Metal and Non-Metal Miners’ Exposure to Crystalline Silica, 1998–2002

ABSTRACT

Background

Crystalline silica is well known to cause silicosis and other diseases. Exposure is common in the mining industry and consequently, the US Mine Safety and Health Administration (MSHA) evaluates miners’ exposure to silica to determine compliance with its exposure limit.

Methods

MSHA exposure measurements were obtained for the 5-year period from 1998 to 2002 and average exposure was calculated classified by occupation and by mine. Evaluation criteria were whether average values exceeded MSHA’s permissible exposure limit or the limit recommended by the National Institute for Occupational Safety and Health (NIOSH), whether there was a risk of exposure to freshly fractured silica, and whether there was a risk of a high rate of exposure to silica.

Results

Miners in certain jobs are exposed to silica above permissible and recommended exposure limits. Some miners may also be exposed at a high rate or to freshly fractured silica.

Conclusions

Known dust control methods should be implemented and regular medical surveillance should be provided.

MATERIALS AND METHODS

Data Set

The results of 8-hr full shift personal air samples taken by MSHA and analyzed for silica at metal and non-metal mines in the US for the 5-year period from 1998 to 2002 were obtained. MSHA takes personal breathing zone samples for a full 8-hr shift with a belt-mounted pump operating at 1.7 L per min. A 10-mm Dorr–Oliver cyclone removes nonrespirable particles, and the respirable fraction is deposited on a 37-mm diameter poly-vinyl chloride filter. The filter is weighed before and after sampling to determine the respirable mass and the sample is analyzed for silica using X-ray diffraction (NIOSH Method 7500). Only samples with greater than 1% silica by mass are included in these data.

The record for each sample contained, among other variables, the date of the sample, concentration of respirable dust (C, mg/m3), exposure limit (L, mg/m3), mine name, owner, and unique identification number, type of mine (underground, surface, or mill), state where mine is located, mine status (active, intermittent, non-producing, or abandoned), and the job name and job code. These data neither include information about mine operators’ use of dust controls nor include information about whether miners were using personal protective equipment.

Exposure Limit

MSHA adopted as its exposure limit the threshold limit value (TLV) set by the American Conference of Governmental Industrial Hygienists (ACGIH) in 1973 (30 CFR 56.5001 [surface mines] and 57.5001 [underground mines]). The TLV for respirable dust containing >1% quartz is calculated using the following formula:

L =  10/ (pQ + 2), in which

L = exposure limit for respirable dust (mg/m3) and
pQ = percent quartz in the sample of respirable dust.

This formula is derived by combining the TLV for mixtures, with quartz and the remaining respirable dust as the two ingredients in the mixture. The TLV for quartz is set at 100 mg/m 3 and the TLV for the remaining dust is treated as nuisance particulate at 5 mg/m 3 [Hearl, 1996].

Because of a concern with evaluating the risk of adverse health effects from exposure to silica per se, the MSHA measurements were converted into equivalent concentrations of respirable silica by taking the product of pQ and C to obtain

Q = 10(10/L -2)C, in which

Q = concentration of respirable quartz (µg/m3)
C = concentration of respirable dust (mg/m3).

The remainder of this report is concerned with the value of Q.

Statistical Evaluation

As expected, the frequency distribution of Q is lognormal. (Fig. 1a,b). Arithmetic and geometric means and standard deviations for the dataset were calculated as a whole and as classified by job, mine, and state. The potential for cumulative exposure and risk of disease was calculated using arithmetic means and trends and differences among subsets were calculated using geometric means. To the extent feasible, a comparison of these data with earlier reports by others was initiated. All calculations and graphics were done using Statistica Version 6.1 or by hand.

Evaluation Criteria

Chronic silicosis typically occurs more than 10 years from first exposure and lung cancer typically occurs with a longer period of latency. Both are associated with cumulative exposure, that is, Exp = ΣCiTi, (Ci = average concentration for time interval i and Ti = its duration) and is typically expressed with the units, mg-years/m3.

