Diesel exhaust exposures in port workers

ABSTRACT 

Exposure to diesel engine exhaust has been linked to increased cancer risk and cardiopulmonary dis-eases. Diesel exhaust is a complex mixture of chemical substances, including a particulate fraction mainly composed of ultrafine particles, resulting from the incomplete combustion of fuel. Diesel trucks are known to be an important source of diesel-related air pollution, and areas with heavy truck traffic are associated with higher air pollution levels and increased public health problems.

Several indicators have been proposed as surrogates for estimating exposures to diesel exhaust but very few studies have focused specifically on monitoring the ultrafine fraction through the measurement of particle number concentrations. The aim of this study is to assess occupational exposures of gate controllers at the port of Montreal, Canada, to diesel engine emissions from container trucks by measuring several surrogates through a multimetric approach which includes the assessment of both mass and number concentrations and the use of direct reading devices. A 10-day measurement campaign was carried out at two terminal checkpoints at the port of Montreal. Respirable elemental and organic carbon, PM1, PM2.5, PMresp(PM4), PM10, PMtot(inhalable fraction), particle number concentrations, particle size distributions, and gas concentrations (NO2, NO, CO) were monitored.

Gate controllers were exposed to concentrations of contaminants associated with diesel engine exhaust (elemental carbon GM=1.6μg/m3; GSD=1.6) well below recommended occupational expo-sure limits. Average daily particle number concentrations ranged from 16,544–67,314 particles/cm³(GM=32,710 particles/cm³; GSD=1.6). Significant Pearson correlation coefficients were found between daily elemental carbon, PM fractions and particle number concentrations, as well as between total carbon, PM fractions and particle number concentrations. Significant correlation coefficients were found between particle number concentrations and the number of trucks and wind speed (R2=0.432;p<0.01). The presence of trucks with cooling systems and older trucks with older exhaust systems was associated with peak concentrations on the direct reading instruments. The results highlight the relevance of direct reading instruments in helping to identify sources of exposure and suggest that monitoring particle number concentrations improves understanding of workers’ exposures to diesel exhaust.This study, by quantifying workers’ exposure levels through a multimetric approach, contributes to thef urther understanding of occupational exposures to diesel engine exhaust.

METHODS

Work description of gate controllers at the port of Montreal

Each port terminal has a gate, or checkpoint, where incoming and outgoing truck traffic is controlled. Three controllers work in rotation at the gate over the course of the day so that there are two controllers at the gate at all times. Controllers work for a total of 5 hr and 20 min over each 8-hr shift; each working period of 2 hr and 40 min spent at the gate is followed by a 1 hr and 20 min break, spent inside the administrative building. The work is carried out in a covered but open environment, similar to a road toll collection booth. The task of the gate controllers consists of inspecting the contents and security seals of containers found on container trucks coming through the gates of the port terminals. If a container is declared empty the controller must examine it according to the port's regulations.

If a container is declared non-empty the controller must make sure that the seal is intact. Time spent inspecting each truck does not exceed 5 min and the controllers rarely take breaks between inspections. When a truck arrives at the gate the driver stops the engine and does not start it again until after the inspection is finished. Trucks equipped with a cooling system have an auxiliary diesel engine that is not stopped during inspection. The controllers wear personal protective equipment comprising of a helmet, safety boots, and an orange vest; controllers may also wear hearing protection in order to minimize noise levels. No respiratory protection is used.

Sampling design

A 10-day measurement campaign was carried out during the month of October 2013 at two terminal checkpoints in the port of Montreal, the first five days at Terminal A and the last five days at Terminal B. A total of six gate controllers were recruited (three per terminal). Workers were recruited through their union as volunteers, following study approval by the ethics committee of the Université de Montréal. A consent form was signed by all of the workers involved. Personal samples (EC, OC, gases) were taken on the controllers (three each day) over their work shift, excluding break periods. Area measurements of PM mass concentrations, particle number concentrations (PNC), and particle size distributions were carried out at both terminals over a minimum period of 6 hr per day.


Particulate measurements

In accordance with the NIOSH Method 5040,[] respirable EC and OC were collected in the workers' breathing zone on 25-mm quartz-fiber filters enclosed in cassettes mounted on 25-mm SKC aluminum cyclones connected via Tygon R tubing to an Airlite pump operated at a flow rate of 2.5 l/min ( Figure 1). Pumps were calibrated before and after the monitoring using a DryCal (DC Lite model, Bios International Corporation), and the pre- and post-sampling flow rates were averaged. A flow rate deviation of up to 5% was considered acceptable. TC concentrations were calculated by adding OC concentrations provided by the laboratory (data not shown) to the EC concentrations. Blank corrections were performed on EC and OC values according to the NIOSH 5040 method. Samples were analyzed by Galson Laboratories (East Syracuse, NY). 

