Ann. ocaip «jv.. Vd 36. No. 4. pp. 363-372. 1992. Printed in Grcmt Britain. t

0003-4878/92 S5 00+0.00 Pergamoo Press Ud 1992 British Occupational Hygiera Socitl).

OCCUPATIONAL EXPOSURE TO CARBON MONOXIDE AND SULPHUR DIOXIDE DURING THE MANUFACTURE OF CARBON BLACK K. GARDINER, W. N. TRETHOWAN, J. M. HARRINGTON, I. A. CALVERT and D. C. GLASS Institute of Occupational Health, University of Birmingham, Edgbaston, Birmingham B15 2TT, U.K.

Abstract—The manufacture of carbon black is known to generate carbon monoxide and sulphur dioxide in the 'production gas' and the pyrolysis products of the 'production gas', respectively. Adverse health effects have been reported as associated with both contaminants (coronary heart disease with carbon monoxide and respiratory morbidity with sulphur dioxide). A major crosssectional and longitudinal respiratory morbidity study is being conducted to assess the effects of exposure to carbon black on lung function, on chest X-rays and on responses to a questionnaire. The questionnaire includes questions on respiratory and cardiovascular symptoms, so that information regarding confounding exposure is essential. The working population of 18 manufacturing plants in seven European Countries was split into 13 job title numbers (1-13) which were then amalgamated into five job categories (A-E), with an appropriate (statistically) number of samples taken from each plant-job category. In total, 1322 carbon monoxide samples and 1301 sulphur dioxide samples were taken, using actively pumped longterm colorimetnc tubes. In the majority of cases, more than half of the samples in each plant-job category were either zero or trace, thereby preventing the accurate estimation of the average exposure. The highest median carbon monoxide concentration was from the amalgamated data from all 18 plants in job number 9 (furnace operators), the highest median sulphur dioxide concentration was only 'trace'. The large number of zero and trace values also precluded the generation of current and retrospective exposure indices.

INTRODUCTION THE HISTORICAL aspects of the manufacture of carbon black have been well documented and are summarized elsewhere (GARDINER et al., 1992a). Carbon black is manufactured by a number of different methods, but in all of the 18 plants involved in this study the oil-furnace process is used (in addition, one plant uses the thermal method, one plant uses the channel method, and one plant uses both the lamp black and channel methods). In the majority of cases the highly aromatic oil feedstock is steam atomized into the reactor which is at a temperature of 1200-1850°C. The feedstock is either pyrolysed under reduction conditions, or is 'cracked' wherein the carbon atoms are dissociated from the other elements present in the oil. As the yield of carbon black is never 100% of the available carbon in the feedstock, the remaining carbon may form both inorganic and organic compounds. The major by-products are carbon monoxide and acetylene, but other noncarbonaceous compounds are also produced, such as hydrogen sulphide and hydrogen. These gases are called 'production gas' and after cooling are separated from the carbon black in large multi-compartmental filters. As a result of legislative restrictions on direct emission and of the high energy content of this gas, it is used in other processes such as power generation and community heating. The combustion 363

Downloaded from http://annhyg.oxfordjournals.org/ at Mount Royal University on July 16, 2015

(Received 17 September 1991 and in final form 12 March 1992)

364

K. GARDINER el

al.

Downloaded from http://annhyg.oxfordjournals.org/ at Mount Royal University on July 16, 2015

