Respirable Crystalline Silica Exposures During Asphalt Pavement Milling at Eleven Highway Construction Sites

Abstract

Asphalt pavement milling machines use a rotating cutter drum to remove the deteriorated road surface for recycling. The removal of the road surface has the potential to release respirable crystalline silica, to which workers can be exposed. This paper describes an evaluation of respirable crystalline silica exposures to the operator and ground worker from two different half-lane and larger asphalt pavement milling machines that had ventilation dust controls and water-sprays designed and installed by the manufacturers.

Manufacturer A completed milling for eleven days at four highway construction sites in Wisconsin, while Manufacturer B completed milling for ten days at seven highway construction sites in Indiana. To evaluate the dust controls, full-shift personal breathing zone air samples were collected from an operator and ground worker during the course of normal employee work activities of asphalt pavement milling at eleven different sites.

Forty-two personal breathing zone air samples were collected over 21 days (sampling on an operator and ground worker each day). All samples were below 50 µg/m³ for respirable crystalline silica, the National Institute for Occupational Safety and Health recommended exposure limit. The geometric mean personal breathing zone air sample was 6.2 µg/m³ for the operator and 6.1 µg/m³ for the ground worker for the Manufacturer A milling machine. The geometric mean personal breathing zone air sample was 4.2 µg/m³ for the operator and 9.0 µg/m³ for the ground worker for the Manufacturer B milling machine. In addition, upper 95% confidence limits for the mean exposure for each occupation were well below 50 µg/m³ for both studies. The silica content in the bulk asphalt material being milled ranged from 7% to 23% silica for roads milled by Manufacturer A and from 5% to 12% silica for roads milled by Manufacturer B.

The results indicate that engineering controls consisting of ventilation controls in combination with water-sprays are capable of controlling occupational exposures to respirable crystalline silica generated by asphalt pavement milling machines on highway construction sites.

INTRODUCTION

Worker exposure to respirable crystalline silica can occur in agriculture, foundry work, hydraulic fracturing, mining, sandblasting, stone and granite work, and construction.^(1–13) Many construction tasks have been associated with overexposure to crystalline silica.^{(14, 15)} Among these tasks are tuck pointing, concrete sawing, concrete grinding, concrete scabbling, jackhammering, installing roof tiles, and abrasive blasting.^(16–24) Road milling has also been shown to result in overexposures to respirable crystalline silica.^{(14, 25, 26)} However, the three road-milling studies do not provide enough information about the operating parameters and engineering controls present on the milling machines to determine if the overexposures were due to a lack of effective controls or poor maintenance of the machines.

A variety of machinery are employed in asphalt pavement recycling, including cold-planers, heater-planers, cold-millers, and heater-scarifiers.⁽²⁷⁾ Cold-milling is the focus of this article. Cold-milling, which uses a toothed, rotating cutter drum to grind and remove the pavement to be recycled, is primarily used to remove surface deterioration on both petroleum-asphalt aggregate and Portland-cement concrete road surfaces.⁽²⁷⁾ Key components of a typical half-lane and larger asphalt pavement milling machine used for cold milling are shown in Figure 1.

Figure 1.

Half-lane and larger asphalt pavement milling machine (Illustration by NIOSH)

Approximately 251,000 U.S. workers are employed in highway, street, and bridge construction.⁽²⁸⁾ A number of these workers use cold-milling machines or work in close proximity to the machine. Dust generated from the machines often contains respirable crystalline silica. This respirable dust can be transported by air currents to worker breathing zones near the milling machine.

The testing described in this article was coordinated by the Silica/Asphalt Milling Machine Partnership and is a result of a collaborative effort by labor, industry, and government to reduce respirable crystalline silica exposure during asphalt pavement milling in highway construction. This Silica/Asphalt Milling Machine Partnership is coordinated by the National Asphalt Pavement Association (NAPA) and includes all U.S. and foreign manufacturers of heavy construction equipment that currently sell pavement-milling machines to the U.S. market. In addition to NAPA and the equipment manufacturers, the Silica/Asphalt Milling Machine Partnership includes numerous paving contractors, the International Union of Operating Engineers, the Laborers International Union of North America, the Association of Equipment Manufacturers, and government organizations including the Occupational Safety and Health Administration (OSHA), the Federal Highway Administration, and the Centers for Disease Control and Prevention’s (CDC’s) National Institute for Occupational Safety and Health (NIOSH).