PM mass concentrations (Dust-Trak DRX, TSI Inc., Shoreview, MN), PNC (P-Trak, TSI Inc., Shoreview, MN) and particle size distributions (Engine Exhaust Particle Sizer EEPS, TSI Inc., Shoreview, MN) were monitored using direct reading instruments (Table 1). The direct reading instruments were equipped with Tygon tubing and were placed between incoming and outgoing traffic lanes in order to sample as close as possible to the gate controllers (Figure 1). All devices were calibrated prior to the 10-day sampling period according to the manufacturer's requirements, and zeroing was performed each day before the monitoring. To improve the relative accuracy between the five mass channels of the Dust-Trak DRX, a user calibration was performed on-site prior to the 10-day sampling period using a 2.5 µm inlet impactor.

Gas measurements

Gases (NO2, NO, CO) were sampled with BW Technologies personal dosimeters placed in the workers' breathing zone, unless the workers preferred to attach them to their pocket (Figure 1). Calibration of all dosimeters was performed before the sampling campaign and on-site bump tests were performed each morning before sampling.

Exposure determinants

Average daily wind speed was estimated throughout the sampling period via the port's nearest on-site meteorological station, and the number of trucks crossing the gates was obtained throughout the sampling period using available data collected by the port, as a part of the regular management of port activities. In addition to these two quantitative determinants of exposure, other field observations relative to working conditions (such as time spent at the gate vs on breaks) and truck characteristics (such as presence of a diesel-fueled cooling system) were noted throughout the sampling campaign.

Statistical analyses

Descriptive statistics were calculated using IHSTAT (American Industrial Hygiene Association - Exposure Assessment Strategies Committee) and treatment of non-detects was performed using NDexpo (School of Public Health, Université de Monteal) . NDexpo is a Web application that implements the robust regression on order statistics censored data treatment approach for dealing with non-detects.

Analysis of variance (ANOVA) was performed on log-transformed PNC, PM1, PM2.5, PMresp, PM10, PMtot, EC, and TC exposure values from both terminals in order to evaluate the homogeneity of the exposure groups (i.e., gate controllers). Pearson correlation coeficients were calculated between the diefrent indicators of exposure and determinants of exposure. ANOVA and calculations of Pearson correlation coecfiients were performed using SPSS (V22, IBM Corporation, Armonk, NY).

RESULTS

Particulate measurements

No significant difference between levels of exposure at the two terminals was observed, confirming that the gate controllers represent a single similar exposure group. Table 2 presents the arithmetic mean (AM), geometric mean (GM) and geometric standard deviation (GSD) of EC, TC, PNC, and PM daily concentrations measured during the 10 days of sampling. GSD for all indicators were lower than or equal to 2. EC concentrations ranged from 0.7- 4 µg/m3 (AM = 1.8 µg/m3; GM = 1.6 g/m3). The highest value obtained was thus 25 times lower than the lowest recommended exposure limit of 100 µg/m3 proposed for an 8 hr/da- 40 hr/week exposure.[Daily PNC measured at both sites ranged from 16,544 to more than 67,000 particles/cm3 (AM = 36,381 particles/cm3; GM = 32,710 particles/cm3). PMresp concentrations ranged from 13-94 µg/m3 (AM = 40.4 µg/m3; GM = 33.8 µg/.

PM mass concentration distribution, calculated using the Dust-Trak, indicated that 85% of the particles were within the respirable fraction and that 79% of those were smaller than 1µm. TC:EC ratios ranged from -30  with an average of 12.

PNC peaks of up to 500,000 particles/cm3 and a PM mass concentration peak of 4 mg/m3 (Figure 2) recorded with the P-Trak and Dust-Trak, respectively, were associated with the observation of the crossing of container trucks equipped with cooling systems.

The P-Trak and EEPS profiles in Figure 3 indicate a PNC peak of 400,000 particles/cm3 and of more than 10,000,000 particles/cm3, respectively. The peak was associated with the observation of the crossing of an older truck with an older exhaust system (pipe attached to the chassis horizontally). As seen in Figure 3 the PTrak and EEPS profiles are of similar form but different magnitude, with the average concentration measured with the P-Trak (67,000 particles/cm³) being two orders of magnitude smaller than the average concentration measured with the EEPS (1,700,000 particles/cm³).

The particle size distribution, based on data from the EEPS, confirms that the large majority of DE particles are on the nanometric scale since more than 99% of particles measured were smaller than 100 nm and that the main particle size fraction was between 20-40nm.

Significant Pearson correlation coefficients were found between daily EC, PM fractions and PNC, as well as between TC, PM fractions and PNC (Table 3). However, EC and TC were weakly correlated (r = 0.34). 

Gas measurements

Controllers were exposed to gas levels lower than the limit of detection of the gas monitors (i.e., < 0.1 ppm), well below the lowest recommended limit values proposed in Sweden to regulate DE exposures, i.e., 20 ppm and 1 ppm for CO and NO2, respectively.

Exposure determinants the number of trucks, and wind speed. Significant correlation coefficients were found between PNC and number of trucks (PCC = 0.621) and between PNC and wind speed (PCC = -0.537). By combining both factors using the ANOVA regression method, 43% of the total variance of the PNC can be explained (R2 = 0.43; p < 0.01). Figure 4 shows the correlation between PNC and number of trucks and between PNC and wind speed.