of this gas, especially of the hydrogen sulphide, gives rise to other contaminants, such as sulphur dioxide. The large cross-sectional and longitudinal study, of which this is a part, aims to assess the respiratory morbidity of individuals exposed to carbon black during its manufacture (GARDINER et al., 1992b). However, in the assessment of respiratory morbidity by means of spirometry and an adapted Medical Research Council (MRC) questionnaire the effects of confounding exposure must not be ignored. As a result of a comprehensive feasibility study, in which all of the plants were visited and a standard questionnaire including a section on confounding exposure was completed, it was decided to measure the concentration of sulphur dioxide to which the workforce were exposed. In addition, since there are reports in the literature of left ventricular hypertrophy (NAVetal., 1976) and right atrial and ventricular strain (NAima/., 1962) from animals exposed to carbon black, and as the questionnaire includes questions relating to symptoms of cardiac disorders, it was also decided to assess exposure to carbon monoxide. Carbon monoxide is a colourless, tasteless, flammable and odourless gas which is easily absorbed through the lungs and combines with haemoglobin to form carboxyhaemoglobin (COHb). The formation of COHb prevents oxygen transport and inhibits the release of oxygen by shifting the dissociation curve to the left. However, what is of concern in this study is the less well known binding of carbon monoxide with myoglobin to form carboxymyoglobin, as this adversely affects muscle metabolism, especially the heart: investigations of these effects and reviews of the literature have been published elsewhere (KRISTENSEN, 1989; GOLDSMITH and ARONOW, 1975; ATKINS and BAKER, 1985). Sulphur dioxide is a colourless, non-flammable gas with a pungent odour at or above 1 ppmand an acid taste (AMOOREand HAUTALA, 1983). It is a highly irritant gas, and owing to its high solubility sulphurous and sulphuric acids are formed on contact with moist mucosae, thereby affecting mainly the upper respiratory tract. As with carbon monoxide, a number of health studies have been undertaken with work of SANDSTROM et al. (1988) and ROM et al. (1986) of importance. In spite of the documented evidence of the health effects both of carbon monoxide and of sulphur dioxide few researchers have measured personal exposures in carbon black manufacturing plants, and those that have did not use the data to assess its potential confounding effect on respiratory or cardiac morbidity. Although a Health Hazard Evaluation (HHE) was conducted by NIOSH at a carbon black plant in Ohio, U.S.A. (NIOSH, 1981) only Komarova investigated exposure to carbon monoxide which she did in a number of Russian manufacturing plants in 1965 and 1973 (KOMAROVA 1965, 1973). The HHE was conducted over a 2-day period and measured airborne concentrations of carbon black, carbon monoxide, welding fume (including iron oxide) and the cyclohexane extractable content of the carbon black. Neither of the Russian papers provided any detailed information as to whether personal samplers or static samplers were used, or methods or results, except that exposure was high enough to result in elevated levels of carboxyhaemoglobin among the exposed workforce. No work has been published on the levels of sulphur dioxide in the work environment of carbon black manufacturing plants, all of the health studies examining the potential adverse respiratory effects of carbon black having neglected this known cause of respiratory symptoms. The authors found one report on the interaction of

Exposure to CO and SO 2 in carbon black manufacture

365

carbon black dust and gaseous sulphur dioxide, but this examines the alterations of lung defence mechanisms against airborne bacteria (RYLANDER, 1969). The health effects of exposure to carbon black dust have been examined in a number of studies. The bulk of these are Eastern European, and suffer from a number of methodological shortcomings, but the two scientifically more valid studies conducted on the European (CROSBIE, 1986) and American carbon black workforces (ROBERTSON et al., 1988) are of greater importance. All of these studies are reviewed in detail elsewhere (GARDINER et al., 1992b).

TABLE 1. AGREED JOB CATEGORIES AND JOB TITLES

Job categories

Job number

Job title

A

1

Administration area workers

B

2 3 4

Laboratory assistant Instrument mechanic Electrician

C

5 6 7 8 9

Process control operator (VDU/ control room) Process foreman Fitter Welder Furnace operator

D

10 11

Process operator Conveyor operator

E

12 13

Warehouse packer/shipping Cleaner (not office)

Downloaded from http://annhyg.oxfordjournals.org/ at Mount Royal University on July 16, 2015

METHODS AND MATERIALS

As described in more detail elsewhere (GARDINER et al., 1992a), a feasibility study was undertaken (1) to assess the likelihood of successfully obtaining sufficient data of a suitable quality from the 18 plants in seven countries; and (2) to identify such information as job title by which individuals could be grouped and sampled (see Table 1). To ensure both representative and random personal sampling, individuals were assigned one of 13 job codes/titles (1-13) which were then amalgamated into five job categories (A-E) within each of which the exposure was estimated to be equal and from these a proportion of each group (chosen to give 90% confidence of sampling an individual in the top 10%) in each plant was randomly chosen for both contaminants (NIOSH, 1977). A Sampling Register was then created for each contaminant and for each plant; this consisted of a randomized list of the unique identification numbers individuals had been given, their job code and the date on which they were sampled, and the Register then returned to the Institute of Occupational Health (IOH). A degree of redundancy was built into the numbers chosen to ensure that the natural rejection rate of individuals to the demands of occupational hygiene sampling over a period did not reduce the number to less than the minimum required. In addition, individuals who left the company or were permanently unavailable were replaced by others randomly selected from the same job category (GLASS, 1990).