One of the aims of the Silica/Asphalt Milling Machine Partnership, and the focus of this article, is to evaluate engineering controls developed to reduce silica exposures among workers on half-lane and larger cold-milling machines. The engineering controls evaluated in this study included ventilation controls and water-spray systems used to cool the cutting teeth on asphalt pavement milling machines. The water-spray dust suppression controls were evaluated during previous field studies.⁽²⁹⁾ The capture efficiency of the ventilation dust controls were evaluated using tracer gas in a factory setting before field testing was conducted.^{(30, 31)} The manufacturers each optimized silica dust controls as part of the Partnership. The purpose of this final phase of testing was to verify the effectiveness of the final engineering control configuration before installation on an entire fleet of milling machines.

Silica Health Effects and Exposure Limits

Inhalation of respirable crystalline silica can cause silicosis, a debilitating and potentially fatal lung disease. Silica exposure has also been associated with lung cancer, chronic obstructive pulmonary disease, renal disease, and other adverse health outcomes.⁽³²⁾ During the period from 1990 through 1999, at least one-third of decedents with silicosis had worked in construction or mining.⁽³³⁾ The NIOSH recommended exposure limit (REL) for respirable crystalline silica is 50 µg/m³ as a time-weighted average (TWA) determined during a full-shift personal breathing zone (PBZ) sample. This REL is applicable for most workers who work up to a 10-hr workday during a 40-hr workweek to reduce the risk of developing silicosis, lung cancer, and other adverse health effects.⁽³²⁾

The current OSHA permissible exposure limit (PEL) for respirable dust containing crystalline silica for the construction industry is measured by impinger sampling. In the construction industry, the PELs for cristobalite and quartz are the same.⁽³⁴⁾

Since the PELs were adopted, the impinger sampling method has been rendered obsolete by gravimetric sampling.⁽³⁵⁾ OSHA currently instructs its compliance officers to apply a conversion factor when converting between gravimetric sampling and the particle count standard when characterizing construction operation exposures.⁽³⁶⁾

On September 12, 2013, OSHA published a Notice of Proposed Rulemaking (NPRM) for occupational exposure to respirable crystalline silica. The NPRM was published in the Federal Register and proposes a PEL of 50 µg/m³ for respirable crystalline silica as an 8-hr TWA exposure.⁽³⁷⁾

The American Conference of Governmental Industrial Hygienists (ACGIH) Threshold Limit Value (TLV) for quartz and cristobalite (respirable fraction) is 25 µg/m³.⁽³⁸⁾ The documentation to the TLV states that “it is the concern about fibrosis (silicosis) and the precedent inflammatory process resulting from silica exposures, and the association of inflammation and fibrosis with lung cancer that leads to this recommendation.”⁽³⁹⁾

Description of the Dust Controls

A NIOSH document Best Practice Engineering Control Guidelines to Control Worker Exposure to Respirable Crystalline Silica during Asphalt Pavement Milling provides detailed recommendations for ventilation controls and water-sprays on asphalt milling machines.⁽⁴⁰⁾ The ventilation control recommendations focus on providing an enclosure around the drum housing and conveyors, proper hood and duct design, airflow capacity, durability of the duct and fan, and measures to prevent clogging of the ventilation control. To contain silica dust, the manufacturers designed their systems to remove air and maintain negative air pressure in the drum housing of the milling machine to contain the source of dust generation. Milling machine manufacturers also optimized water-sprays along the primary and secondary conveyor in addition to the original water applications applied only to the drum to cool cutting teeth.

The equipment provided by Manufacturer A and used during the first 11 days at four sites was an asphalt milling machine with dual diesel engines capable of producing 534 kilowatt (kW) (716 horsepower (HP)) that comply with emissions standard EC Stage 3b / US Tier 4i. The Manufacturer A asphalt milling machine had a 2 m 79-inch) wide cutter drum with a ventilation control and water spray dust suppression. The ventilation control used a fan and flexible hose system to draw air through a hood through the top cover of the primary conveyor nearest the drum housing which created negative air pressure in both the primary conveyor and drum housing areas. The ventilation control did not include any filtration but exhausted the air at the end of the secondary conveyor away from any workers. The water spray dust suppression system included water applied to the drum housing to cool the teeth and suppress dust as well as additional water spray nozzles in the primary conveyor to suppress dust along the conveyed path of recycled asphalt pavement. The maximum water flow rate was 18 gallons per minute but was adjusted by the operator depending on milling machine speed and depth of cut.

