Applicability of Concentrations Obtained by Working Environment Measurement to Assessment of Personal Exposure Concentrations of Chemicals

Shinobu Yamamoto; Shuichiro Natsumeda; Kunio Hara; Satoru Yoshida; Haruhiko Sakurai; Masayoshi Ichiba

doi:10.1539/joh.12-0243-OA

Abstract

Objective: This study determined the applicability of Japanese working environment measurements to assessment of personal exposure concentrations of chemicals by comparing both levels of concentrations. Methods: The chemicals measured in this study comprised eight kinds of vaporous chemicals as well as two kinds of chemicals in dust. Personal exposure measurements, Japanese working environment measurements and spot sampling measurements were undertaken in 70 companies. Results: Personal exposure concentrations and the arithmetic mean value (E_A2) of the working environment measurement concentrations obtained according to the Japanese working environment control system had statistically positive correlations (r=0.732−0.893, p<0.01) after logarithmic transformation. The 5th to 95th percentile values of personal exposure concentrations divided by E_A2 ranged from 0.17 to 7.69 for vaporous chemicals and from 0.27 to 18.06 for dust. There was a relatively large difference between the personal exposure concentrations and the E_A2 obtained in weighing, forming and bonding use-processes. In such cases, the B-value measured in ten minutes in the Japanese working environment control system, which is almost the same as the spot measurement concentration in this study, is supposed to be substituted for the E_A2 value. Conclusions: Ten times the E_A2 of the working environment measurement concentrations, or ten times the B-value, obtained according to the Japanese working environment control system can be used to conservatively estimate the personal exposure concentrations in EU workplaces as well as in occupational exposure scenarios of the Regulation on Registration, Evaluation, Authorisation and Restriction of Chemicals.

(J Occup Health 2014; 56: 85-92)

Introduction

Control of chemical hazards and physical agents in the working environment is of prime importance for protecting workers from health impairments due to hazardous agents. The number of chemicals for which occupational exposure limits have been determined is at most 1,000 all over the world. In contrast, the numbers of chemicals registered with the Chemical Abstracts Service Registry and in accordance with the Industrial Safety and Health Law in Japan were over 130 million by the end of 2012¹⁾ and 62,000²⁾, respectively. Although the number of chemicals used in industries has been reported to be several tens of thousands³⁾, the hazardous properties of most chemicals have not been characterized. It appears almost impossible for regulatory agencies to assess the true risks of chemicals. Therefore, in 2006, the EU introduced the Regulation on Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH). REACH requests manufacturers and importers that produce or import more than or equal to 10 tons of chemicals per year to conduct risk management measures based on exposure scenarios themselves. Assessment of occupational exposure scenarios use actual measured exposure concentrations or estimated exposure concentrations of the targeted chemicals in each use-process category.

The Japanese legislative system for control of chemical hazards and physical agents in the workplace is different from that of the EU and the US. The Japanese system is exemplified by “workplace control”, which is a risk control system using concentrations obtained by working environment measurements as “area monitoring results”. The hazardousness of working environments contaminated with hazardous chemicals is assessed by measurement of airborne chemicals in the workplace. Chemical risk management in Japan is based on the idea of decreasing the risk by improving the working environment and then decreasing worker exposure to harmful chemicals.

Cherrie et al.⁴⁾ reviewed the studies on comparisons between personal exposure and workplace concentrations that were performed between 1992 and 2002^5-16). They concluded that the geometric mean of personal exposure concentrations divided by the corresponding data, obtained by static samples, ranged from 1.2 to 8.5.

Although many researchers have reported differences between personal sampling results and area monitoring results, the Japanese working environment control system results can be used as an alternative to personal exposure concentrations and can be useful in estimating personal exposure concentrations in both EU workplaces and the occupational exposure scenarios of REACH if there is a correlation between the two. This study determined the applicability of Japanese working environment measurements to assessment of personal exposure concentrations of chemicals by comparing working environment measurements of concentrations with personal exposure concentrations.