366

K. GARDINER el at.

All of the factory hygienists attended a 2-day training session at the IOH et aL, 1992a) where detailed instructions were given as to the methods and difficulties of using and reading Draeger tubes. To minimize inter-plant variability, all plants and their personnel were instructed to interpret any colour change as being that of the contaminant of interest, and that if there was an uneven 'front' across the tube, then the maximum value should be taken. The factory hygienists were also instructed to make comments on the Sample Record Sheet if one or both of the above events occurred. The concentrations of carbon monoxide and sulphur dioxide were measured using calibrated SP15 low-flow pumps (Casella Ltd, U.K.) to draw air through long-term colorimetric tubes [types 10/a-L and 2/a-L, respectively (Draeger Ltd, U.K.)] . The contaminants were both measured over approximately 4 h with a sample volume of about 4 1. The sample tubes were placed in a holder which was then secured onto the lapel of the person involved. At the end of the sample period the sampling train was recalibrated, the stain length of the tube read and the other details of the procedure such as I.D. Number, date, duration, volume, etc., were recorded on the colour-coded Record Sheet. Because of the possible change in stain length over time, tubes were not returned to the IOH for verification. The tubes are graduated in microlitres (jil), that for CO from 10 to 100 /il and that for SO 2 from 2 to 20 fi\, giving concentration ranges of 2.5-25 ppm and 0.5-5, respectively, with a 4 1. sample. Any discernible stain below the first gradation was recorded as 'trace'. The colour change of the tubes is from white to brown for carbon monoxide and from bluish violet to yellow for sulphur dioxide, the basis of the reactions being: (GARDINER

SeO 2 + H 2 S 2 O 7

(I2 + 5CO) 2

(catalysts) (white)

> (brown)

SO 2 + 2NaOH (violet)

(Na 2 SO 3 + H 2 O) • (yellow).

Simultaneous exposure to benzene and acetylene (but not to hydrogen sulphide) may give rise to a cross-sensitivity at high concentrations, for the CO detector tubes and exposure to other acid reacting compounds such as hydrochloric and acetic acid, chlorine and nitrogen dioxide may give false readings for the SO 2 tubes (DRAEGER, 1989). Of these contaminants only acetylene is likely to confound the measurements at low concentrations. RESULTS

In all, 1322 carbon monoxide and 1301 sulphur dioxide samples were taken in the 18 European plants between June 1987 and June 1989. The median duration both of the carbon monoxide and of the sulphur dioxide samples was 240 min, with interquartile ranges of 11.25 and 5 min, respectively. The median carbon monoxide and sulphur dioxide concentrations are shown in Table 2. The data from all 18 plants have been amalgamated and are shown by job number. (The reason for using the median concentration is explained in the Discussion.) For CO the median

Downloaded from http://annhyg.oxfordjournals.org/ at Mount Royal University on July 16, 2015

5CO + I 2 O 5

Exposure to CO and SO 2 in carbon black manufacture

367

TABLE 2. MEDIAN CO AND SO2 CONCENTRATION FOR AMALGAMATED DATA FROM ALL 18 PLANTS

Job category

Job number

Median CO (ppm)

Median SO 2 (ppm)

1

Trace (285)

0 (277)

B

2 3 4

Trace (152) Trace (58) 1.2(56)

Trace (145) Trace (59) 0(56)

C

5 6 7 8 9

Trace (49) Trace (79) 1.9 (95) Trace (44) 2.1 (50)

0(51) 0(71) 0(90) 0(33) 0(57)

D

10 11

Trace (175) 1.2(57)

Trace (181) Trace (55)

E

12 13

Trace (155) 0.9 (44)

0(153) 0(43)

(22)

(3D

Non-assigned

Note: total number of samples for each job number is given in brackets.