The equipment provided by Manufacturer B and used during the last 10 days at 7 sites was an asphalt milling machine with an 2.2 m 86-inch) wide cutter drum and a diesel engine that provides 462 kW (620 HP) at 1850 rpm. The Manufacturer B asphalt milling machine was fitted with a water spray system and ventilation controls consisting of a hydraulic powered 5.2 kW (7-hp) Ilmeg fan connected to a 6-inch (15 centimeter (cm)) diameter duct leading to a manifold that split the flow into two 4-inch (10 cm) diameter ducts that exhausted air at the top of the secondary conveyor. The ventilation control used a fan and flexible hose system to draw air through multiple slots through the top cover of the primary conveyor near the drum housing and near the conveyor transition area which created negative air pressure in both the primary conveyor and drum housing areas. The ventilation control did not include any filtration but exhausted the air at the end of the secondary conveyor away from any workers. The water spray dust suppression system included water applied to the drum housing and along the primary conveyor and secondary conveyors as shown in Figures 5, 6, and 7 of NIOSH Publication No. 2015-105.⁽⁴⁰⁾ The maximum water flow rate was 23 gallons per minute but was adjusted by the operator depending on milling machine speed and depth of cut.

Description of the Operator and Ground Worker Tasks

PBZ air samples for respirable crystalline silica were collected from the milling machine operator and ground worker during the course of normal employee work activities of asphalt pavement milling. The milling machine operator would typically spend the entire shift on the operator bridge of the milling machine which is located above the cutter drum housing. The operator is responsible for adjusting controls such as milling machine speed, steering, depth of cut, water flow to cutting teeth, and maintaining communications with dump truck drivers using hand signals to fill trucks with recycled asphalt pavement. The ground worker operates the ground controls on the back sides of the milling machine and would typically spend most of the shift walking next to the machine at ground level several feet away from the drum housing. The primary task of the ground worker is to operate ground level milling machine controls associated with the quality of cut for the road surface being milled. The ground worker may also be responsible for tasks such as coordinating traffic flow around the machine especially when milling through intersections, connecting the hose from the water truck to the milling machine, and various additional tasks.

METHODS

PBZ air samples for respirable crystalline silica were collected from the milling machine operator and ground worker using respirable dust cyclones (model GK2.69, BGI Inc., Waltham, MA) at a flow rate of 4.2 liters/minute (L/min) with battery-operated sampling pumps (Gilian model GilAir^® Plus, Sensidyne^®, Clearwater, FL) calibrated before and after each day’s use. A sampling pump was clipped to each sampled employee’s belt and connected via Tygon^® tubing and a tapered Leur-type fitting to a pre-weighed, 37-mm diameter, 5-micron (µm) pore-size polyvinyl chloride filter supported by a backup pad in a three-piece filter cassette sealed with a cellulose shrink band (in accordance with NIOSH Methods 0600 and 7500).^{(41, 42)} The front portion of the cassette was removed and the cassette was attached to a respirable dust cyclone and placed in the breathing zone of the worker.

The filter samples were analyzed for respirable particulates in accordance with NIOSH Method 0600,⁽⁴¹⁾ for which the limit of detection (LOD) was 30 µg/sample, and the limit of quantitation (LOQ) was 110 µg/sample. The results were blank corrected with the average of the media blanks.

Crystalline silica analysis of filter samples was performed using X-ray diffraction in accordance with NIOSH Method 7500.⁽⁴²⁾ The LODs for quartz, cristobalite and tridymite are 5 µg/sample, 10 µg/sample, and 10 µg/sample, respectively. The LOQs for quartz, cristobalite, and tridymite are 17 µg/sample, 33 µg/sample, and 33 µg/sample, respectively.

Bulk samples were analyzed in accordance with NIOSH Method 7500, for which the LODs for quartz, cristobalite, and tridymite in bulk samples were 0.3%, 0.3%, and 0.5%, respectively. The LOQs for quartz, cristobalite, and tridymite in bulk samples were 0.83%, 0.83%, and 1.7%, respectively.

Statistical Methodology

The statistical criterion used for effective performance of each manufacturer’s control system is that the upper 95% confidence limit for the arithmetic mean of each occupation’s PBZ air sampling results should be less than the NIOSH REL of 50 µg/m³. These upper confidence limits for each occupation and manufacturer are computed assuming the lognormal distribution for the PBZ samples.⁽⁴³⁾ Tolerance limits for 95% of the population of PBZ air samples are also provided. The goal of the study was to compare results with the NIOSH REL and not to compare manufacturers against each other.

The PBZ air sampling results were analyzed separately for each occupation and manufacturer. The methods available for obtaining confidence limits for the arithmetic mean of lognormal distributions are limited to relatively simple models.^(44–46) The methodology used is for models with a mean and two variance components and is described in more detail in Appendix C of NIOSH Publication 2015-105 which uses the “Method of Variance Estimates Recovery” (MOVER).⁽⁴⁰⁾

The arithmetic mean of the log normally distributed PBZ air sampling results for each manufacturer’s occupation is shown in Equation 1:

$$\text{exp}(\mathrm{\mu }+0.5({{\mathrm{\sigma }}^2}_{\mathrm{s}}+{{\mathrm{\sigma }}^2}_{\mathrm{s}\mathrm{d}})),$$ (1)

where,

• µ =log scale mean;

• σ²_s = between site variance;

• σ²_sd = within site variance.