When setting permissible exposure limits, the conventional practice is to assume a working lifetime of 47 years (from age 18 to 65) as the total expected duration of exposure. A time-weighted average exposure limit for a shift is then selected that would reduce risk of disease to an acceptable level. The MSHA permissible exposure limit is roughly equivalent to a TWA limit of 100 µg/m3 of quartz.

In 1972 and again in 2000, the National Institute for Occupational Safety and Health (NIOSH) recommended an exposure limit for respirable silica of 50 µg/m3 for a 10-hr work shift as necessary to prevent silicosis [NIOSH, 1972, 2000]. Several risk assessments support this recommendation [Rice and Stayner, 1995; Steenland and Brown, 1995; Finkelstein, 2000; Chen et al., 2001; Rice et al., 2001; Steenland et al., 2001; Mannetje et al., 2002; Park et al., 2002; Attfield and Costello, 2004]. Because it is well supported by available data, it was concluded that an average exposure (over the entire 5-year period considered in these data) greater than 50 µg/m3 was a risk factor for silicosis and for lung cancer.

A second evaluation criterion arises from the apparent elevated risk of disease associated with a high rate of exposure. The exposure limit above assumes a linear relationship between cumulative exposure and the risk of silicosis. This linear relationship appears to be valid in the absence of episodes of very high exposure. However, with instances of very high exposure, even when average exposure over time is at or below an acceptable exposure limit, the risk of silicosis appears to be higher than a linear model would predict [Smith, 1992; Seixas et al., 1993; Wagner, 1997; Buchanan et al., 2003]. The level (per shift), frequency, and duration of very high exposure at which a departure from linearity occurs have not been estimated.

Nevertheless, as a precautionary consideration, the occurrence of exposure to ‘‘high’’ levels is worth noting. Therefore, as a conservative approximation, we suggest that the occurrence of more than 10% of samples for a job exceeding 150 µg/m3 constituted a second risk factor.

A third risk factor is the potential for exposure to freshly fractured silica. Risk of silicosis (and perhaps lung cancer) may be related to whether respirable dust has been freshly fractured from its parent rock. Several published reports have shown that freshly fractured quartz particles generate free radicals in vitro [Castranova, 1994; Shoemaker et al., 1995; Vallyathan et al., 1997]. Many characteristics of these free radicals, such as their pathogenic potential in vivo, the importance of the mode of fracture (e.g., drilling, crushing, blasting), the activity level related to the percent silica in the parent rock, and others have not been assessed. Nevertheless, and also as a precautionary consideration, we note those circumstances when such exposure is likely. These are situations when miners are present at the time and place that parent rock is being broken and respirable particles generated.

A positive or nil trend in exposure, in combination with any of the first three risk factors, constitutes an additional risk factor for disease. The dataset in this report cover only 5 years but even so, there may be some trends in exposure over this relatively short time interval. In addition, data from the same source evaluated earlier by others can also be used to estimate trends. Consequently, trends for the time interval encompassing these data were analyzed. These measurements, to the extent feasible, were compared with those from prior reports.

To summarize, the risk factors we used as criteria to evaluate exposure are:

- Mean exposure (5-year) above 50 µg/m3,
- The occurrence of more than 10% of samples above 150 µg/m3 for any particular job or mine,
- Potential exposure to freshly fractured quartz, and
- A positive or nil trend in exposure for jobs with other risk factors.
RESULTS

Data Editing

From 1998 to 2000, MSHA took a total of 16,578 measurements for respirable quartz at 4,726 mines operating in every state in the US for at least part of this time period. There were 114 different job codes, 50 states, and 3 types of mines (Strip/Open Pit, Mill/Prep Plant, and Underground).