concentration in job numbers 1-13 were'trace'except for electricians (B4), fitters (C7), furnace operators (C9), conveyor operators (D11) and cleaners (not office) (E13) where the median concentrations were 1.2,1.9,2.1, 1.2 and 0.9 ppm, respectively (well below the European 8 h-TWA Occupation Exposure Limit of 50 ppm). The SO 2 median concentrations in job numbers 1-13 were zero, except for laboratory assistants (B2), instrument mechanics (B3), process operators (D10) and conveyor operators (Dl 1), where the median concentrations were all 'trace'. The median concentrations for CO and SO 2 were also calculated for each job in each plant. The range of plant median CO and SO 2 concentrations are shown in Table 3, which presents the minimum and maximum median concentrations for each plant and for each job. For example, the minimum median concentration for CO was zero for process foremen (C6) in both plants 5 and 7. The minimum median concentration for both CO and SO 2 was zero, except that for CO these concentrations were measured in only a few plants (i.e. in job category/number A 1-only plants 2 and 5 had a minimum median concentration of zero) whereas all plants in all job numbers had a minimum median concentration of zero for SO 2 (i.e. in job category/number A1-E13 and in plants 1-18). The plant maximum median concentrations for CO ranged from 2.4 ppm for process control operators (C5), plant 1, to 19.8 ppm for electricians (B4), plant 16, and for SO 2 the range is from 0.5 ppm for process foremen (C6), plants 6 and 18, to 1.9 ppm for process control operators (C5), plant 9 (this last SO 2 concentration is very close to the European 8 h-TWA Occupational Exposure Limit of 2 ppm). The range of concentrations from amalgamated plant data by job numbers are presented in Tables 4 and 5 for CO and SO 2 , respectively. In addition, the percentage of samples above trace, above half of the Occupational Exposure Standard and above the Standard itself are given for both CO and SO 2 (25 and 50 ppm for CO, and 1 and 2 ppm for SO 2 ). The maximum concentration of CO was 83.3 ppm for furnace

Downloaded from http://annhyg.oxfordjournals.org/ at Mount Royal University on July 16, 2015

A

368

K. GARDINER et al. TABLE 3. RANGE OF PLANT MEDIAN CO AND SO,

Job category

Minimum median CO concentration (ppm)

Job number

CONCENTRATIONS

Maximum median CO concentration (ppm)

Minimum median SO 2 concentration (ppm)

4.2(16)

0

1.7(9)

Maximum median SO 2 concentration (ppm)

1

0 (2, 5)

B

2 3 4

0 (2, 5, 7) 0 (5, 7) 0 (5, 7)

4.3(1) 5.2(17) 19.8(16)

0 0 0

1.6(9) 1.0(9) 1.5(9)

C

5 6 7 8

2.4(1) 5.2(16) 13.9(17) 7.2(17)

0 0 0 0

1.9(9) 0.5(6, 18) 0.8 (9) 0.8 (9)

9

0(13) 0 (5, 7) 0 (5, 7) Trace (1,6, 10, 11, 14) 0(7)

16.7(6)

0

1.6(6)

D

10 11

0(10) Trace (8, 14, 17)

3.7 (3) 5.2(16)

0 0

1.9(15) 0.6 (6)

E

12 13

0(5) Trace (1,2, 8,9, 11)

5.2(16) 2.8 (3)

0 0

0.7 (9) 1.7(15)

Note: plant number is given in brackets. Trace has been assigned to readings where a colour change between zero and the first graduation was noted.

TABLE 4. RANGE OF CONCENTRATIONS WITH THE PERCENTAGE OF SAMPLES GREATER THAN TRACE, 25 AND 5 0 p p m FOR CARBON MONOXIDE AMALGAMATED BY PLANT

Job category

Job Minimum Maximum number concentration concentration

Percentage of samples > trace

Percentage of samples > 25 ppm

Percentage of samples > 50 ppm

ppm

Total number of samples

A

1

0

8.3

23.5

0

0

285

B

2 3 4

0 0 0

28.3 12.5 22.7

32.2 46.6 50.0

0.7 0 0

0 0 0

152 58 56

C

5 6 7 8 9

0 0 0 0 0

5.0 16.7 41.5 17.5 83.3

32.7 44.3 54.7 47.7 68.0

0 0 I.I 0 4.0

0 0 0 0 4.0

D

10 11

0 0

25.9 18.2

44.0 52.6

0.6 0

0 0

175 57

E

12 13

0 0

18.7 9.3

49.0 50.0

0 0

0 0

155 44

Note: 22 samples were assigned a job code but not a job number.