The upper confidence limits in Tables II and IV were calculated using Equation 2, which is based on a statistical model with one mean and two variance components, as in Equation 1 above:.

$$\mathrm{U}\mathrm{C}\mathrm{L}2=(\overline{\overline{x}}+0.5s_{{\overline{x}}_s}^2+0.5(1-n\mathit{\text{bar}})ast{d}_{ws}^2+\sqrt{{\left(s_{{\overline{x}}_s}\frac{\mathrm{t}(0.95,\mathrm{d}\mathrm{f}1)}{\sqrt{n_{\mathit{\text{sites}}}}}\right)}^2+{0.5}^2+s_{{\overline{x}}_s}^4{(\frac{df1}{\text{chisq}(0.05,\mathrm{d}\mathrm{f}1)}-1)}^2+{(0.5(1-n\mathit{\text{bar}}))}^{2}ast{d}_{ws}^4{(\frac{df2}{\text{chisq}(1-0.05,\mathrm{d}\mathrm{f}2)}-1)}^2}(2)$$ (2)

where,

• x̿ = mean of log-scale site means of PBZ samples; Tables II and IV geometric means obtained by exponentiation;

• $s_{{\overline{x}}_s}^2$ = sample variance of log-scale site means of PBZ samples;

astd²_ws = pooled within-site variance on log-scale, computed by weighting within-site variances by their degrees of freedom, summing, and dividing by sum of degrees of freedom; geometric standard deviation for within sites in Tables II and IV were obtained by exponentiating the square root of this value;

• n_sites = number of sites;

• df1= n_sites −1;

• df2 = number of measurements-number of sites;

• nbar = average of reciprocals of number of days at each site;

• t(0.95,df1) = 95^th percentile of t distribution, df1 degrees of freedom;

• chisq(0.05,df1) = 5^th percentiles of the chi square distribution, df1 degrees of freedom;

• chisq(1–0.05,df2) = 95^th percentiles of the chi square distribution, df2 degrees of freedom;

• UCL_o=e^UCL2 is the upper confidence limit for the arithmetic mean of occupation o

Table II.

Manufacturer A PBZ statistics in µg/m³ with estimates

Occupation	Number of samples	Number samples < LODA	Maximum (µg/m3)	Number of sites	AM (µg/m3) Original scale	GM (µg/m3)	Between Site GSD	Within Site GSD	Total GSD	GSD- based total RSD (relative standard deviation)	Upper 95% confidence for arithmetic mean (µg/m3)B	%Confidence that 95% of population < RELC
Operator	11	1	13	4	7.2	6.2	1.50	1.77	2.02	80%	28.2	87%
Ground man	11	1	10	4	6.6	6.1	1.20	1.67	1.72	58%	13.5	>95%

^AMDC /sqrt(2) was substituted for these two samples for use in the calculations.

^BEquation 2 was used separately for each occupation.

^CEquation 11 in Krishnamoorthy et. al was used separately for each occupation.⁽⁴⁴⁾

Table IV.

Manufacturer B PBZ statistics in µg/m³

Occupation	Number of Samples	Number < LODA	Maximum (µg/m3)	Number of sites Used	AM (µg/m3) Original scale	GM (µg/m3)	Between Site GSD	Within Site GSD	Total GSD	GSD-based total RSD, Relative standard deviation	Upper 95% confidence limit for arithmetic mean (µg/m3)B	%Confidence, 95% of population <RELC
Operator	10	1	11	6	4.9	4.2	1.70	1.43	1.90	71%	11.8	>95%
Ground worker	10	1	24	6	10.8	9.0	1.86	1.35	1.99	78%	29.8	80%

^AThe two values for site 3 were omitted from the statistical analysis because the work done was atypical.

^BEquation (2) was used separately for each occupation.