Of the total, 356 samples from the data set were deleted because they were outside the bounds of possible values. From equation (1) above, the upper and lower limits to the exposure limit for total respirable dust are 3.33 mg/m3 and 0.098 (rounded to 0.10) mg/m3 , respectively, corresponding to implied concentrations of quartz of less than 1% and greater than 100%. Thirty-three samples had an exposure limit greater than 3.33 and 323 had exposure limits less than 0.10. It was assumed that these values resulted from errors that occurred during data processing. Also deleted were 15 samples from this data set that were analyzed for cristobalite. (Analysis was conducted on the remaining 16,207 samples with >1% quartz (Table I).

Percent Silica

For these dust samples, the percent silica increases with the number of samples, from 1% to 100% (Fig. 2). This distribution differs significantly from the frequency distribution noted in a report jointly prepared by the Bureau of Mines and the Public Health Service [Ankney and Heiman, 1963]. It is reported that 98% of over 20,000 samples from both surface and underground metal mines had less than 40% quartz. This difference may be a consequence of MSHA enforcement policy of sampling more frequently those mines and jobs with greater potential for exposure to quartz. In 1963, the federal government had a significantly weaker enforcement mandate.
Summary Statistics

The grand geometric mean concentration was 27.2 µg/m3 . The geometric mean minus and plus one geometric standard deviation were 10.2 µ/m3 and 72.4 µg/m3, respectively. The grand arithmetic mean was 47.2 +- 129.3 µg/m3 and the grand median 25.9 µg/m3 (Table II).

Most (85%) samples were taken at strip or open pit mines, 7% were taken at mills or preparation plants, and 8% were taken in underground mines (Table II). The differences in geometric mean did not differ significantly when classified by type of mine. The percent of samples above 150 µg/m3 ranged from 4.4% at preparation plants and mills to 5.5% in underground mines.

Samples Exceeding Exposure Limits

From among these samples taken together, 11.5% exceeded MSHA’s permissible exposure limit as described above and among these, 8.8% exceeded MSHA’s citation threshold value (CTV). This is the value above which MSHA issues a citation for non-compliance. More than one out four samples (27.2%) exceeded NIOSH’s recommended exposure limit of 50 µg/m 3 (Table III).

Higher proportions of samples exceed exposure limits in specific jobs. For stone polishers and bagging operators, approximately half of samples exceed the NIOSH REL and one out of four exceed the MSHA PEL. For cleanup workers, two out of five samples exceed the REL and one out of five exceed the PEL. For crusher operators and laborers, one out of three samples exceed the REL and one out of seven exceed the PEL (Table IV).

Concentration by Job

The mean concentration ranged from 271.7 µg/m3 for continuous miner operators (five samples taken at three mines) to 6.1 µg/m3 . For jobs with 100 or more samples, the mean concentration ranged from 75.5 µg/m 3 (stone polisher/ cutter) to 28.4 µg/m3 . More than 10% of samples exceeded 150 µg/m 3 at 14 jobs, five of which had more than 100 samples.

Distribution of values for 17 jobs with more than 100 samples is shown in Figure 3 using a box and whiskers display. This graph shows the median value near the midpoint of the box, the 25th and 75th percentiles at the lower and upper end of the box and the 5th and 95th percentiles at the lower and upper whiskers. Extreme and outlier values—less than the 5th and greater than the 95th percentile values are not shown. Variation is high for all jobs, with values extending over two orders of magnitude for all but four of these jobs. The median values range from 56 µg/m3 to 17 µg/m3 ; the 75th percentile exceeds 100 µg/m3 at one job (stone polisher/ cutter) and the 95th percentile exceeds 200 µg/m3 for four jobs (stone polisher/cutter; rotary drill operator; supervisor; and mechanic) and is near this level for one more job (bagging operator).

Risk Factors for Selected Jobs

Three risk factors (mean >50 mg/m3 , >10% of samples >150 µg/m3, and exposure to freshly fractured quartz) were present for stone polishers/cutters and ball mill operators (Table V). Two risk factors (mean >50 µg/m3 and exposure to freshly fractured quartz) were present for rotary drill operators and crusher operators. Mean exposure exceeded 50 µg/m3 for six more jobs and exposure to freshly fractured quartz occurred without other risk factors for one job (drill operator).

