Diesel Exhaust Exposure in the Canadian Railroad Work Environment

ABSTRACT

An investigation of occupational exposure to diesel exhaust, in terms of elemental carbon, was conducted as part of a feasibility study in the Canadian railroad industry. Both personal and area samples were collected from three major operating divisions of the railways: mechanical service, transportation, and engineering. A total of 255 elemental carbon samples have been described. The results show that all but six elemental carbon concentrations, expressed as size-selective respirable air samples taken using a 10 mm nylon cyclone, are well below the 2001 proposed American Conference of Governmental Industrial Hygienists’ (ACGIH) threshold limit value (TLV) of 20 μg/m3.The concentration of diesel exhaust, expressed as elemental carbon, in the railroad industry is much lower than that in some other major industries such as mining and forklift truck operations. If the TLV is to be applicable to a broad range of workplace settings such as railroad, construction, and mining, the use of a TLV that is based on an elemental carbon measurement of size selective respirable samples, as recommended in the 2001 ACGIH proposal, would appear to be the most valid strategy for control of exposure to diesel exhaust.



MATERIALS AND METHODS

Workplace Description

The two major participants were CN, with approximately 21,000 employees, and CPR, with approximately 18,000 employees, in Canada and the United States. Other participants in the exposure monitoring phase included Ontario Southland Railways, Goderich and Exeter Railway (Rail Link Canada), Waterloo–St. Jacobs Railway, and the Quebec-Gatineaux Railroad/Chemin de Fer Quebec-Gatineaux.

The three main operating divisions at the railway companies are mechanical-services, transportation, and engineering. All other divisions including management, human resources, and hospitality services, were not included in the study.

Mechanical Services


These employees are responsible for the maintenance and re- pair of rolling-stock (railways). They work in diesel shops where locomotives are repaired, and car shops where freight cars are repaired. Job occupations found in the mechanical services divi- sion include pipe fitters, machinists, welders, mechanics, electricians, boiler-makers, carpenters, laborers, car men, hostlers, and engine attendants.

Transportation

Employees operate the trains (also know as the Running Trades), or perform traffic control and transportation planning. Occupations include engineers/train drivers, conductors, foremen, trainmen, and helpers. Supervisory staff are involved in quality control, planning, and training. Other clerical occupations in this division include traffic controllers, station masters, and yard masters. Transportation employees work in yards, on main lines, in traffic control centers, and stations. Before the 1990s, train crews may have included brakemen, trainmen, and firemen (coal stokers). However, in the last ten years this has been reduced to just two employees: conductors and engineers. Yard crews can include an engineer, foreman, and helper or brake person.

Engineering Services

This group builds, maintains, and repairs property, buildings, bridges, and track using specialized equipment and other construction vehicles. Engineering services is also responsible for constructing non-railway buildings such as storage sheds and bunk houses. Occupations include field workers, gang laborers, equipment operators, signal maintainers, and track maintainers.

Supervisory staff include track maintenance foremen, production supervisors, and construction engineers. Workers are organized into crews, such as rail gangs, tie gangs, welding gangs, and bridge and structures gangs. They can have anywhere from 4–30 workers and 1–12 pieces of equipment. Smaller section crews or signal maintenance crews may have as little as 2 to 4 employees and a high rail truck.

Sampling and Analytical Methods

The methods of sampling and analysis were essentially those that have been applied in our previous railroad study and have been described in detail elsewhere. The EC samples were taken during the feasibility study using 37-mm-diameter open-face cassettes with precleaned 37 mm quartz fiber filters at the flow rate of 2.0–4.0 liters per minute (Lpm). In the earlier study, the results of which have been included in this article, the air samples were taken using a 10 mm nylon cyclone with precleaned 37 mm quartz fiber filter at a flow rate of 1.7 Lpm for the respirable fraction, and a small number using a Marple two-stage impactor with a final cutpoint of 0.5μm. The samples from the feasibility study were analyzed for elemental carbon and organic carbon by two laboratories: CANMET Natural Resources Canada, Sudbury, Ontario, and DATA CHEM Laboratory in Salt Lake City. The bulk of the samples were analyzed by the CANMET lab. Other elemental carbon results included from previous studies were those analyzed by the Sunset Laboratory,Forest Grove, Oregon,(10)and the CANMET Laboratory. All samples were analyzed using NIOSH Method 5040 Elemental Carbon Method. 