^CEquation 11 in Krishnamoorthy et. al was used separately for each occupation.⁽⁴⁴⁾

Geometric standard deviations between sites are the exponentiated values $\sqrt{s_{{\overline{x}}_s}^2-n\mathit{\text{bar}}\times ast{d}_{ws}^2}$. GSD-based RSD values in Tables II and IV were calculated using the formula sqrt(e^{[ln(GSD) × ln(GSD)]}−1).⁽⁴³⁾

The upper tolerance limits in Tables II and IV also use the MOVER. The aim is to determine the largest confidence level (1−α) for which 95% of PBZ measurements from the entire population (from which the chosen sites and workers come) are less than the REL. The upper tolerance limits use the exponentiated values of equation 11 in Krishnamoorthy et. al.⁽⁴⁴⁾ The quantities used in eq (2) are also used for these limits. Ogden et al. recommend using between 70% and 80% upper tolerance limits for 95% of the population in compliance decisions for small sample sizes.⁽⁴⁷⁾

The air sampling data for the two manufacturers differ in number of days at sites. Manufacturer A included three sites with three days of sampling and a fourth site with two days. Manufacturer B included five sites with one day of sampling and two sites with multiple days. For Manufacturer B, within sites variance is based on the two sites with multiple days.

For Manufacturer A, there were three different operators and five different ground workers. Some individuals worked at one site, but others worked at multiple sites. Manufacture B used the same operator for all sites, and the same ground worker for the first six sites with a different one at the seventh site.

SAS Proc Mixed was used to evaluate the between worker variance components separately for each study.⁽⁴⁸⁾ The dependent variable in each study model was all log scale PBZ samples for the two occupations. Including both occupations allows for more statistical power. Whereas the model used for Equation 2 had one mean and two variance components, this new model had two means, one for each occupation, and four variance components: between site and within site variances, as in expression (1), and the variance between workers in occupations and the residual variance. For both Manufacturer A and B, the test of between worker variance components did not yield statistically significant results (5% level) via F test⁽⁴⁸⁾ or likelihood ratio⁽⁴⁹⁾; the between worker variance component was therefore removed. That variance component is also not needed in the model for Equation 2. Thus, for each occupation in each study, the between site and within site variance components were the two components used. The need to use relatively simple models restricts the number of components to two.

Manufacturer A had one day with sample values less than the limit of detection (LOD) for both workers, for which the substitution MCD/√2 was made, where MDC is the concentration corresponding to the sample when the LOD is used as the reported mass.⁽⁵⁰⁾

In multiple regression with seven explanatory variables (discussed in a section below), the log scale average of the two PBZ samples per day was used as the dependent variable. All subsets regression via the Cp criterion⁽⁵¹⁾ was used to reduce the number of variables in the model. Each retained numerical variable’s effect was quantified by taking the difference, 75^th – 25^th percentile values of the variable, multiplying the difference by the variable’s regression coefficient, and exponentiating. This is the effect of the 75th relative to 25^th percentile of this variable when all other variables are held constant (Table VI).

Table VI.

Estimates of Explanatory Variable Regression Coefficients

Variable	Estimated Regression Coefficient	25th percentile of variable	75th percentile of variable	Regression Coefficient × (75th percentile −25th percentile)	Ratio of 75th to 25th Predicted Values, When Other Variables Are Held Constant (Exponentiated Value of Column 5)
Average Machine Speed (ft/min)	0.047	40	60	0.949	2.583
Average Cut Depth (in)	0.291	1.5	4.5	0.874	2.397
%Silica in Bulk	0.078	7.83	17.25	0.738	2.092
Day relative to night work	−0.659				0.518A

^AThe exponentiated value of −0.659 is 0.518. This means that the average day TWAs are about 0.518 of the average night TWAs, based on the statistical model.

RESULTS

NIOSH researchers conducted 42 full-shift PBZ air sampling for respirable crystalline silica from the operator and ground worker on two different asphalt pavement milling machines. The sampling was conducted over 21 days at 11 highway construction sites during the course of normal employee work activities of asphalt pavement milling. A milling machine provided by Manufacturer A was used during the first 11 days at 4 sites, and a milling machine provided by Manufacturer B was used for milling during the remaining 10 days at 7 sites.

Diversity in asphalt pavement milling conditions were present. The selected sites included diversity in day and night milling with milling during night shifts occurring in 8 of the 21 shifts. The study included diversity of shift lengths with 79% of the shifts lasting between 8 to 12 hours. Of the 42 PBZ air samples, 38 of the 42 PBZ air samples were from shifts lasting longer than 7 hours, 33 of the 42 PBZ air samples were from shifts lasting longer than 8 hours, 20 of the 42 PBZ air samples were from shifts lasting longer than 9 hours, 11 of the 42 PBZ air samples were from shifts lasting longer than 10 hours, and 5 of the 42 PBZ air samples were from shifts lasting longer than 11 hours.

