Ann. occup. Hyg., Vol. 47, No. 5, pp. 343-348, 2003
© 2003 British Occupational Hygiene Society
Published by Oxford University Press
Commentary: Two Seminal Contributions of S. A. Roach to the Evaluation and Control of Hazardous Substances in Air
School of Public Health, University of North Carolina, Chapel Hill, NC 27599-7431, USA
Received 18 February 2003; in final form 1 March 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT A BRIEF SUMMARY OF...‘A MOST RATIONAL BASIS...‘ON THE ROLE OF...REFERENCES |
|---|
S. A. Roach was a pioneer in the assessment and control of hazardous substances in the working environment during the second half of the 20th century. The two papers discussed in this commentary are generally regarded as his most important scientific contributions. The first paper (Roach, 1977) dealt with the determinants of the body burdens of toxic air contaminants. Using simple kinetic models, he showed how levels of toxicants rise and fall in the body according to the patterns of airborne exposures received during relevant time windows. This led to several useful rules of thumb, including the timing of grab samples for ‘fast acting’ substances, the appropriate duration of air samples relative to the biological half time, how to deal with unusual work schedules, and how to integrate exposure assessment with control. He also offered sage advice regarding the meaning and interpretation of exposure limits, the importance of repeated monitoring, and the extent to which unacceptable levels of exposure might be reduced. In concluding this work, Roach emphasized that the hygienist can fulfill a central role in occupational health simply by intervening to reduce the body burden. The second paper (Roach, 1981) dealt with the design of effective ventilation systems to control worker exposure to toxic airborne contaminants. By developing a series of simple differential equations, Roach evaluated the impact of turbulent diffusion upon industrial ventilation. He emphasized that the stationary contaminant concentration was proportional to the contaminant generation rate and that velocity alone was not a sufficient design criterion to control exposures. Rather, he argued that the equivalent ventilation rate (the ratio of the contaminant generation rate to the steady concentration in the breathing zone) should be the guiding criterion for ventilation design. Throughout both papers, Roach used fundamental principles to tie together exposure assessment and engineering control, and pointed the way to a science for occupational hygiene. The profession can still learn a great deal from these seminal contributions.
Keywords: air contaminants; air turbulence; body burden; control; exposure; ventilation
|
A BRIEF SUMMARY OF THE WORK OF S. A. ROACH |
|---|
|
TOPABSTRACTA BRIEF SUMMARY OF...‘A MOST RATIONAL BASIS...‘ON THE ROLE OF...REFERENCES |
|---|
Stanley A. Roach read mining engineering at Sheffield University and physics at London University in the late 1940s. This blending of theory with practice would surface repeatedly in Stan’s career as an occupational hygienist. His professional work began in 1949, as a scientist with the BMRC Pneumoconiosis Research Unit. There he studied dust exposures in the collieries of South Wales, where colliers suffered the ravages of anthrasilicosis. While it was known that long-term inhalation of coal dust led to this condition, the quantitative link between exposure and disease was not understood. Working with statistician Peter Oldham, Stan randomly sampled dust levels among the colliers with repeated breathing-zone measurements. Roach and Oldham used these data to investigate the relationship between exposure to coal dust and the progression of lung disease (Oldham and Roach, 1952; Roach, 1953, 1959). They also applied the lognormal distribution to characterize air levels, the first such application to occupational data (Oldham, 1953).
In 1959, Stan became a reader in occupational hygiene at the London School of Hygiene and Tropical Medicine (LSHTM) under Professor Richard Schilling. There he established the first graduate course in occupational hygiene in the UK. During his tenure at the LSHTM, Stan earned a Ph.D. in occupational health (London University) with a dissertation entitled The Theory of Random Clumping (Roach, 1968). This work on the theory of clustering phenomena demonstrated a grasp of statistical theory that Stan would later apply in his studies of occupational exposures. During the 1960s, Stan initiated theoretical work on the interrelations of exposure to hazardous substances in air, the build-up of these substances in the body, and industrial ventilation, themes that he would pursue for the rest of his life. His 1966 paper, ‘A more rational basis for air sampling’ (Roach, 1966), was the first to apply a random input to a toxicokinetic model in considering the impact of exposure variability upon the dose received over time. While at the LSHTM, Stan also began to explore the setting of occupational exposure limits, and contributed a sophisticated sampling strategy to the BOHS’s 1968 standard for asbestos (British Occupational Hygiene Society, 1968; Ogden, 2003).
