Exposure to Crystalline Silica in Abrasive Blasting Operations Where Silica and Non-Silica Abrasives Are Used

ABSTRACT

Exposure to respirable crystalline silica is a hazard common to many industries in Alberta but particularly so in abrasive blasting. Alberta occupational health and safety legislation requires the consideration of silica substitutes when conducting abrasive blasting, where reasonably practicable. In this study, exposure to crystalline silica during abrasive blasting was evaluated when both silica and non-silica products were used. The crystalline silica content of non-silica abrasives was also measured. The facilities evaluated were preparing metal products for the application of coatings, so the substrate should not have had a significant contribution to worker exposure to crystalline silica. The occupational sampling results indicate that two-thirds of the workers assessed were potentially over-exposed to respirable crystalline silica. About one-third of the measurements over the exposure limit were at the work sites using silica substitutes at the time of the assessment. The use of the silica substitute, by itself, did not appear to have a large effect on the mean airborne exposure levels. There are a number of factors that may contribute to over-exposures, including the isolation of the blasting area, housekeeping, and inappropriate use of respiratory protective equipment. However, the non-silica abrasives themselves also contain silica. Bulk analysis results for non-silica abrasives commercially available in Alberta indicate that many contain crystalline silica above the legislated disclosure limit of 0.1% weight of silica per weight of product (w/w) and this information may not be accurately disclosed on the material safety data sheet for the product. The employer may still have to evaluate the potential for exposure to crystalline silica at their work site, even when silica substitutes are used. Limited tests on recycled non-silica abrasive indicated that the silica content had increased. Further study is required to evaluate the impact of product recycling on crystalline silica content for non-silica abrasives. Measurement of blaster exposure was challenging in this study as the blasters evaluated conducted this task intermittently throughout the work shift, frequently removing their blasting helmets. In spite of the challenges in accurately measuring blaster exposure, the measurements were still, for the most part, over the 8-h OEL. Further work is required to develop more effective sampling strategies to evaluate blaster exposure over the full work shift when task-based monitoring is not practical.

METHODOLOGY

There were two components to this work: collection of occupational samples to evaluate exposure to crystalline silica, as well as the collection and analysis of bulk samples of silica substitutes used in abrasive blasting in Alberta.

Occupational samples for respirable crystalline silica were collected at four work sites in the summer and fall of 2011, where abrasive blasting was done on metal equipment to prepare the surfaces for the application of coatings. The samples were collected and analysed according to the National Institute for Occupational Safety and Health (NIOSH) Method 7500 and the United States Occupational Safety and Health Administration (OSHA) Method ID-42. Samples were collected over a full work shift. The sample collection method was modified by increasing the volume of air drawn through the filter from 2.5 to 2.75 l min−1. This allowed for a lower concentration to be detected so that the exposure results could be compared with the legal limit even where the OEL had been adjusted to compensate for a work shift longer than 8h (i.e. 10% of the adjusted OEL).

The air samples were collected in the workers’ breathing zone (the equipment was clipped to the collar) using personal sampling pumps pre- and post-calibrated to draw a known amount of air at an average flow rate of 2.75 l min−1 through a 37-mm 5-µm PVC filter cassette attached to an SKC GS-3 cyclone. The GS-3 cyclone is designed to meet the American Conference of Governmental Industrial Hygienists, International Organization for Standardization/European Standardization Committee size-selection curve and has a 50% cut-point of 4.0 µm aerodynamic diameter when operated at a flow rate of 2.75 l min−1. To prevent deposit of over-sized material from the cyclone grit-pot and body onto the filter, the sampler assembly was carefully handled to ensure that the cyclone was never inverted at any time. All grit-pots were emptied and cyclones were washed in hot soapy water and air-dried between each sampling session. One field blank was submitted for each sampling location. About 20% of the samples consisted of duplicates. Prior to analysis, the laboratory treated the filters with an acid wash to eliminate calcium carbonate bias.

There were challenges in measuring the full-shift exposure of the blasters. They wear a blast hood assembly that consists of a helmet attached to a cape that falls over the shoulders. Air from a compressor is blown into the helmet and flows across the blaster’s face, exiting at the gap between the helmet and the worker’s neck. This equipment was worn while the blaster was actively engaged in abrasive blasting, which comprised about half of their total activities over the work shift. The helmet was taken on and off numerous times, even while blasting, when the worker inspected their work, or while other activities were completed. Some initial measurements were taken with the sampler positioned on the outside of the helmet in previous work conducted in 2010; however, these samples were overloaded and had to be voided. After due consideration, the filter cassette and cyclone assembly were attached to the worker’s collar, as for the other workers evaluated, but under the cape. The equipment was clipped to stay in place to sample within the worker’s ‘quasi-breathing zone’ when the hood was removed. Just as when sampling for welding fumes, the filter cassette is placed inside the welding helmet to obtain a measurement of the employee’s exposure, the author’s believe that the quasi-breathing zone sample placement was the best solution, under the circumstances, to provide an estimate of the blasters’ potential for exposure to silica over the full work shift. The authors acknowledge that the sampler placement on the blasters may bias the exposure results as measuring under the cape could dilute the samples (from air flowing from the hood) or shield the sampling equipment while the helmet was worn. More discussion on the accuracy of the blaster sample results is provided later in this paper.