One-hundred sixty elemental carbon total (EC-T) samples were taken during the feasibility study and an additional 100 elemental carbon samples comprised of both EC-T (using a 37 mm diameter open-face cassette at 2.0–4.0 Lpm), EC-C (using a 37 mm diameter filter with a 10 mm nylon cyclone at 1.7 Lpm for the respirable fraction), and EC-M (elemental carbon–Marple, using a 37 mm filter with a two-stage impactor called a Marple sampler at 2 Lpm) were assembled from a previous publication. Fifty-nine sample results were extracted from company files. This provided a total of 319 elemental carbon samples for analysis.
The vast majority of the samples were taken as long-term samples (6–8 hrs). Six samples were spoiled due to errors in sampling rate, equipment failure, or tampering. Eighteen samples were excluded as they were experimental by nature or were not representative of work conditions. In addition, because it was desirous to review shift-based exposures and not task-based exposures, a further 28 samples were excluded because conclusions regarding time-weighted average exposures could not be interpreted fairly from the results. Finally, 12 sample results taken from company reports were excluded because necessary information such as sampling duration, flow rate, and so on, could not be found. None of the excluded samples had concentrations that exceeded 20μg/m3 and only 4 were above 10μg/m3. A total of 64 samples were thus excluded, leaving 255 elemental carbon samples for analysis. All EC-T and EC-M values have been converted to equivalent EC-C values using the relationship obtained in the side-by-side sampling program of the earlier railroad study as: EC-C=0.84×EC-T and EC-C=1.33×EC-M. All values reported in this paper can then be compared to the most recent TLV recommendation of 20μg/m3as EC-C.

Quality Control

In an earlier study, the validity of the analytical method was evaluated by using spike QA samples because at that time only one laboratory was performing the analysis. During the feasibility study, we relied on the fact that the two laboratories that analyzed the samples were among 11 participants in an inter-laboratory QA program for elemental carbon,(15)and they provided proof of their proficiency in that program. Additional standard quality control procedures included use of blanks and replicates. Ten paired samples taken during the feasibility study were analyzed by both laboratories. In general, the results were similar. Also, duplicate analysis on a total of seven filters was performed by analyzing two different wedges from the same filter. These were found to give similar results. The weights ofEC detected in all these QA samples were very low and close to the detection limit. 

Statistical Analysis

Each elemental carbon sample has been expressed in terms of EC-C. As stated earlier, EC-T and EC-M samples were converted using the relationship observed in the earlier study. Every sample was classified according to the type of sample(personal or area), location (turnaround, heavy repairs, on board locomotive), by various trades and ventilation situation. Many of the samples contained concentrations of elemental carbon below the detection limit (BDL) of the analytical method. It was hypothesized that the sampling results would follow a logarithmic normal distribution. The Maximum Likelihood Estimation (MLE) statistical method has been shown to produce the best estimate of both the mean and standard deviation of an industrial hygiene data set containing values below the limit of detection. The data were summarized by the geometric mean and standard deviation. These were computed by the MLE method, which is implemented in an Excel spreadsheet.

RESULTS

The results of 255 samples of all types are given in Table I. In Table II, personal samples are summarized, and in Table III, area samples are shown. In Table II, 23 area samples (also listed in Table III for leading locomotive) have been used as surrogate personal exposure samples for engineers/train drivers. Seven of the same 23 samples have also been included as surrogate personal exposure samples for conductors/trainmen. Furthermore, 7 of the total 62 personal samples listed in Table I from both turnaround and heavy repairs could not be included in Table II because their jobs could not be identified from the company reports. In Table IV, EC results from other studies are tabulated for comparison, and in Table V, the ratio of EC to TC representing various work environments has been shown where the EC is reported as percent of TC