Methods

Measurements of ten chemicals in 70 companies

The surveys in this study were conducted as the “Survey of Actual Exposure Conditions”, which was established by the Ministry of Health, Labour and Welfare (MHLW) during 2009 and 2010. The chemicals measured comprised eight vaporous chemicals, (1) ethyl acrylate, (2) acetaldehyde, (3) ethylbenzene, (4) vinyl acetate, (5) 1,2-dibromoethane, (6) 2-chloro-1,3- butadiene, (7) propylene oxide and (8) 1,4-dichloro-2- butene, and two chemicals in dust, indium and cobalt, in accordance with the “Guidelines on Exposure Assessment”¹⁷⁾ of the MHLW. Personal exposure measurements, working environment measurements and measurements using spot sampling were undertaken in 70 companies, each of which either produced or handled 500 kg per year; the companies were selected through a type of control banding that indicated companies where high exposure concentrations were expected.

The Japanese legislative system for control of chemical hazards and physical agents in the workplace^{18, 19)}

The Japanese government has developed “an administrative control level” to indicate acceptable chemical concentration levels in the workplace. Approximately 100 administrative control levels were decided by examining acceptable Occupational Exposure Limit and Threshold Limit Values recommended by the Japan Society for Occupational Health (JSOH) and the American Conference of Governmental Industrial Hygienists (ACGIH). The concept of the “unit workplace” was introduced in order to clarify what should be controlled in the working environment. The unit workplace is defined as the area wherein the airborne concentrations are presumed not to fluctuate greatly or wherein the concentrations are presumed to be random with regard to each worker, even if the average airborne concentration fluctuates. A minimum of five sampling points are selected by random systematic sampling at even intervals in a unit workplace; sampling of airborne chemicals is conducted for a minimum of one hour. The Industrial Safety and Health Act requires companies to conduct no fewer than five measurements of the working environment on two different days. If it is difficult to conduct these measurements on two different days, then the working environment assessment standard, as provided for by an ordinance of the MHLW, regulates using a geometric standard deviation on the second day, i.e., a σ₂ of 1.95, which represents the 90th percentile of the standard deviation value obtained in the research on Japanese workplaces conducted by Koshi et al.¹⁸⁾ (Fig. 1).

Fig. 1.

Measurement points A (① to ⑨): Five or more measurement points are, in principle, located at the intersections between lengthwise and crosswise lines drawn at equal intervals of 6 m or less. This measurement is carried out to identify the 95th percentile value (E_A1) and arithmetic mean (E_A2) of the hazardous substance in air in terms of place and time. Measurement point Ⓑ: When work is carried out near the emission source of a hazardous substance, the concentration is measured at the place and time that the workers face the maximum exposure. ⌧: Emission source of the hazardous substance.

After the samples are analyzed and a logarithmic normal distribution of the airborne concentrations is estimated from the data, the 95th percentile value (E_A1) and the arithmetic mean (E_A2) are calculated. Adding to these values, the concentration-B-value-is obtained in ten-minute sampling if there are some spots where worker exposure seems considerably higher compared with other spots. The B-value is not always measured because airborne concentrations in some unit workplaces are relatively uniform. B-values were not measured in this study. The B-value was thought to be less than or equal to the highest value of the spot samples collected adjacent to workers during the use-processes performed for a short duration in this study because the B-value must be measured for ten minutes, even when the targeted use-process continues for less than ten minutes. When the sampling times for spot measurements are longer than ten minutes, the spot measurement concentration is supposed to be almost the same as the B-value²⁰⁾. When the sampling times of the spot measurements were longer than ten minutes owing to a longer duration of the targeted use-process, the spot samples in this study were collected in 20 minutes or less during the targeted use-process. Both the E_A1 and the E_A2 were compared with the administrative control level, and the B-value was compared with values that were 1.5 and 1.0 times the administrative control levels. A unit workplace in which both the E_A1 and B-value are less than the administrative control level is assessed as being in good condition, and unit workplace in which the E_A2 is more than the administrative control level or the B-value is more than 1.5 times the administrative control level is assessed as being in a condition that needs countermeasures to improve the working environment.

Collection and analysis of samples

The personal sampling time periods ranged from 8:00 AM to 5:00 PM during a full shift of approximately 8 hours. Sampling in the workplace was conducted within the time frame according to the “Japanese Working Environment Measurement Standards”. Spot measurements were conducted at a fixed point on the windward and leeward sides of the source neighborhood through a similar measurement method of personal exposure concentrations of use-process categories, which were likely to have high exposure concentrations (within 20 minutes).