49 79 95 44 50

Downloaded from http://annhyg.oxfordjournals.org/ at Mount Royal University on July 16, 2015

A

Exposure to CO and SO 2 in carbon black manufacture

369

TABLE 5. RANGE OF CONCENTRATIONS WITH THE PERCENTAGE OF SAMPLES GREATER THAN TRACE, 1 ppm AND 2 p p m FOR SULPHUR DIOXIDE AMALGAMATED BY PLANT

Job category

Minimum Maximum Job number concentration concentration

Percentage of samples > trace

Percentage of samples > 1 ppm

Percentage of samples > 2 ppm

Total number of samples

1 2 3 4

0 0 0 0

4.3

31.4

10.1

3.2

277

8.3 2.1 2.1

40.0 35.6 37.5

14.5 5.1 1.8

6.9 1.7 1.8

145 59 56

C

5 6 7 8 9

0 0 0 0 0

3.8 1.0 4.2 1.7 3.8

41.2 33.8 24.4 15.2 28.1

15.7 0 5.6 3.0

7.8 0 2.2 0 3.5

51 71 90 33 57

D

10 11

0 0

4.3 4.2

42.0 41.8

13.8 5.5

7.2 3.6

181 55

E

12 13

0 0

2.5 1.7

19.0 27.9

3.3 4.7

2.0 0

153 43

Note: 31 samples were assigned a job category but not a job number.

operators (C9) and that of SO 2 was 8.3 ppm for laboratory assistants (B2). Although the percentage of CO samples above 'trace' ranges from 23.5% in administrative workers (Al) to 68% in furnace operators (C9), there are only five job numbers where there are 50% or more samples above 'trace', and there are no job numbers for SO 2 where there are more than 50% of the samples above 'trace'. This makes data analysis very difficult (see Discussion). Only four job titles have data greater than half the CO Occupational Exposure Limit and these are laboratory assistants (B2), fitters (C7), furnace operators (C9) and process operators (D10) where 0.7, 1.1, 4.0 and 0.6%, respectively, exceed 25 ppm with the same 4% of process operators (C9) exceeding 50 ppm. Despite the SO 2 data having a lower percentage of samples above 'trace', there are a much greater number of job numbers in which half the Occupational Exposure Limit and the Occupational Exposure Limit are exceeded. Only process foremen (C6) did not provide any samples above 1 or 2 ppm, and welders (C8) did not provide any above 2 ppm. The highest percentage above 1 ppm and above 2 ppm were both from process control operators (C5). The data presented has not been amended or adjusted to take into account the effect of control measures, as the respiratory protection equipment used was exclusively for particulates—usually filtering facepiece respirators—whereas the machinery on which local exhaust ventilation is used (packing machines) is not a source of either CO or SO 2 . DISCUSSION

As a result of the very high number (> 50%) of both CO and SO 2 samples of zero or 'trace' (i.e. less than the relative detection limits) it is not valid to calculate either means and standard deviations or geometric means and geometric standard deviations depending in distributional form. A number of workers have published papers on how

Downloaded from http://annhyg.oxfordjournals.org/ at Mount Royal University on July 16, 2015

A B

370

K. GARDINER et al.

Downloaded from http://annhyg.oxfordjournals.org/ at Mount Royal University on July 16, 2015

