RSP concentrations for the 13 non-smoking subjects ranged from < 12.5 to 47.9 µg/m3, with a mean concentration of 2 3.6 µg/m3 (median 22.8 µg/m3). Mean RSP levels were slightly higher on April 13th (26.4 µg/m3) compared to the second day of measurement (21.2 µg/m3).

The UVPM analysis showed concentrations ranging between <2.4 and 32.6 µg/m (mean 16.9 µg/m3; median 12.9 µg/m3). Analysis of the ratios between RSP and UVPM concentrations suggests that, on average, between 60 and 75 % of the determined RSP mass was associated with combustion-related processes, including ETS.

FPM concentrations ranged from 1.0 to 12.2 µg/m3 (mean 4.6 µg/m3, median 4.3 µg/m3). The ratio between RSP and FPM estimated that typically between 20 and 50% of the RSP mass was associated with ETS and other combustion-related processes.

Comparison of the FPM and UVPM results shows substantially lower FPM concentrations, a finding consistent with Ogden et al. [26] who showed that UVPM overestimated ETS-related particulate by up to 30% in an experimental chamber and that FPM provided a more accurate estimate of ETS-related particulate matter.

Mean solanesol concentrations were 0.13 µg/m3 on April l 3th and 0.17 ,µg/m3 on April l 4th, with a range of concentrations from <0.06 to 0.49 µg/m3. The weight ratios for solanesol: RSP ranged from.0.33 to 1.46%, with a mean weight ratio of 0.61 %. These results are similar to those reported in `real world' work environments, where a less than 1% weight ratio for solanesol: RSP was reported [27). The solanesol concentrations suggest that 35-55% of the RSP mass may be attributable to ETS, an estimate similar to that determined from the FPM analysis.

Nicotine concentrations ranged from 0.3 to 4.7 µg/m3 (mean 2.0 µg/m3, median 1.6 µg/m3). 3-EP concentrations ranged from 0.5 to 1.5 µg/m3, with a mean concentration of 0.9 µg/m3.

TVOC concentrations for 12 of the 13 subjects who had personal monitors, ranged from < 1.3 to 26.2 µg/m3.

However, a concentration of 502.6 µg/m3 was recorded for 1 subject. As the tracers of particle-phase and vapour-phase ETS exposure for the subject were not elevated above those determined for the others, the elevated TVOC concentration does not appear to be related to ETS exposure. Chromatographic analysis of the TVOC samples Identified low-molecular-weight compounds in the C3-C8 range, including toluene and xylene, and higher molecular-weight compounds in the C9-C 11 range. These higher-molecular-weight compounds were qualitatively identified as common indoor VOCs such as terpenes, d-limonene and a-pinene, which are typically found in inks, adhesives, air fresheners, and cleaning products.

The salivary cotinine analysis showed cotinine levels of between < 1.0 and 5.6 ng/ml for all except 3 of the subjects: 2 subjects showed levels of 32.9 and 79.6 ng/ml, respectively; and one sample was spoiled in transit to the laboratory. Previous research has shown that salivary cotinine levels of less than 6 ng/ml are typical of non-smokers and has also suggested a "cut-off" point between current smokers and non-smokers as ranging between 20 and 100 ng/ml [17-19]. The higher cotinine concentrations for the two subjects, while elevated above the others, were both within this range. Review of the measured particle-phase and vapour-phase tracers for the two subjects did not show substantial differences in exposure compared to the other subjects. Consequently, the 2 subjects were assumed not to have smoked at work. A possible explanation for the higher cotinine concentrations for the 2 subjects may be dietary intake of nicotine-containing foods, such as leafy vegetables, potatoes and tomatoes.

The impact of dietary nicotine on cotinine levels has been demonstrated by several investigators [28, 29].

Smoking Conditions

An estimate of the prevalence and frequency of smoking can be determined from (a) the subject activity logs and (b) the original selection procedures for the subjects.

The mean reported smoking frequency was 1.4 cigarettes/h, with a range from 0 to 5 cigarettes/h. From the original survey to identify subjects, the proportion of non-smokers in building 1 was estimated at 22%. The estimated smoking rate of 1.4 cigarettes/h and smoking frequency of 22% is slightly higher than the 1992 US national average smoking rate in office workplaces of 1.21 cigarettes/h and 20.2% smokers reported in the National Health Interview Survey [30].

[...]

Discussion

Comparison of Personal and Fixed-Location Monitoring Methods to assess ETS exposure

The findings from this research provide new information about the comparability of data gathered through personal monitoring and fixed-site measurements. Research on non-smokers' exposure to ETS in office workplaces has rarely been conducted by personal monitoring.

ETS exposure has predominantly been assessed through fixed-location monitoring, due to the non-obtrusive e methods creating minimal disruptions to normal workplace activities. However, the question may be posed as to whether fixed monitoring data are representative of an individual non-smoker's exposure to ETS. Comparison of the personal and fixed location data in the study buildings provides some insight into this question.