Metals in dust collected on membrane filters at a rate of 1.0–10.0 l/minutes were extracted with acid solutions. An aliquot of supernatant was injected into an atomic absorption spectrometer or into an inductively coupled plasma mass spectrometer. The vaporous chemicals in passive and active samplers (flow rate of 0.1–1.0 l/min) were desorbed in desorption solution. An aliquot of supernatant was injected into either a gas chromatograph or a high-performance liquid chromatograph.

A total of 343 samples of personal concentrations, 75 samples from workplaces and 253 samples from spot measurements were collected. Since some sample results were below detection limits, their values were excluded. Table 1 briefly shows the characteristics of the investigated data: 35 samples of vapor data and 19 of dust data were used to estimate a working environment measurement. Similarly, 95 samples of vapor data and 57 of dust data were used to estimate the spot measurement.

Table 1. Brief characteristics of data

	Chemicals	Boiling point (°C)	Companies	Personal exposure monitoring	Working environment measurements	Spot measurements
Vapor	Ethyl acrylate	99	7	18	3	16
	Acetaldehyde	20	4	14	3	19
	Ethylbenzene	136	14	81	19	34
	Vinyl acetate	73	12	74	14	59
	1,2-dibromoethane	132	1	2	0	1
	2-chloro-1,3-butadiene	59	1	10	2	11
	Propylene oxide	34	6	15	2	11
	1,4-dichloro-2-butene	159	1	2	2	3
Dust	Indium	2,000	8	59	16	59
Dust	Cobalt	2,927	14	68	14	40
	Total		68	343	75	253

Statistical tests were performed using a 4-Step Excel statistical statcel 2²¹⁾. The personal exposure, working environment and spot measurement concentrations were subjected to log transformation because the values were lognormally distributed, and a Chi-square test for fitness was performed to determine whether data were normally distributed. The relationship among personal exposure concentrations, working environment measurement concentrations and spot measurement concentrations was investigated by Pearson's correlation coefficient test.

Results

Relationship among personal exposure concentrations, working environment measurement concentrations and spot measurement concentrations

Figures 2 (A) and 2 (B) show the relationships between personal exposure and working environment measurement concentrations of vaporous chemicals and chemicals in dust, respectively. Because the dispersion mechanisms of vaporous chemicals and dust in the working environments were different, vaporous chemicals and chemicals in dust were analyzed separately. Statistically significant positive correlations (r=0.732–0.893, p<0.01) were determined between personal exposure concentrations and working environment concentrations. Figures 3 (A) and 3 (B) show the relationships between personal exposure and spot measurement concentrations of vaporous chemicals and chemicals in dust, respectively. Statistically significant positive correlations (r=0.688–0.708, p<0.01) were determined between personal exposure and spot measurement concentrations.

Fig. 2.

Relationship between personal exposure concentration and workplace concentration (E_A1 is the 95th percentile value; E_A2 is the arithmetic mean) for (A) 35 workers exposed to vaporous chemicals and (B) 19 workers exposed to dust.

Fig. 3.

Relationship between personal exposure concentration and spot measurement concentration (SPOT-AM indicates the arithmetic mean concentrations; SPOT-MAX indicates the highest concentrations) for (A) 95 workers exposed to vaporous chemicals and (B) 57 workers exposed to dust.

Values of exposure concentrations divided by the E_A1 and E_A2

Table 2 shows the percentile values of personal exposure concentrations divided by the E_A1 and E_A2 which were calculated based on the working environment measurement concentration. The 5th to 95th percentile values of the personal exposure concentrations of vaporous chemicals divided by the E_A2 ranged from 0.17 to 7.69, and those for dust ranged from 0.27 to 18.06. The use-process categories in which there was a relatively large difference between personal exposure concentrations and the E_A2 were weighing, forming and bonding use-processes for which local exhaust ventilation was set up and wet equipment was operated. Weighing, forming and bonding use-processes were conducted repeatedly during sampling.