The purpose of selecting jobs with more than 100 samples is to focus attention on those jobs that are both numerous and have higher levels of exposure. However, this should not obscure the occurrence of risk factors on less frequently sampled jobs. These jobs are tabulated (in alphabetical order, Table VI) showing the number of samples, the average exposure, the percent of samples greater than 150 µg/m3 , and the likelihood of being exposed to freshly fractured quartz. Similar jobs are grouped together and listed if any job in the category has an average greater than 50 µg/m3 . Such jobs are cutting machine workers, drill operators, jobs listed as "miners," operators of mobile machinery, and sand fillers. The list of drill operators also includes roof bolters and drilling jobs with more than 100 samples. Average concentration of respirable quartz for nearly all drill operators is greater than 50 µg/m3 and this dust, by the usual proximity of drill operators to the drilling operation, likely also contains freshly fractured quartz. This list of jobs includes several in which rock is cut or broken (continuous miner operator, cutting machine operator or helper, and jackhammer operator). These data should be interpreted with caution because as the number of samples decreases so does the relevance of the measure and the stability of the mean and of the percent of samples greater than 150.

Concentration by Job and by Type of Mine; Two-Way ANOVA

A two-way analysis of variance (not shown) of log-transformed values classified by job (with >100 samples) and type of mine showed that the geometric mean concentration varied significantly (P < 0.001) by job and by mine. The concentration for specific jobs is somewhat higher at strip and open pit mines than for the same jobs at other types of mines. However, the small P-value is more a consequence of the large number of data points than of meaningful differences.

Concentration by Mine

During this 5-year time period, there was only one sample per mine for 69% of all mines and there were fewer than 10 samples per mine for 93% of mines. Such few samples do not provide reliable estimates of exposure for specific mines. The distribution in the number of samples per mine is shown in Figure 4a.

Mine-specific mean and geometric mean values for mines with more than 10 samples show that mean exposure was above 100 µg/m3 for 2 mines, was between 50 µg/m3 and 100 µg/m3 for 29 mines, and was <50 µg/m3 for the remaining 88 mines. The maximum exposure was greater than 1,000 µg/m3 for 2 of these mines, it was between 500 µg/m3 and 1,000 µg/m3 for 20 mines, was between 100 µg/m3 and 500 µg/m 3 for 130 mines, and was less than 100 µg/m3 for the remainder. Examination of job-specific values within mines with very high concentration shows that the high values almost always occurred at jobs listed in Figure 3.

There are a large number of mines with high mean concentration but few samples (Fig. 4b). A small number of samples at mines with very high levels of exposure is not consistent with a policy of taking more samples at mines where exposure is elevated. There were 191 samples that exceeded MSHA’s CTV and were the only samples at these mines during the time interval of these data. This suggests there were no follow-up inspections at these mines although there could have been other types of follow-up. It is conceivable that these samples were taken towards the end of the 5-year time interval and that follow-up inspections occurred after the end of this interval and therefore were not included in this data set. These inspections, however, occurred throughout this time interval and were not concentrated at the end.

Concentration by State

The mean concentration of respirable silica by state (all mines combined) ranges from 84.8 µg/m3 to 9.1 µg/m3 . The mean value exceeds 50 µg/m 3 in 15 states. Five of the 10 states with the highest mean concentration are Rocky Mountain States, Nevada, Utah, Montana, Colorado, and Idaho.

A two-way ANOVA of log-transformed values for mountain states and type of mine showed significant differences between states, by type of mine, along with significant interaction. The highest geometric mean concentration occurred for underground mines, particularly in Nevada, Utah, and Idaho. Levels are generally higher in Nevada and Utah and lower in New Mexico and Arizona.