Stan left the LSHTM in 1969 and, after a period of private consultancy, became the Corporate Occupational Hygienist for Imperial Chemical Industries, UK, in 1975. There he developed a program for evaluating and controlling hazardous substances in factories world-wide, and became the industry representative to many boards and commissions involved with exposure limits. Although no longer an academician, Stan continued his theoretical studies of exposure, body burden, and ventilation with two major papers in the Annals of Occupational Hygiene, namely ‘A most rational basis for air sampling programmes’, in 1977 (Roach, 1977), and ‘On the importance of turbulent diffusion in industrial ventilation’, in 1981 (Roach, 1981). These writings are generally regarded as Stan’s most important contributions to the science and practice of occupational hygiene, and are the subject of the following reviews.
After retiring from ICI in 1988, Stan spent a year with one of us (S.M.R.) at the University of California, Berkeley. There he explored the underpinnings of the Threshold Limit Values, exposure limits that were (and probably still are) the most influential in the world. Although this inquiry had been initiated during Stan’s years at ICI, it could not be completed there due to the controversial nature of his findings (Roach and Rappaport, 1990). Stan retired to Cambridgeshire in 1989 but continued to write about occupational hygiene and exposure limits (Roach, 1992) until his death in 1996. He remained witty and irascible to the end.
|
‘A MOST RATIONAL BASIS FOR AIR SAMPLING PROGRAMMES’ (COMMENTARY BY S. M. RAPPAPORT) |
|---|
|
TOPABSTRACTA BRIEF SUMMARY OF...‘A MOST RATIONAL BASIS...‘ON THE ROLE OF...REFERENCES |
|---|
When I first entered academia in 1976, I struggled with the theory and practice of exposure assessment. I knew from experience that air levels of industrial contaminants varied by as much as 100-fold from one measurement to the next. (Perhaps this explained why many hygienists restricted their air sampling to one measurement.) Since some substances produced toxicity in a single day while others required many years, there was also little agreement as to how exposure variability might impact the health of workers. It seemed naïve to assume that one assessment strategy would serve all purposes, but such was the situation at the time (Leidel et al., 1977). Given the paucity of thoughtful writing on this conundrum, I remember the excitement that winter day when I first perused Roach’s treatment of the subject (Roach, 1977). I would return to this paper many times.
Roach recognized that variability of industrial exposures was dictated by random processes plus the fixed effects of environmental characteristics. Of the environmental factors, he viewed ventilation of the workroom as the most important; this he characterized by a time constant (designated as 
). Likewise, Roach understood that the body burden of an inhaled substance was influenced by exposure variability plus the fixed effect of biological elimination, also characterized by a time constant (designated as a) (Roach, 1966). Roach argued in his 1977 paper that the interplay of and a should be used, along with the nature of the toxic effect, to guide the hygienist’s sampling strategy.
Turning first to the effect of ventilation, Roach modeled the workroom as a continuously stirred volume, with being the number of air changes per hour. He focused upon the smoothing effect that poor ventilation might exert upon a series of instantaneous air levels. He showed that if the interval between sequential measurements was small (a few minutes say), and if was <5, then air levels became highly autocorrelated. If the hygienist wished to predict the likelihood that short-term exposure to a ‘fast acting’ substance might exceed a ceiling limit, this autocorrelation would significantly bias estimation of the mean and variance, leading (generally) to underestimation of the exceedance. To avoid this problem, Roach encouraged the hygienist to consider the ventilation rate when assessing exposures to ‘fast acting’ substances; for well-ventilated rooms ( > 5), grab samples can be collected at intervals of 15 min or less without significant autocorrelation, but for poorly ventilated spaces ( = 1), samples should be spaced at least 2.3 h apart. Roach’s simple ventilation model for short-term autocorrelation predicted behaviors that are consistent with empirical observations from workplaces many years later (Kumagai et al., 1993; Kumagai and Matsunaga, 1994).