Table 2. Percentile values of personal exposure concentration divided by workplace concentration or spot measurement

		Percentile
		n	5	25	50	75	95
Vapor	EC / E_A1	35	0.06	0.13	0.29	0.79	2.76
	EC / E_A2	35	0.17	0.38	0.84	2.47	7.69
	EC / A-GM	35	0.22	0.63	1.43	4.06	10.34
	EC / A-MAX	35	0.10	0.30	0.63	1.09	4.99
	EC/SPOT-AM	95	0.04	0.14	0.39	1.02	2.05
	EC/SPOT-MAX	95	0.02	0.10	0.30	0.86	1.65
Dust	EC / E_A1	19	0.10	0.28	0.58	1.70	6.00
	EC / E_A2	19	0.27	0.94	1.88	6.21	18.06
	EC / A-GM	19	0.87	2.03	3.89	15.48	79.85
	EC / A-MAX	19	0.14	0.55	1.26	5.80	21.79
	EC/SPOT-AM	57	0.05	0.29	1.28	4.08	9.65
	EC/SPOT-MAX	57	0.03	0.21	0.84	2.92	6.06

EC: personal exposure concentration. A-GM: The geometric mean concentration of the A_value. A-MAX: The highest concentration of the A_value. SPOT-AM: The arithmetic mean concentration of the spot_value. SPOT-MAX: The highest concentration of the spot_value.

Figure 4 (A) shows the values of personal exposure concentrations divided by the E_A2. The values of personal exposure to vaporous chemicals divided by the E_A2 were within ten times the E_A2, but the three values for dust divided by the E_A2 were greater than ten times the E_A2. The three highest values for dust were observed for weighing, forming, and bonding use-processes. The sampling time for the highest spot measurement was 8 minutes, and that for the second highest spot measurement was 15 minutes. If the values of the aforementioned weighing and forming use-processes were excluded, the personal exposure concentrations were within ten times that of the E_A2 of the working environment measurements.

Fig. 4.

(A) Personal exposure concentrations divided by workplace concentrations (E_A2) and (B) spot measurement concentrations (sampling time).

Percentile values of exposure concentrations divided by spot measurement concentrations

Table 2 shows percentile values of personal exposure concentrations divided by the arithmetic mean concentrations of the spot measurements (SPOT-AM) and the highest concentration of the spot measurements (SPOT-MAX) obtained in the working environment measurements. The 5th to 95th percentiles of the distribution estimated by the data for the personal exposure concentrations divided by the SPOT-MAX for vaporous chemicals ranged from 0.02 to 1.65, and those for dust ranged from 0.03 to 6.06.

Figure 4 (B) shows the values of the personal exposure concentrations divided by the SPOT-MAX. The two highest values for dust in Fig. 4 (B) were obtained for different use-processes than those with the highest values for dust in Fig. 4 (A). The sampling time for the highest spot measurement was 20 minutes, and that for the second highest spot measurement was 15 minutes. Personal exposure concentrations were within ten times those of SPOT-MAX concentrations of working environment measurements.

Discussion

Statistically significant positive correlations were determined between personal exposure concentrations and arithmetic mean value (E_A2) of the working environment measurement concentrations obtained according to the Japanese working environment control system. The 5th to 95th percentiles of the values of personal exposure concentrations of vaporous chemicals divided by the E_A2 values were less than ten, except in the cases of the weighing, forming, and bonding use-processes. Therefore, ten times the E_A2 can be used to estimate personal exposure concentrations in most cases. In this study, the spot measurement concentrations were obtained instead of B-values because the objective of the survey conducted by the MHLW was to determine the exposure concentrations of each use-process. Usually, the B-value measured according to the Japanese working environmental control system is less than or equal to the spot measurement concentration because the B-value must be measured just for ten minutes, even when the targeted use-process continues for less than ten minutes. Since the sampling times for the two highest values for dust in Fig. 4 (B) were 20 minutes and 15 minutes, the spot measurement concentrations are supposed to be almost equal to or larger than the B-value²⁰⁾, which was not measured in this study. The 5th to 95th percentiles of the distribution estimated by the data for personal exposure concentrations divided by the highest concentration of the spot measurements (SPOT-MAX) for dust ranged from 0.03 to 6.06. In weighing, forming and bonding use-processes, personal exposure concentrations were greater than 10 times the E_A2. In these cases, ten times the B-value can be used as the personal exposure concentration instead the 10 times of E_A2. Substitution of B-values can be applied to the working environment in which worker exposure at some spots seems considerably higher than at other spots. And the E_A2 value or B-value obtained from Japanese working environment measurements is applicable to assessment of the personal exposure concentrations of chemicals.