to estimate the average concentration in the presence of non-detectable values, but this is possible only where more than 50% of values are greater than the limit of detection (HORNUNG and REED, 1990; PERKINS et al., 1990; COHEN and RYAN, 1989). HORNUNG and REED (1990) stated that, "when the majority of samples are below the limit of detection, reporting a mean and standard deviation is a questionable practice. A better description of the data may simply be obtained by reporting the percentage of the samples below the limit of detection and the range of the remaining samples", and this is why the data in this paper are reported in this manner (Tables 4 and 5) as well as in the form of the percentages above half the OES and OES. To alleviate this problem a method with a lower limit of detection is required, however, with a type of survey in which there are 18 different hygienists in seven countries undertaking the sampling and analysis a simple technique like colorimetric detector tubes is essential. Unfortunately, no manufacturer makes either an active or a passive tube with a sufficiently high degree of sensitivity, probably because it is not relevant to the Occupational Exposure Limit. It is clear from the results presented that in the main the airborne concentrations of CO and SO 2 in the European manufacturing plants are low; however, on occasions concentrations above the Occupational Exposure Limits are measured. These are usually due to short-term elevated levels from interaction with the relatively enclosed plant process (see results for instrument mechanics, electricians, fitters and welders) or from other industrial machinery not specific to the carbon black industry such as fork lift trucks, cranes, compressors, etc. This information has been gained by a combination of data supplied on the Record Sheets and on the personal experience of one the authors (K. Gardiner) of the industry. By the nature of the sampling strategy used, it is not possible to assess this type of one-off exposure systematically, as the next individual on the sampling register should be sampled regardless of the task(s) to be performed. The alternative would be to choose people specifically, if the work is likely to generate high exposure, as one would do with compliance testing. It is worth noting that the feasibility study aimed to identify and group job titles mainly for the purpose of sampling carbon black dust, rather than either CO or SO 2 • In fact it is likely that different job titles would be grouped together for CO and again for SO 2 , but it was not practical to have three separate lists and groupings of them, particularly as 18 plants were involved. However, because the concentrations measured and the limit of detection of the technique used it is most unlikely that a different sampling strategy would have provided more meaningful data. If separate groupings were made both for CO and forSO 2 , then those job numbers most likely to have the highest exposure for CO would be, in descending order, the furnace operators, fitters, conveyors and electricians, and for SO 2 , the process operators, conveyor operators, laboratory staff and fitters. As mentioned previously, in addition to assessing symptoms of respiratory morbidity, tests of lung function were also performed. The sampling strategy was designed to estimate average exposure by job category; however, the known effect of sulphur dioxide on lung function is acute and therefore in terms of the spirometric investigation, the relevant measurement should have been taken in the 4 h previous to the test. As with the difficulty in terms of job titles and category groupings discussed in the previous paragraph, this was not practical as, first, a great many medical examinations were carried out in the first 4 h of the day thereby preventing a complete sample from being taken, and second, the creation of two sampling registers (one for

Exposure to CO and SO 2 in carbon black manufacture

371

REFERENCES AMOORE, J. E. and HAUTALA, E. (1983) Odour as an aid to chemical safety: odour threshold compared with threshold limit values and volatilities for 214 industrial chemicals in air and water dilution. / . appl. Toxic. 3, 272-290. ATKINS, E. H. and BAKER, E. L. (1985) Exacerbation of coronary artery disease by occupational carbon monoxide exposure: A report of two fatalities and a review of the literature. Am. J. ind. Med. 7, 73-79. COHEN, M. A. and RYAN, P. B. (1989) Observations less than the analytical limit of detection: a new approach. J . Air Pollut. Control Ass. 39, 328-329. CROSBIE, W. A. (1986) The repiratory health of carbon black workers. Archs Environ. Hlth 41, 346-353. DRAEGER (1989) Detector Tube Handbook (7th Edn) (Complied by LHCHNITZ, K.). Dragerwerk, A.G., Lubeck, Germany. GARDINER, K., TRETHOWAN, W. N., HARRINGTON, J. M., CALVERT, I. A. and GLASS, D. C. (1992a)