Roach then considered the relationship between airborne exposure and a contaminant’s body burden, using theory from his earlier treatment of toxicokinetics (Roach, 1966). He showed that the rate of build-up of a substance in the body was dictated by the first-order elimination rate constant a (with units of h–1), which can be related to the elimination half-time T = 0.693/a (with units of h). With continuous exposure to a chemical, the body burden increases until an equilibrium is achieved, where the rate of uptake equals the rate of elimination. Substances with small half-times (seconds to an hour or so) reach equilibrium within a single work shift, while those with long half-times (several months to years) require years to do so. In both cases, Roach suggested that the smoothing effect of biological elimination should be considered in designing sampling strategies, a notion that I elaborated upon some years later (Rappaport, 1985; Rappaport and Spear, 1988).
Roach differentiated ‘fast acting’ substances from those producing long-term effects. For ‘fast acting’ substances, the body burden is achieved following transient exposures during a single work shift. Roach reasoned that if the averaging time for grab samples were too long, then the measurements would not accurately reflect the ‘maximum tolerated air concentration’. However, he showed that this problem could be avoided by selecting an averaging of 0.3T, since peak air concentrations of shorter duration would be of little physiological consequence. Regarding substances producing long-term effects, Roach showed that the body burden varied over time about the mean burden, just as air levels varied from shift to shift about the mean concentration. However, the smoothing effect exerted by biological elimination effectively reduced the coefficient of variation (CV) of the body burden compared to that of the air levels. By selecting an averaging time of 3T, Roach showed that the CV of air measurements would be of the same magnitude as that of the body burden (provided that > 1 air change/h). Here, Roach implied that the mean exposure over intervals of 3T should be the object of concern, not the large, single-shift level that motivates compliance testing. It also follows that slowly eliminated substances are good candidates for biological monitoring, because of the reduction in variability of the body burden relative to air levels (Rappaport, 1988; Droz et al., 1991; Rappaport et al., 1995; Symanski et al., 2000).
Concluding the section on body burdens, Roach considered exposure regimens other than the usual work schedule of 8 h/day and 5 days/week. Using his toxicokinetic model, Roach offered useful guidelines for adjusting exposure limits for unusual work schedules, based upon the value of T for a particular substance. This is apparently the first such analysis, being published earlier in the same year as a similar work by Hickey and Reist (Hickey and Reist, 1977).
In the remainder of his paper, Roach commented upon the programmatic aspects of assessing exposures. He prefaced the final section with his interpretation of occupational exposure limits (at least those based upon industrial experience), as average air levels. Specifically, he stated on p. 79 of his paper:
The hygiene standard sets an upper limit to the job exposure, which it is currently believed, will produce responses in people exposed which are just tolerable medically, socially and industrially. It follows from the underlying procedure by which hygiene standards are derived that ‘exposure’ is the average contaminant concentration in the breathing zones of individuals undertaking a similar job.
Roach’s contention that occupational exposure limits related to average conditions within a given job, rather than to air levels during a single work shift, has certainly influenced my work in this area (Rappaport, 1991; Rappaport et al., 1995, 1999). But the idea remains controversial, at least in part, because regulators have been reluctant to address the issue (see e.g. Hewett, 1997a,b, 1998; Rappaport et al., 1998).
From his experience as a corporate occupational hygienist, Roach recognized the dynamic nature of the working environment, where processes and equipment were always changing. Thus, he viewed exposure assessment as a continuous process, where repeated measurements were used to evaluate the air environment in much the same manner that statistical quality control ensured acceptable products for commerce. In situations where exposures were found to be unacceptable, Roach recognized the need for reductions in air levels, but cautioned that dramatic improvements were impractical; that is, on p. 80 he stated that:
...it is reasonably practicable in industry to bring about a reduction in airborne contaminant concentration at an average rate of 25% per annum, but it is unrealistic to suppose that improvements may be made at a much faster rate.
This maximum reduction of 25% per year was undoubtedly derived from Roach’s intuition and experience. Nonetheless, it proved to be remarkably accurate when Symanski et al. showed (two decades later) that 90% of annual reductions in air levels were <25% (Symanski et al., 1998).