Diverse types of highway construction sites were selected. The roads milled included rural highways, city streets, a parking lot, and a major freeway. Recycled asphalt pavement removal depths included typical 1.5 to 3 inch (3.81 to 7.62 cm) mill and fill removals as well as full-depth removals of up to 11 inches (28 cm) of recycled asphalt pavement. Mill and fill removals accounted for 13 of the 21 shifts with typical milling machine speeds ranging from 40 to 80 feet per minute (fpm) (0.2 to 0.41 meters per second (m/s). Removals of 4 to 6 inches (10 to 15 cm) of recycled asphalt pavement accounted for 3 of the 21 shifts with typical milling speeds ranging from 40 to 50 fpm (0.2 to 0.25 m/s). Full depth removals of 9 to 11 inches (23 to 28 cm) of recycled asphalt pavement accounted for 4 of the 21 shifts with milling machine speeds ranging from 20 to 30 fpm (0.1 to 0.15 m/s). On one of the 21 days of milling, the machine was used to remove only 1-inch (2.54 cm) of asphalt pavement at an average speed of 12 fpm (0.06 m/s) which was not considered to be representative of typical asphalt pavement milling for that 6-hour shift.

Weather

The weather was recorded from the National Oceanic and Atmospheric Administration (NOAA) fixed weather station nearest to each evaluated site. Ambient temperatures ranged from 0°C to 33°C and average wind speeds per shift ranged from 0.9 to 5.0 m/s.

Manufacturer A PBZ Air Sampling Results

Full-shift PBZ silica air sampling results during the 11 days of air sampling at 4 sites for the operator and ground worker using the Manufacturer A asphalt milling machine are shown in Table I along with the silica content in the bulk and filter samples for each day. At the 4 sites studied, the percent bulk silica content in the roads being milled ranged from 7 and 23%, with an average of 16%. Silica content in the PBZ filter air samples for the operator and ground worker using the Manufacturer A asphalt milling machine ranged from below the LOD to 14%. The 22 full-shift PBZ air sampling results for the operator and ground worker using the Manufacturer A milling machine ranged from below the minimum detectable concentration to 13 µg/m³.

Table I.

Manufacturer A Full-shift PBZ silica air sampling results in µg/m³

	Highway 92 (Mount Horeb, Wisconsin)			Eclipse Center Parking lot (Beloit, Wisconsin)		Highway 50 (Kenosha, Wisconsin)			Multiple residential city streets (Menomonee Falls, Wisconsin)
Silica Results	Tuesday	Wednesday	Thursday	Friday	Saturday	Wednesday night	Thursday night	Friday night	Monday	Tuesday	Wednesday
	26-Jun	27-Jun	28-Jun	29-Jun	30-Jun	11-Jul	12-Jul	13-Jul	30-Jul	31-Jul	1-Aug
Operator
Concentration	NDA	3	6	9	10	6	9	11	13	3	8
% silica	NDA	11%	12%	14%	12%	6%	7%	10%	8%	4%	6%
Sample time	561	700	284	530	455	493	436	480	537	650	550
MDCB	2.1	1.7	4.2	2.2	2.6	2.4	2.7	2.4	2.3	1.9	2.2
Ground Worker
Concentration	NDA	5	7	8	7	6	9	10	10	4	7
% silica	NDA	11%	13%	14%	12%	6%	6%	9%	7%	4%	5%
Sample time	560	518	279	527	455	500	440	476	525	646	552
MDCB	2.1	2.3	4.3	2.3	2.6	2.4	2.7	2.5	2.3	1.9	2.2
% silica bulk	18%	19%	20%	23%	23%	9.6%	15%	17%	16%	7%	10%

^AND=non-detect (there was no quartz detected in the filter samples on June 26)

^BField limit of detection or Minimum Detectable Concentration (MDC) = (LOD of 5 µg/sample)/(volume in cubic meters)

Table II shows the means and upper 95% confidence limits for the two occupations and other summary statistics. The geometric mean respirable crystalline silica exposure for the operator was 6.2 µg/m³ with an upper 95% confidence limit for the arithmetic mean of 28.2 µg/m³. The geometric mean respirable crystalline silica exposure for the ground worker was 6.1 µg/m³ with an upper 95% confidence limit for the arithmetic mean of 13.5 µg/m³. The upper confidence limits are less than the NIOSH REL of 50 µg/m³.

Table II also shows the confidence of being less than the NIOSH REL for 95% of the population of values from which manufacturer A data were drawn. For the ground worker, 95% of the population is less than the REL with greater than 95% confidence, and the confidence for the operator is about 87%.

Manufacturer B PBZ Air Sampling Results

Full-shift PBZ air sampling for silica during the ten days of sampling at seven sites for the operator and ground worker using the Manufacturer B asphalt milling machine are shown in Table III along with the silica content in the bulk and filter samples for each day. At the 7 sites studied, the percent bulk silica content in the road material being milled ranged from 5 and 12%, with an average of 8%. Silica content in the PBZ filter air samples for the operator and ground worker using the Manufacturer B asphalt milling machine ranged from below the LOD to 9%.