According to the workplace exposure assessment rating criteria for exposure scenarios in R14.4.2 of REACH²²⁾, the kind of exposure data that can be used for exposure assessment is that shown in Table 14-1 for exposure scenarios in R14.4.2 of REACH. Since personal exposure concentration can be estimated by multiplying the measured value of an area monitoring system in Japan by a safety factor, our results suggest that the data of the Japanese method can be used as data of Table 14-1, and many Japanese factory data can be used for the creation of exposure scenarios. Thus, the E_A2, or B-value, obtained according to the Japanese working environment control system can be used to estimate personal exposure concentrations as data of medium or higher quality in both EU workplaces and REACH.

Cherrie et al.⁴⁾ concluded that the geometric mean of the personal exposure concentrations divided by the corresponding data obtained from static samples ranged from 1.2 to 8.5. This roughly agrees with our results, excluding the weighing, forming, and bonding use processes for chemicals in dust. Several other researchers^23-28) showed that personal exposure levels varied in occupational “homogeneous exposure groups”, which were identified as having the same job title, the same location and working environments with the identifiable features. For example, there were 15-fold differences among 95% of the individuals in these homogeneous groups²⁴⁾. Rappaport et al.^{24, 26)} suggested that industrial hygienists should adopt methods of statistical sampling and analysis owing to the variability of personal exposure levels among workers in a given occupational homogeneous exposure group. The Japanese workplace control system requires that companies using chemicals regulated by the Industrial Safety and Health Act conduct no fewer than five measurements of the workplace environment in one specified area on two different days. This minimum of five samples taken on different days ensures that the variability in exposure levels among workers is adequately represented. Our results show that the variations in the E_A2 value or B-value and in the geometric mean of the personal exposure concentrations are similar to the variation in the personal exposure levels in occupational homogeneous exposure groups²⁴⁾.

Because this study was conducted in representative workplaces in Japan, the results are considered to be typical conditions in Japanese workplaces; however, as the number of samples was small, other use-process categories may require further study.

In conclusion, ten times the arithmetic mean value (E_A2) of working environment measurement concentrations, obtained according to the Japanese working environment control system, can be used to conservatively estimate personal exposure concentrations, except for dust in some use-process categories. The concentrations of some chemicals in dust among some types of use-process categories, such as weighing, forming and bonding, were greater than ten times the E_A2. In such cases, ten times the B-value measured in the Japanese working environment control system, which was almost the same as the spot measurement concentration obtained in this study, can substitute for ten times the E_A2.

Acknowledgments: This study was a part of the Japanese Ministry of Health, Labour and Welfare's risk evaluation project for chemicals in the workplace. We would like to thank the engineers of the Japan Industrial Safety and Health Association for performing measurements and analyses.