Trend Analysis

For the data set as a whole there is no discernable trend by simple linear regression of log-transformed data on the date of the sample. Although the very large number of samples produced results with P < 0.05, the adjusted value of R2 was only 0.004. Similarly, there was no trend for selected jobs (bagging, stone polishing, crusher operator, cleanup, and laborer).

These data were compared with prior measurements by other investigators. As it was not possible to match jobs exactly, comparisons should be interpreted with caution. Published reports of prior exposure to silica among metal and non-metal miners show higher geometric mean exposure for bagging operators, lower levels for workers in crushing or mill operations, and approximately the same level for other jobs (Table VII). As part of a risk assessment of silicosis among gold miners in the US, Steenland and Brown [1995] evaluated miners’ exposure from 1940 to 1965. They reported a median level of 50 µ/m3 with 150 µg/m3 for workers hired before 1930. Steenland and Sanderson [2001], in a study of lung cancer among industrial sand workers, evaluated over 4,000 samples collected from 1974 to 1995 and reported, as in the study of gold miners, a median exposure level of 50 µg/m3. Separately, these same authors provided additional information for the same population and reported a geometric mean of 25.9 µg/m3, with some exposure measurements as high as 11,700 µg/m3 [Sanderson et al., 2000]. (The grand geometric mean in the present data is 27.2 µg/m3/.) More than a third of samples exceeded the MSHA exposure limit and about half exceeded the NIOSH REL. In the present data analysis, 11.5% of samples exceeded MSHA’s exposure limit and 27.2% exceeded the REL, suggesting substantial improvement.

Kullman et al. [1995] reported on exposure to crushed stone operations from 1979 to 1982. They reported that one in seven samples exceeded the MSHA permissible exposure limit (i.e., the 1973 TLV) and one in four exceeded the NIOSH recommended limit of 50 µg/m3. Mill operators and  mill laborers consistently had the highest and most frequent overexposure to silica. In our data analysis, one of seven (14.7%) samples of crusher operators exceeded MSHA’s exposure limit and one of three (32.6%) exceeded the REL, suggesting an increase in exposure for this job. In our data, stone polishers consistently had the highest levelof exposure; we had no job comparable to mill operators or laborers.

For three out of four jobs, the geometric mean is higher in the more recent period and for one job (bagging), recent exposure measurements are lower by a third. In the 1974-1995 period, the geometric mean for bagging operator was 60.2 µg/m3 but this had been reduced to 44.6 µg/m3 by 1998-2002.

Watts et al. [1989] analyzed nearly 12,000 samples taken at all metal and non-metal mines from 1974 to 1981. Data were obtained from the same source as these data, that is, they were taken by MSHA and its predecessor agencies for compliance purposes. Results were compared to the 1973 TLV for silica. Over this time period, they found significant decline in the percent of samples greater than the MSHA exposure limit, from 46% in 1974 to as low as 14% in 1979. There were significant differences between jobs and types of mines.

For baggers, Watts et al. [1989] found that 39.8% of samples exceeded the exposure limit, as did 21.9% of laborers’ samples, 18.5% of crushing machine operators, and 19.4% of samples for rotary drill operators for the period 1978–1981. Higher levels were reported for the period 1974–1977, prior to the creation of MSHA.

For the period considered in this report, 25% of samples taken for stone-cutters, 23% of those for bagging, 20% of samples for cleanup workers, 17% of samples for general laborers, and 15% of samples taken at crushing jobs exceeded the exposure limit. For drill operators, 15.7% of samples for rotary (air) drill operators exceeded the exposure limit as did 11.0% of samples for rotary drill operators (Table VIII). Reduction in exposure for drill operators may result from dust control regulations promulgated by MSHA [1994]. There was a significant reduction in exposure for drill operators at coal mines followed this rule [Weeks, 2002].

These comparisons (Table VIII) show substantial reduction in the number of samples above the exposure limit for baggers, laborers, and drill operators and modest reduction for stone crushing workers.
