenny LC, Aitken RJ, Beldwin PEJ, et al. (1999) The sampling efficiency of personal inhalable aerosol samplers in low air movement environments. J Aerosol Sci 30:627–38.
Li S-N, Lundgren DA, Rovell-Rixx D. (2000) Evaluation of six inhalable aerosol samplers. Am Ind Hyg Assoc J 61:506–16.
Llewellyn DM. (1995) Exposure during application of anti-fouling paints: copper and organotin compounds. Internal report, Health and Safety Executive, Redgrave Court, Bootle L20 7HS, UK.
Machera K, Goumenou M, Kapetanakis E, et al. (2003) Determination of potential dermal, inhalation operator exposure to malathion in greenhouses with the whole body dosimetry method. Ann Occup Hyg 47:61–70.
OECD. (1997) Guidance document for the conduct of studies of occupational exposure to pesticides during agricultural application, series on testing and assessment no. 9 (OCDE/GD(97)148, Paris, France: Organisation for Economic Co-operation and Development.
Omae I. (2003) Organotin antifouling paints and their alternatives. Appl Organometal Chem 17:81–105.
Schneider T, Vermeulen R, Brouwer DH, et al. (1999) Conceptual model for assessment of dermal exposure. Ann Occup Hyg 56:765–73.
Soutar A, Semple S, Aitken RJ, et al. (2000) Use of patches, whole body sampling for the assessment of dermal exposure. Ann Occup Hyg 44:511–8.
Tatum VL, Ray AE, Rovell-Rix DC. (2001) The performance of personal inhalable dust samplers in wood-products industry facilities. Ann Occup Hyg 16:763–9.
TNsG. (2002) Human exposure to biocidal products—guidance on exposure estimation. Prepared under contract B4-3040/2000/291079/MAR/E2 for the European Commission, DG Environment (available from the ECB website http://ecb.jrc.it/biocides). The user guidance for the Technical notes for Guidance is also available from this website.
Van Hemmen JJ, van der Jagt KE, Warren ND, et al. (2003) 62–5 Worker exposure assessment for biocidal products under European legislation. PharmaChem; October:.
Vaughan NP, Chalmers CP, Botham RA. (1990) Field comparison of personal samplers for inhalable dust. Ann Occup Hyg 34:553–73.
WHO. (1982) Field surveys of exposure to pesticides. Standard protocol. VBC/82.1(WHO, Geneva, Switzerland).
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  E. Tielemans, D. Noy, J. Schinkel, H. Heussen, D. Van Der Schaaf, J. West, and W. Fransman Stoffenmanager Exposure Model: Development of a Quantitative Algorithm Ann. Hyg., August 1, 2008; 52(6): 443 - 454. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. C. Alexopoulos and Y. Alamanos Secondary sex ratio in Greece: evidence of an influence by father's occupational exposure Hum. Reprod., November 1, 2007; 22(11): 2999 - 3001. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||