In concluding his paper, Roach embraced the role of the occupational hygienist as a central figure in the prevention of disease. Indeed, he emphasized that the hygienist can fulfill this role even if the mechanism of disease progression is not known, simply by intervening to reduce the body burden. As he stated in the last two lines of his paper:
Adverse health effects may appear long after the body burden which caused them has disappeared but they cannot occur before it has arrived. It is this principle which is at the heart of the matter.
If we were to embrace the ideas at the heart of Roach’s important contribution, i.e. relating exposure to ventilation and body burden and using repeated measurements to guide our efforts, we might yet achieve the special place that he held for us.
|
‘ON THE ROLE OF TURBULENT DIFFUSION IN VENTILATION’ (COMMENTARY BY M. R. FLYNN) |
|---|
|
TOPABSTRACTA BRIEF SUMMARY OF...‘A MOST RATIONAL BASIS...‘ON THE ROLE OF...REFERENCES |
|---|
It is an honor and pleasure to comment on this classic work by Stanley Roach. I had the great fortune of meeting Stan when I was a doctoral student studying the performance of local exhaust hoods. I greatly enjoyed my conversations with him on this subject. Several years later he visited me at the University of North Carolina and showed me some of his original notes on the capture efficiency of local exhaust hoods. The papers looked like old pirate maps—the pages brown and aged—and I was impressed (and surprised) that years ago he had made many of the same types of measurements that I, and others, had done only recently. He was truly one of the great pioneers in the science of occupational hygiene as shown in this classic work.
One of the great problems in occupational health engineering is the design of effective ventilation systems to control worker exposure to toxic airborne contaminants. This problem remains a challenge today because, as Roach remarked on p. 105 of this paper, ‘the effectiveness of hood ventilation varies markedly with the external ventilation conditions and air turbulence’. The physics of this problem are extremely complex and intractable, even with today’s supercomputers. Yet, in his paper, Roach brought together some of the classic solutions for diffusion equations that shed light on this difficult problem. He demonstrated shortcomings in the design concepts of capture velocity and face velocity, and provided some practical guidance on how to deal with the problem.
The paper begins with a theoretical development of concentration as a function of the contaminant generation rate, m, the distance from the point source, r, and the turbulent diffusivity, D. Drawing on the classic analogy between heat and mass transfer, Roach developed a series of simple differential equations, and their analytic solutions, for various problems of significance in occupational hygiene. Throughout the paper a constant value of the turbulent diffusion coefficient is assumed; Roach acknowledged that this was not the case in reality, but proceeded with the development accepting its limitations. Several interesting examples and useful conclusions resulted from this analysis.
First, Roach illustrated that diffusion resulted in the air concentration being inversely proportional to the distance from a point source, rather than to the square of distance as often assumed. Furthermore, theoretical developments illustrated how the location of air supply and extracts can enhance turbulent mixing and reduce concentration. Roach used a particularly interesting solution, combining uniform airflow and a point source of contaminant, to illustrate the effects of diffusivity on hood performance and to stress the counter effect that turbulence has on transporting the contaminant back against the flow. He used this idea to examine the effect of cross draughts on hood performance.
Through theoretical examples, laboratory experiments, and field studies, Roach emphasized two points.
1. The stationary concentration of contaminant is proportional to the contaminant generation rate.
2. Velocity alone is not a sufficient design criterion to control exposures.
He then proceeded to cite an alternative performance index, the equivalent ventilation rate, a term attributed to Lidwell (1960) that was defined as the ratio of the contaminant generation rate to the steady concentration at the point of interest (breathing zone). Roach referred to equivalent ventilation rate (on p. 119) as ‘the volume flow-rate of air in which all the released contaminant would have to be evenly mixed to produce the stationary concentration at the measuring point’. He then presented measurements from laboratory fume hoods to support the utility of this concept, using standard tracer gas equipment.
The equivalent ventilation rate is seldom identified as such today, although the terms ‘local purging flow rate’ and ‘local net flow rate’ are equivalent. Today, rapid real-time detectors make tracer gas studies more flexible and many different concepts (e.g. age distributions of air and contaminant) and measurements (decay curves) are possible. However, the fundamental relationship of tying a known contaminant generation rate to a measure of local concentration is a key concept that Roach developed in his paper from both theoretical and practical points of view.