The results from building 1 show statistical agreement in the measured exposure to particle-phase and vapour-phase constituents of ETS in the personal and fixed-location monitoring. For building 2, the results of the vapour-phase analyses (nicotine and 3-EP) also showed similar concentrations determined in both the personal and fixed-location monitoring. However, the particle-phase indicators were significantly different, due to the `cluster' of non-detected levels from the fixed-monitoring locations.

Despite this anomaly, the overall results from the two study buildings indicate that fixed-location monitoring provides a close approximation to an individual's exposure to ETS, as determined through personal monitoring.

This finding suggests that the `real world' data obtained by past researchers, primarily through fixed-location monitoring, is appropriate for estimating occupant exposure to ETS.

Comparison of the Study Findings with Other `Real World' Data

The data from the two buildings in Richmond add to the growing archive of objective data from the measurement of ETS-related constituents in office buildings. The most widely measured indicators of ETS exposure from the literature have been total RSP and nicotine [6]. Comparison shows that the measured RSP and nicotine concentrations of the vapour-phase and particle-phase indicators of ETS exposure from both the personal and fixed-location monitoring are generally consistent with the levels reported in the research literature.

From fixed-location monitoring in mechanically ventilated off ce environments in which smoking is permitted, research has shown RSP levels typically to range from 20 to 80 µg/m3 [6, 7, 33, 34]. Nicotine concentrations typically have ranged between 1 and 6 µg/m3 [6, 7, 35].

Personal exposure monitoring studies to assess ETS exposure in office environments are less common. Coultas et al. [36] reported a mean RSP concentration of 56.7 µg/m3 and a mean nicotine level of 4.8 µg/m3, for 5 subjects in office buildings in I 'e Mexico. Muramatsu et al. [37] reported nicotine levels ranging from 5 to 19 µg/m3 in Japanese office buildings. While the measured concentrations in these studies are higher than those determined in the Richmond study buildings, no information was provided regarding either (a) ventilation conditions or (b) smoking prevalence.

Some researchers have suggested a consistent ratio between measured RSP and nicotine concentrations, and have applied this consistent ratio to predict nicotine from reported RSP concentrations [38]. The results from the study presented here question the conclusion of a consistent ratio between nicotine and RSP levels. Within the overall data set, RSP:nicotine ratios varied from approximately 4:1 to 75:1. The inconsistency of RSP:nicotine ratios has also been reported in a survey of the research literature by Guerin et al. [6], who observed variable ratios from 4:1 to 100: 1.

The present study also provides valuable information about tracers of ETS exposure other than total RSP and nicotine. To date, limited `real world' data have been reported in the research literature for (a) the three methods to estimate the contribution of ETS to the total RSP mass (UVPM, FPM and solanesol), or (b) the use of 3-EP as a tracer for exposure to vapour-phase constituents of ETS.

The UVPM, FPM and solanesol analyses show a high degree of variability associated with the analytical methods. In general, the highest estimates of the proportions of total RSP associated with ETS were obtained from the UVPM analysis. Lower (and less variable) estimates were obtained from the FPM and solanesol analyses. In general, the results are consistent with the conclusions from experimental chamber research that UVPM concentrations tend to be higher than FPM levels, and that both FPM and solanesol may provide a better estimate of ETS-related particulate than UVPM [26].

. The measurable levels of 3-EP in the study buildings suggest that it may provide an alternative tracer to nicotine under field conditions. Questions have been raised regarding the appropriateness of nicotine as a tracer for exposure to vapour-phase constituents as it displays non-typical and unpredictable decay kinetics [3,12]. 3-EP has more predictable decay kinetics and may be more typical of other vapour-phase constituents. The results show that while 3-EP was less abundant than nicotine, it was clearly quantifiable under `real world' conditions.

Regulatory Implications of the Research

The research also provides an important case study of non-smokers' exposure to ETS in an of office environment supplied with outside air ventilation rates in accordance with current ventilation standards, with the results providing information on the effectiveness of dilution ventilation as a regulatory option to control ETS in the workplace.

The HVAC performance assessments determined that the HVAC systems serving both buildings were providing outside air to the occupied space at ventilation rates nominally in accordance with the ASHRAE/ANSI ventilation standard 62-1989. Smoking conditions in both study buildings were slightly greater than `average' conditions in US office workplaces [30]. However, they are representative of `moderate' amounts of smoking assumed for the ventilation rates described in ASHRAE/ANSI Standard 62-1989.

Given HVAC system performance in accordance with ASHRAE/ANSI Standard 62-1989 and a `moderate' amount of smoking, the concentrations of the various tracers of ETS exposure measured in the two study buildings demonstrate that ETS-related constituent levels may be effectively controlled to low concentrations through general dilution ventilation. This finding suggests that dilution ventilation can provide an appropriate option to accommodate smoking in the workplace, a view contrary to currently proposed regulations for US workplaces. The US Occupational Safety and Health Administration (OSHA) has proposed sweeping indoor air quality regulations which reject the traditional engineering .practice of dilution ventilation for the control of ETS and other indoor sources [39]. In the rationale for the proposed OSHA regulations, no research data are cited to justify the rejection of dilution ventilation. The results of the research presented here demonstrate that dilution ventilation can provide an effective means of controlling non-smokers' exposure to ETS and providing acceptable indoor air quality.