Bulk sampling and analysis of silica substitute products used in abrasive blasting in Alberta were also done. Most of the abrasive samples were collected directly from Alberta suppliers. Where a non-silica abrasive was used by the company and a sample had not already been provided by the supplier (ground glass sample collected in 2010 and garnet used at Company 1), bulk samples were collected at the work site. A 100-g sample of clean, unused abrasive was collected directly from the product bag or tote and analysed according to NIOSH Method 7500 for the presence of quartz silica down to 0.1% w/w. This method includes protocols for analysing bulk or settled dust samples. The 0.1% concentration was selected because it is the cut-off concentration above which disclosure of crystalline silica is required on a material safety data sheet (MSDS) under the Canadian Workplace Hazardous Materials Information System (WHMIS) legislation (Government of Canada, 1987). The MSDS for each product was obtained and reviewed to identify whether the supplier had disclosed if the product contained crystalline silica, and if so, how much was present. Two samples of used abrasive were collected at Company 1 and analysed as described above for comparison with the new, unused product. These samples were collected from product on the ground in the surface preparation area and on the wall in the liquid coating area.
RESULTS

The companies apply protective coatings to unpainted metal equipment, industrial structures, vessels, or large diameter piping. Abrasive blasting is done to prepare the metal surfaces prior to the application of the coatings. These operations were selected to ensure that the substrate being blasted would not significantly contribute to airborne silica exposure. Of the four companies assessed, one used garnet only (Company 1); one used silica sand, nickel slag, and a vitreous smelter slag product (Company 2); and two used silica sand only (Companies 3 and 4). At the time of the monitoring, Company 2 was using the slag products.

The exposure monitoring results are summarized in Table 1. Samples were collected from 27 workers; eight of whom were blasters (33 samples in total including duplicates). In addition to conducting abrasive blasting, blasters were also responsible for material handling (moving new abrasive, removing used abrasive), quality control, and work site clean-up. Abrasive blasting was conducted intermittently for about half of the total work shift; this activity was interrupted frequently while the blasters inspected their work. The blast helmets were worn only while actively conducting abrasive blasting. The other 19 workers had a variety of job classifications ranging from labourers/shop hands to painters to office personnel. Labourers and shop hands had the widest range of work activities, including material handling, equipment and vehicle operation, coating preparation, painting, and work site clean-up. For the purposes of the data analysis, labourers and shop hands are grouped together. In addition to the application of coatings by brush or spray, painters also cleaned surfaces (usually with compressed air), did surface preparation (touch ups, filled dents on surfaces, and grinding) and did work site clean-up. Work site clean-up was generally conducted using dry sweeping, shovelling, or compressed air. Company 2 also used compressed air to clean out the abrasive system prior to changing abrasives (although this did not occur at the time of the assessment).

Accumulations of used abrasive were found throughout all of the facilities to some degree; however, this appeared to be better controlled at Companies 1 and 4 as abrasive accumulation was mostly restricted to the areas where blasting was done. Blasting areas were always separated from other work areas by hoarding (screens or other hard barriers), curtains, or physical walls; however, these systems were observed to be ineffective at containing the dust as dust plumes were observed travelling beyond the barrier or were compromised (curtains tied back, doors between work areas left open). It was not unusual for other workers to enter blasting areas while blasting was being done. All of the companies provided coveralls, but they were not always worn by the workers. Dirty coveralls were often left in lockers with no laundering between uses. Although shower facilities were available, they were generally not used by workers. The companies had respirator programs, but respirator use appeared to be inconsistent. Five of the blasters were observed wearing half-face, disposable N95 respirators under the helmets (blasters at Companies 1 and 2). Among the labourers and shop hands, five were observed wearing respirators at the time of the assessment, but they were half-face air-purifying respirators equipped with organic vapour cartridges. Company 4 was the only work site where trades workers (apart from the blasters) were provided with respirators suitable for particulate (half-face, disposable P100 respirators); however, only one welder was observed wearing a respirator at the time of the assessment.

A summary of the exposure data by worker category (excluding administrative and management personnel) is provided in Table 2. Five of the eight blasters evaluated were using silica substitutes at the time of the assessment. All of the exposure samples for the blasters were over the applicable adjusted exposure limit for silica; for seven of the blasters, the samples exceeded the 8-h OEL, regardless of the abrasive in use. Because of the small size of the blaster worker group using silica sand, the geometric mean and geometric standard deviation for the airborne exposures were not calculated. Similarly, there was a high potential for over-exposure for the labour/shop hand group, regardless of the abrasive used, although exposures were somewhat lower for the blasters and labourers at the work sites where silica substitutes were used. Trades workers (painter, mechanic) also had a high potential for over-exposure at all of the work sites, even though they were not directly involved in abrasive blasting activities. Four of the samples collected for other worker categories had to be voided due to overloading or equipment failure and are not included in the results.

The results of the bulk analysis for the non-silica abrasives are provided in Table 3. A total of eight products were sampled. Only four of the products were used at the companies at the time of the exposure monitoring as most of the substitute products are used on a project-specific basis. These were the garnet used by Company 1, an aluminium oxide abrasive used in a blasting cabinet at Company 4, and vitreous smelter slag and nickel slag used by Company 2. The substitutes analysed, apart from the aluminium oxide abrasive, contained more than 0.1% w/w silica. Since most are natural products or produced from mining waste, it is not surprising that the products could contain some crystalline silica given how common this mineral is in the earth’s crust. The silica content in the ground glass product was likely the result of contamination arising during the manufacturing process (contamination of the glass by sand and gravel). The disclosure of silica content on the MSDSs for the silica substitutes was not always consistent with the bulk analysis results. For the garnet, the MSDS indicated that the crystalline silica content was <0.5%, whereas the bulk analysis of the product indicated that it was 0.76%. For the nickel slag and vitreous smelter slag products, the MSDSs indicated that the products had no crystalline silica, but the bulk analysis found 0.30% and 0.28% of crystalline silica, respectively. Company 1 also recycles the garnet product. Bulk analysis of the used material indicated that the amount of crystalline silica in the used product was substantially higher compared with the unused product (0.76% initially, increasing to 3.5% and 6.9% in samples of the used product).