Finally in the Discussion section (on p. 127) Roach offered the following advice to the practising hygienist:
...when a successfully controlled source of air contaminant has been established, it would be helpful to maintain records of at least six factors: the configuration of any exhaust hoods, their volume flow rates, a survey of the velocity fields around the hood, the ventilation provisions in the surroundings and the workshop configuration. A compilation of such practical information from different applications would go far in helping raise the design of ventilation arrangements from the level of an art to that of a science. The direct assessment of the concentration field arising from a known rate of release of contaminant is advocated wherever possible...
This statement embodies the true significance of Roach’s work. He used fundamental principles of science to tie together the important issues of exposure and engineering control, and pointed the way to a science for occupational hygiene. It is a shame that this sage advice, offered over 20 yr ago, has not received wider consideration in many of the applied studies published today. We often see exposure assessment or engineering evaluations that do not contain the ‘complete package of information’ that Roach provided for making intelligent decisions about control.
One measure of the value of a paper is the citations that subsequently reference it, the breadth of applications employing it, and the extent to which it provides an ongoing basis for further work. An on-line citation search identified 20 references to this paper from fields as diverse as veterinary medicine (Milligan and Sablan, 1982) and metal working (Averill et al., 1999). Authors in the fields of occupational hygiene and risk assessment, who have drawn on this work, include Nicas (Nicas, 1996, 2000, 2001) Flynn (Flynn and Ellenbecker, 1986; Flynn and Miller, 1988) and Jayjock (Jayjock and Hawkins, 1993; Shade and Jayjock, 1997). It is encouraging and significant to see that, after more than 20 yr, this work is still being referenced in the literature.
|
REFERENCES |
|---|
|
TOPABSTRACTA BRIEF SUMMARY OF...‘A MOST RATIONAL BASIS...‘ON THE ROLE OF...REFERENCES |
|---|
Averill AF, Ingham JM, Nolan PFA. (1999) A study of the dispersion of solvent vapour in the workspace during wipe cleaning of metal components with organic solvents—a Monte Carlo uncertainty analysis. T I Metal Finishing;77: 204–8.
British Occupational Hygiene Society. (1968) Hygiene standards for chrysotile asbestos dust. Ann Occup Hyg; 11: 47–69.[Medline]
Droz PO, Berode M, Wu MM. (1991) Evauation of concomitant biological and air monitoring results. Appl Occup Environ Hyg; 6: 465–74.
Flynn MR, Ellenbecker MJ. (1986) Capture efficiency of flanged circular local exhaust hoods. Ann Occup Hyg; 30: 497–513.
Flynn MR, Miller CT. (1988) Comparison of models for flow through flanged and plain circular hoods. Ann Occup Hyg; 32: 373–84.
Hewett P. (1997a) Mean Testing: I. Advantages and disadvantages. Appl Occup Environ Hyg; 12: 339–46.
Hewett P. (1997b) Mean testing: II. Comparison of several alternative approaches. Appl Occup Environ Hyg; 12: 347–55.
Hewett P. (1998) Re: A log-normal distribution-based exposure assesment method for unbalanced data. [Letter to the Editor.] Ann Occup Hyg; 42: 413–22.
Hickey JLS, Reist PC. (1977) Application of occupational exposure limits to unusual work schedules. Am Ind Hyg Assoc J; 38: 613–21.[ISI][Medline]
Jayjock MA, Hawkins NC. (1993) A proposal for improving the role of exposure modeling in risk assessment. Am Ind Hyg Assoc J; 54: 733–41.[ISI][Medline]
Kumagai S, Matsunaga I. (1994) Approaches for estimating the distribution of short-term exposure concentrations for different averaging times. Ann Occup Hyg; 38: 815–25.
Kumagai S, Matsunaga I, Kusaka Y. (1993) Autocorrelation of short-term and daily average exposure levels in workplaces. Am Ind Hyg Assoc J; 54: 341–50.[ISI][Medline]
Leidel NA, Busch K, Lynch J. (1977) Occupational Exposure Sampling Strategy Manual. NIOSH 77-173. Washington, DC: US Government Printing Office.