Table III.

Manufacturer B Full-shift PBZ silica air sampling results in µg/m³

	US-31N Carmel, IN	Washington and S HS Rd, Indianapolis, IN	East McCarty, Indianapolis, IN	South Ford Rd, Zionsville, IN	I-465 between Allisonville Rd and I-69, Indianapolis, IN		US-24, Wabash, IN	US-24, Ft. Wayne, IN
Silica (µg/m3)	18-Sep	19-Sep	20-Sep	21-Sep	27-Sep	28-Sep	4-Oct	10-Oct	11-Oct	12-Oct
Operator
Concentration	4	5	NDA	11	2	2	2	7	6	3
% silica	6%	4%	NDA	6%	2%	3%	2%	3%	4%	3%
Sample time	484	611	338	540	612	678	676	533	559	519
MDCB	2.5	2.0	3.5	2.2	1.9	1.8	1.8	2.2	2.1	2.3
Ground worker
Concentration	12	24	NDA	12	4	5	4	6	13	8
% silica	9%	6%	NDA	8%	3%	3%	3%	3%	3%	5%
Sample time	482	614	370	539	620	677	685	533	556	519
MDCB	2.5	1.9	3.2	2.2	1.9	1.8	1.7	2.2	2.1	2.3
% silica bulk	5.0%	5.4%, 12%B	7.2%	11.4%	5.5%	6.3%	5.8%	8.1%	8.9%	9.1%

^AND=non-detect (there was no quartz detected in the filter samples on September 20^th, site 3); this was not a typical work day and results were not used in the statistical analysis.

^BField limit of detection or Minimum Detectable Concentration (MDC) = (LOD of 5 µg/sample)/(volume in cubic meters)

^C12% silica content on Washington Street and 5.4% silica content on High School Road

The 20 full-shift PBZ air sampling results for the operator and ground worker using the Manufacturer B milling machine ranged from below the LOD to 13 µg/m³ with the exception of second day of sampling where the PBZ air sample for the ground worker was 24 µg/m³. The third day of sampling (site 3) resulted in non-detectable concentrations and was not used in the statistical analysis partly because the 6-hour shift included 3-hours of downtime and partly because the low milling speed and 1-inch (2.54 cm) removal depth are not typical of asphalt milling.

Table IV shows the means and upper 95% confidence limits for the two occupations and other summary statistics. The geometric mean respirable crystalline silica exposure for the operator was 4.2 µg/m³ with an upper 95% confidence limit for the arithmetic mean of 11.8 µg/m³. The geometric mean respirable crystalline silica exposure for the ground worker was 9.0 µg/m³ with an upper 95% confidence limit for the arithmetic mean of 29.8 µg/m³. Thus, the upper 95% confidence limits are statistically significantly less than the NIOSH REL of 50 µg/m³.

Table IV also shows the confidence of being less than the REL for 95% of the population of values from which manufacturer B data were drawn. For the operator, 95% of the population of silica measurements were less than the REL with more than 95% confidence, and for the ground worker the confidence is about 80%.

Investigation of Effect of Explanatory Variables

The seven explanatory variables (Table V) were manufacturer (A or B), average cut depth, average machine speed, whether work was day or night, average wind speed, average temperature, and % silica in bulk samples. Table V has 20 rows, since 11 days were sampled for manufacturer A and 9 days for manufacturer B (after omitting one day). The dependent variable was the log scale average of the operator and ground worker PBZ samples for each day. The explanatory variables were used in multiple regression.

Table V.

Explanatory Variables

Study	Site	Day at site	Average Cut Depth (inches)	Average Machine Speed (ft/minute)	Day or night	Average Wind Speed(m/s)	Temperature Average (degrees C)	%Silica in Bulk Samples
A	1	1	2	40	d	1.5	26	18
A	1	2	2	40	d	4.9	25	19
A	1	3	2	40	d	3.6	28	20
A	2	1	1.5	60	d	1.2	21	23
A	2	2	1.5	40	d	1.2	21	23
A	3	1	2	45	n	3.1	23	9.6
A	3	2	2	45	n	3.7	27	15
A	3	3	2	40	n	1.8	24	17
A	4	1	6	40	d	1.5	24	16
A	4	2	4	40	d	2.5	22	7
A	4	3	4	50	d	1	21	10
B	1	1	9	30	d	4.8	16	5
B	2	1	11	20	d	2.7	11	8.7
B	4+	1	1.5	80	d	0.9	11	11.4
B	5	1	9	20	n	1	19	5.5
B	5	2	9	20	n	1.3	16	6.3
B	6	1	1.5	80	d	4	16	5.8
B	7	1	1.5	60	n	3.3	4	8.1
B	7	2	1.5	60	n	4.3	12	8.9
B	7	3	1.5	60	n	2.4	5	9.1

^*Average of the operator and ground worker samples

⁺Site 3 data were not used because work conditions were not typical.