References

1) American Chemical Society (ACS): Chemical Abstracts Service Registry, Database counter. [Online]. 2013 [cited 2013 Mar 3] ; Available from: URL: http://www.cas.org/cgi-bin/cas/regreport.pl
2) Ministry of Health, Labour, and Welfare (MHLW), Japan: Update information of the chemical substance information. [Online]. 2013 [cited 2013 Mar 3]; Available from: URL :http://anzeninfo.mhlw.go.jp/user/anzen/kag/ankgc01.htm (in Japanese).
3) European Chemicals Agency (ECHA): Summary of Work Programme 2011. [Online]. 2011 [cited 2011 Dec 18]; Available form: URL: http://echa.europa. eu/documents/10162/17210/summary_wp2011_en.pdf.
4) Cherrie JW. The Beginning of the Science Underpinning Occupational Hygiene. Ann Occup Hyg 2003; 47: 179-85.
5) Purdham JT, Bozek PR, Sass-Kortsak A. The evaluation of exposure of printing trade employees to polycyclic aromatic hydrocarbons. Ann Occup Hyg 1993; 37: 35-44.
6) Hori H, Higashi T, Fujino A, et al., Measurement of airborne ceramic fibres in manufacturing and processing factories. Ann Occup Hyg 1993; 37: 623-9.
7) Pengelly MI, Groves J, Foster RD, Ellwood PA, Wagg RM. Development of a method for measuring exposure to resin acids in solder fume. Ann Occup Hyg 1994; 38: 765-76.
8) Scheeper B, Kromhout H, Boleij SM. Wood-dust exposure during wood-working processes. Ann Occup Hyg 1995; 39: 141-54.
9) Cherrie JW, Gibson H, McIntosh C, Maclaren WM, Lynch G. Exposure to fine airborne fibrous dust amongst processors of para-aramid. Ann Occup Hyg 1995; 39: 403-25.
10) Plato N, Krantz S, Andersson L, Gustavsson P, Lundgren L. Characterization of current exposure to man-made vitreous fibres (MMVF) in the prefabricated house industry in Sweden. Ann Occup Hyg 1995; 39: 167-79.
11) Lange JH, Lange PR, Reinhard TK, Thomulka KW. A study of personal and area asbestos concentrations during asbestos abatement: a statistical evaluation of fibre concentration data. Ann Occup Hyg 1996; 40: 449-66.
12) Moen BE, Hollung BE. Exposure to organic solvents among car painters in Bergen, Nrway. Ann Occup Hyg 2000; 44: 185-9.
13) Groves J, Cain JR. A survey of exposure to diesel engine exhaust emissions in the workplace. Ann Occup Hyg 2000; 44: 435-47.
14) De Raeve H, Van Cleemput J, Nemery B. Airborne polyvinyl alcohol (PVA) and cellulose fibre levels in fibre-cement factories in seven European countries. Ann Occup Hyg 2001; 45: 625-30.
15) Watson SI, Cain JR, Cowie H, Cherrie JW. Development of a push-pull ventilation system to control solder fume. Ann Occup Hyg 2001; 45: 669-76.
16) Thorne J, Beijer L, Jonsson T, Rylander R. Measurement strategies for the determination of airborne bacterial endotoxin in sewage treatment plants. Ann Occup Hyg 2002; 46: 549-54.
17) Ministry of Health, Labour, and Welfare (MHLW), Japan. Guideline of exposure evaluation of worker'- s' harm: MHLW; 2009 (in Japanese).
18) Koshi S. Proposition for method of evaluating the work environment with regard to air-borne toxic substance. Ind Health 1980; 10: 179-86.
19) Koshi S. A basic framework of working environment control for occupational health in Japan. Ind Health 1996; 34: 149-65.
20) Kumagai S. Autocorrelation of short-term and daily average exposure levels in workplaces. Am Ind Hyg Assoc J 1993; 54: 341-50.
21) Four-step Excel statistical statcel 2 Statcel: The Useful Add-in Software Forms on Excel, 2nd ed. The publisher OMS. ISBN 798-4-434-04060-3 2004 (in Japanese).
22) European Chemicals Agency (ECHA), Guidance on information requirements and chemical safety and assessment, Chapter R14: Occupational exposure estimation version 2. ECHA; 2010.
23) Kromhout H, Symanski E, Rappaport SM. A comprehensive evaluation of within- and between- worker components of occupational exposure to chemical agents. Ann Occup Hyg 1993; 37: 253-70.
24) Rappaport SM, Kromhout H, Symanski E. Variation of exposure between workers in homogeneous exposure groups. Am Ind Hyg Assoc J 1993; 54: 654-62.
25) Symanski E, Kupper LL, Kromhout H, Rappaport SM. An investigation of systematic changes in occupational exposure. Am Ind Hyg Assoc J 1996; 57: 724-35.
26) Lyles RH, Kupper LL, Rappaport SM. A lognormal distribution-based exposure assessment method for unbalanced data. Ann Occup Hyg 1997; 41: 63-76.
27) Weaver MA, Kupper LL, Taylor D, Kromhout H, Susi P, Rappaport SM. Simultaneous assessment of occupational exposures from multiple worker groups. Ann Occup Hyg 2001; 45: 525-42.
28) Liu S, Hammond SK, Rappaport SM. Statistical modeling to determine sources of variability in exposures to welding fumes. Ann Occup Hyg 2011; 55: 305-18. Epub 2011 Feb 25. Review. PubMed PMID: 21355083.

Corresponding author

Register with J-STAGE for free!