Lidwell OM. (1960) The evaluation of ventilation. J Hyg Camb; 58: 297–305.[Medline]
Milligan JE, Sablan JL. (1982) Waste anesthetic gas concentrations in a veterinary recovery room. J Am Vet Med Assoc; 181: 1540–1.[ISI][Medline]
Nicas M. (1996) Estimating exposure intensity in an imperfectly mixed room. Am Ind Hyg Assoc J; 57: 542–50.[ISI][Medline]
Nicas M. (2000) Markov modeling of contaminant concentrations inindoor air. Am Ind Hyg Assoc J; 61: 484–91.
Nicas M. (2001) Modeling turbulent diffusion and advection of indoor air contaminants by Markov chains. Am Ind Hyg Assoc J; 62: 149–58.
Ogden TL. (2003) Commentary: The 1968 BOHS Chrysotile Asbestos Standard. Ann Occup Hyg; 47: 3–6.
Oldham P. (1953) The nature of the variability of dust concentrations at the coal face. Br J Ind Med; 10: 227–34.
Oldham P, Roach SA. (1952) A sampling procedure for measuring industrial dust exposure. Br J Ind Med; 9: 112–9.
Rappaport S. (1988) Biological considerations for designing sampling strategies. In John W, editor. Advances in Air Sampling. Chelsea, MI: Lewis. pp. 337–52.
Rappaport SM. (1985) Smoothing of exposure variability at the receptor: implications for health standards. Ann Occup Hyg; 29: 201–14.
Rappaport SM. (1991) Assessment of long-term exposures to toxic substances in air. Ann Occup Hyg; 35: 61–121.
Rappaport SM, Spear RC. (1988) Physiological damping of exposure variability during brief periods. Ann Occup Hyg; 32: 21–33.
Rappaport SM, Lyles RH, Kupper LL. (1995) An exposure-assessment strategy accounting for within- and between- worker sources of variability. Ann Occup Hyg; 39: 469–95.
Rappaport SM, Kupper LL, Lyles RH. (1998) Re: A log-normal distribution-based exposure assessment method for unbalanced data. [Letter to the Editor.] Ann Occup Hyg; 42: 417–22.
Rappaport SM, Weaver M, Taylor D, Kupper L, Susi P. (1999) Application of mixed models to assess exposures monitored by construction workers during hot processes. Ann Occup Hyg; 43: 457–69.
Roach SA. (1953) A method of relating the incidence of pneumoconiosis to airborne dust exposure. Br J Ind Med; 10: 220–6.
Roach SA. (1959) Measuring dust exposure with the thermal precipitator in collieries and foundries. Br J Ind Med; 16: 104–22.
Roach SA. (1966) A more rational basis for air-sampling programs. Am Ind Hyg Assoc J; 27: 1–12.[ISI][Medline]
Roach SA. (1968) The theory of random clumping. London: Methuen.
Roach SA. (1977) A most rational basis for air sampling programmes. Ann Occup Hyg; 20: 65–84.
Roach SA. (1981) On the role of turbulent diffusion in ventilation. Ann Occup Hyg; 24: 105–32.
Roach SA. (1992) Health risks from hazardous substances at work. Oxford: Pergamon Press.
Roach SA, Rappaport SM. (1990) But they are not thresholds: a critical analysis of the documentation of Threshold Limit Values. Am J Ind Med; 17: 727–53.[ISI][Medline]
Shade WD, Jayjock MA. (1997) Monte Carlo uncertainty analysis of a diffusion model for the assessment of halogen gas exposure during dosing of brominators. Am Ind Hyg Assoc J; 58: 418–24.[ISI][Medline]
Symanski E, Kupper LL, Rappaport SM. (1998) Comprehensive evaluation of long-term trends in occupational exposure: Part 1. Description of the database. Occup Environ Med; 55: 300–9.[Abstract]
Symanski E, Sallsten G, Barregard L. (2000) Variability in airborne and biological measures of exposure to mercury in the chloralkali industry: implications for epidemiologic studies. Environ Health Perspect; 108: 569–73.[ISI][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  D. Kriebel, H. Checkoway, and N. Pearce Exposure and dose modelling in occupational epidemiology Occup. Environ. Med., July 1, 2007; 64(7): 492 - 498. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. CARTER British Occupational Hygiene Practice 1720-1920 Ann. Hyg., June 1, 2004; 48(4): 299 - 307. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||