The reduced statistical model excluded average temperature and average wind speed. Except for the variable that allowed for difference in study means, the estimated regression coefficients are shown in Table VI. The variable concerning study means, though needed in the model, was not statistically significant at the 10% level, whereas all other included variables were. For the three continuous variables, the estimated 75^th to 25^th percentile factor effect is between 2 and 3. Also, day milling leads to about half the PBZ levels as night milling. The data set is small and these results need verification from additional work.

DISCUSSION

Both manufacturers designed their dust controls using recommendations provided as part of the Silica/Asphalt Milling Machine Partnership. These recommendations are documented in the NIOSH Publication No. 2015-105.⁽⁴⁰⁾ A goal of the Partnership was to statistically compare the air sampling results during field testing with the NIOSH REL so that milling machine manufacturers would have statistical confidence that the dust controls would protect workers when implemented on all new asphalt milling machines of similar model type for that manufacturer. Appendix C of NIOSH Publication No. 2015-105 provides the full development of the statistical method and Appendix B of NIOSH Publication 2015-105 provides site selection considerations for field testing of engineering controls on asphalt milling machines before those machines are implemented across a manufacturer’s entire fleet. Site selection recommendations such as weather, silica content, downtime, and shift length were met for all test days except for day 3 of Manufacturer B. Each manufacturer chose to have sampling conducted for more days and sites than the minimum recommend by NIOSH Publication No. 2015-105.⁽⁴⁰⁾

The evaluated ventilation controls on both asphalt pavement milling machines performed well at capturing dust generated in the drum housing and releasing the dust at the top of the secondary conveyor at a location away from the workers. The ventilation controls were effective at maintaining PBZ air samples to levels well below the NIOSH REL of 50 µg/m³ for respirable crystalline silica. However, three out of the 21 evaluated shifts included times when the crew was milling into the wind while dust released at the outlet of the ventilation control blew back toward the operator and ground worker. It was estimated that dust blew back toward the machines for less than one hour during two of the shifts and approximately three hours for another shift. Dust blowing back toward the operator and ground worker for a portion of the three shifts did not appear to influence exposures, but was annoying for the operator and the ground worker. When milling into the wind, the operators decided to temporarily turn off the ventilation control until wind conditions changed. Turning off the ventilation control when milling into the wind worked well for the limited amount of time that unfavorable wind conditions were present during three of the 21 days of sampling and did not appear to influence exposures. It is possible that future milling studies could occur during longer periods of unfavorable wind conditions or the operator could forget to turn the ventilation control back on after short periods of unfavorable wind conditions. After this study, both milling machine manufacturers made plans to equip their ventilation controls with technology to automatically turn the dust control back on after it has been off for 1 hour.

CONCLUSIONS

The 42 PBZ air samples were collected during 21 days at 11 sites and included diversity in typical asphalt pavement milling conditions. The roads being milled included rural highways, city streets, a parking lot, and a major freeway. Milling depths ranged from typical one to three inch (2.54 to 7.62 cm) mill and fill removals as well as full-depth removals of up to eleven inches (28 cm) of recycled asphalt pavement. The selected sites included diversity in day and night milling. Weather conditions included a wide range of ambient temperatures and wind conditions. The study included diversity of shift lengths with the majority of shifts lasting between 8 to 12 hours. Respirable crystalline silica PBZ air samples were below 50 µg/m³ for the operator and ground worker over all 21 evaluated shifts covering a wide range of typical conditions for asphalt pavement milling. For each occupation and each manufacturer the 95% upper confidence limits for the arithmetic mean exposure were less than 50 µg/m³. This suggests that for the population from which the sites were drawn, there is high probability that the average exposure of workers is less than 50 µg/m³.

The results indicate that the evaluated ventilation dust controls in combination with water-sprays are capable of controlling occupational exposures to respirable crystalline silica during a wide range of typical highway construction jobs using asphalt pavement milling machines. Following this testing, the manufacturers of both asphalt pavement milling machines made plans to include the ventilation as a standard feature on all new half-lane and larger asphalt pavement milling machines. Additional asphalt milling machine manufacturers who are Silica/Asphalt Pavement Milling Machine Partnership members also made commitments to perform similar field testing and implement silica dust controls on their milling machines.