Occupational exposure to carbon black in its manufacture. In press. GARDINER, K., TRETHOWAN, W. N. HARRINGTON, J. M., ROSSITER, C. E. and CALVERT, I. A. (1992b) The

respiratory health effects of carbon black—a survey of European carbon black workers. In press. GLASS, D. C. (1990) An assessment of the exposure of water reclamation workers to hydrogen sulphide. Ann. occup. Hyg. 34, 509-519. GOLDSMITH, J. R. and ARONOW, W. S. (1975) Carbon monoxide and coronary heart disease: A review. Environ. Res. 10, 236-248. HORNUNG, R. W. and REED, L. D. (1990) Estimation of average concentration in the presence of nondetectable values. Appl. occup. Environ. Hyg. 5, 46-51. KOMAROVA, L. T. (1965). [The effect of air contamination in the production of carbon black on the morbidity and health of the workers.] Nauchn. Tr. Omsk. Med. /nst. 61, 115-121 (in Russian). KOMAROVA, L. T. (1973) [Hygienic evaluation of the production of carbon black by the furnace method.] Gig. Tr. Prof. Zabol. 17, 32-36 (in Russian). KRISTENSEN, T. S. (1989) Cardiovascular diseases and the work environment. A critical review of the epidemiologic literature on chemical factors. Scand. J. Wk Environ. Hlth 15, 245-264. NAU, C. A., NEAL, J., STEMBRIDGE, V. A. and COOLEY, R. N. (1962) Physiological effects of carbon black IV Inhalation. Archs Environ. Hlth 4, 415-431. NAU, C. A., TAYLOR, G. T. and LAWRENCE, C. H. (1976) Properties and physiological effects of thermal carbon black. J. Occup. Med. 18, 732-734. NIOSH (1977) Occupational exposure sampling strategy manual. National Institute for Occupational Safety and Health, Cincinnati, Ohio, U.S.A. NIOSH (1981) Health hazard evaluation report No. HHE-80-203-960. Phillips Chemical Company, Toledo, Ohio, U.S.A. Hazard Evaluations and Technical Assistance Branch, NIOSH, Cincinnati, Ohio, U.S.A. PERKINS, J. L., CUTTER, G. N. and CLEVELAND, M. S. (1990) Estimating the mean, variance and confidence limits from censored ( < limit of detection), lognormally-distributed exposure data. Am. ind. Hyg. Ass. J. 51,416-419.

Downloaded from http://annhyg.oxfordjournals.org/ at Mount Royal University on July 16, 2015

estimating average exposure for the symptoms analysis and one for the immediate adverse effect in lung function) for SO 2 would have been undesirable for both the IOH and the 18 plants. The results of this study strongly suggest that the effect would not have matched the benefit. The main purpose of the measurement of these two gaseous contaminants was to assess the potential confounding effect both on symptoms and on lung function, and if necessary to generate both current and retrospective exposure indices against which the symptoms and lung function tests could be analysed. The very low levels measured enabled the elimination of these two contaminants as confounders due to the highly censored nature of the results (i.e. more than 50% below the limit of detection) and it was neither possible nor necessary to analyse against the health data. This is the only study to have examined the concentration of these contaminants in carbon black manufacturing plants but, most importantly, it is the only such study to have been conducted concurrently with the measurement of health effects (GARDINER et al., 1992b) with which they are associated.

372

K. GARDINER et a\.

ROBERTSON, J. M C D . , DIAZ, J. F., FYFE, I. M. and INGALLS, T. H. (1988) A cross-sectional study of

pulmonary function in carbon black workers in the United States. Am. ind. Hyg. Ass. J. 49, 161-166. ROM, W. N., WOOD, S. D., WHITE, G. L., BANG, K. M. and READING, J. C. (1986) Longitudinal evaluation of

pulmonary function in copper smelter workers exposed to sulphur dioxide. Am. Rev. resp. Dis. 133, 830-833. RYLANDER, R. (1969) Alterations of lung defense mechanisms against airborne bacteria. Archs Environ. Hlth 18, 551-555. SANDSTROM, T , KOLMODIN-HEDMAN, B., STJERNBERG, N., ANDERSSON, M. C. and LOFVENIUS, G. (1988)

Challenge test for sulphur dioxide symptom and lung function measurements. Scand. J. Wk Environ. Hlth 14, Suppl. 1, 77-79.

Downloaded from http://annhyg.oxfordjournals.org/ at Mount Royal University on July 16, 2015

Occupational exposure to carbon monoxide and sulphur dioxide during the manufacture of carbon black.

The manufacture of carbon black is known to generate carbon monoxide and sulphur dioxide in the 'production gas' and the pyrolysis products of the 'pr...
509KB Sizes 0 Downloads 0 Views