In 1995, the ACGIH proposed that the particulate fraction of diesel exhaust emissions (also known as "diesel particulate matter" or DPM) be categorized as an A2 level carcinogen (suspected human carcinogen) and suggested a TLV - TWA of 150 micrograms/m³ (mg/m³) for particulates with an aerodynamic diameter of less than 1 micrometre. The document (Ref. 1) supporting the proposed TLV did not specify what measuring method should be used. Below are the three methods that were discussed:

• gravimetric analysis of the respirable fraction,
• elemental carbon measurement using thermo-optical analysis of the respirable fraction, and
• analysis of respirable combustible dust (RCD).

In 1999, the ACGIH amended its previous proposal and lowered the proposed TLV to 50 mg/m³, while continuing to categorize the product as an A2 carcinogen. The document that accompanied the amendment (Ref. 2) contained a somewhat more detailed discussion of measuring methods and a table comparing the characteristics of the different methods used to measure diesel exhaust particulates. The data in this table clearly established that DPM measurement results vary by a certain number of orders of magnitude, depending on the method used. However, the document did not clearly establish which method should be used to compare the measurements to the proposed TLV. Review of the data available at the time seems to show that elemental carbon measurement was the most appropriate method. The main problem with all the methods was the lack of selectivity for diesel particulate matter. In other words, they did not distinguish diesel particulate matter from any other dust present in the air sample. Indeed, as shown by a number of articles in industrial hygiene journals (Ref. 3 and 4), most of the hygienists measuring DPM assumed that TLV should be based on elemental carbon analysis.

I recently learned that the ACGIH has again changed the proposed TLV for DPM. It is now proposing a threshold limit of 20 mg/m³ in the form of elemental carbon. The most recent document (Ref. 5) states that the TLV proposed earlier (50 mg/m³) in fact applied to total DPM, which includes both organic and elemental carbon. However, organic carbon is not a substance particular to DPM (there are many other sources of organic carbon) and using it a marker (or as a marker component) for DPM could produce misleading results. Studies by NIOSH have clearly demonstrated that elemental carbon is a substance fairly characteristic of DPM; in other words, in most workplaces where workers are exposed to diesel exhaust, there are few other sources of elemental carbon. It has also been shown that elemental carbon constitutes about 50% of total carbon in DPM. The ACGIH has thus established, based on the same method of risk analysis used in the earlier documents, a proposed TLV of 20 mg/m³ in the form of elemental carbon.

As you know, we took measurements on board locomotives in the winter of 1999 and the summer of 2000 to determine the exposure of engine crew members to diesel exhaust emissions and noise. The evaluation of exposure to diesel exhaust emissions included measurements of elemental carbon concentrations in the cabs of leading and trailing locomotives. The two reports we delivered in January 2001 (separate reports for the two companies that participated in the study) compare the concentrations that were measured with the TLV of 50 mg/m³ proposed in 1999. The most recent amendment to the value proposed by the ACGIH obviously led me to review the data from the measurements. The results of my review are given below.

A total of 48 series of measurements were taken in the course of the study: 24 under winter conditions when the cab windows were closed, and 24 under summer conditions when the cab windows were open. The measurements were taken in the lead locomotive and the first trailing locomotive. For the purposes of this review, we can reasonably omit all the measurements taken in winter and the ones taken in the lead locomotives in summer. Elemental carbon (EC) concentrations were measured for each route in three locations: on the engineer’s console, on the conductor’s console and on the fire wall (middle) at the back of the cab. The two attached tables show the measurements for which EC concentrations at any of the above locations exceeded 20 mg/m³ or 10 mg/m³. EC exceeded 20 mg/m³ for 8 of the 24 measurements taken in summer: 6 measurements taken on board Canadian National locomotives and 2 taken on CP Rail locomotives. EC concentrations exceeded 10 mg/m³ in 18 of 24 cases: 11 measurements taken on Canadian National locomotives and 7 on board CP Rail locomotives. The value of 10 mg/m³ was selected because it corresponds to 50% of the proposed TLV. It is considered an action level to be taken into account in hazard investigations, under paragraph 10.4(2)(h) of the Canada Occupational Safety and Health Regulations. I realize that these regulations do not apply on board trains, but the concept of "action level" is nonetheless a useful guideline for employers, indicating discrepancies between employee exposure values and the prescribed limits.