Atlanta, GA: U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, R01-OH-007626, 2008 Mar; :1-97
The ability to estimate worker exposure is essential for evaluating workplace hazards and protecting workers. In research, however, exposure assessment is often the weakest element in examining the relationship between contaminant exposure and occupational disease; thus, the development and improvement of exposure estimation models and methods is extremely important. Here experimental and mathematical methods were used to explore important determinants of exposure to airborne contaminants, particularly worker presence and activity. This research addresses the inherent challenge presented by the variation of concentration with workroom location. The effects of five factors - air flow rate, temperature, air inlet type, worker location, and worker activity - on contaminant distribution and worker exposure were investigated. Mathematical models for exposure estimation were evaluated including simple deterministic models, zonal models, and computational fluid dynamics (CFD). CFD simulations were used to investigate the effects of physical factors and the performance of simple deterministic models, and CFD estimates were compared with measured contaminant concentrations in a manufacturing work area. Methods. The study was conducted in an experimental room with mixing ventilation and a tracer gas was injected at a constant rate. Concentration was monitored at 144 points with a photoionization detector attached to an automated sampling system. The research was designed to use three constant dilution air flow rates, 5.5, 3.3, and 0.88 m3/min; but use of the lowest air flow rate was not feasible due to excessive monitoring time requirements. The number of sampling points required to characterize concentration distribution within the experimental room was determined by comparing concentration isopleths from subsets of data containing from 100 to 150 points. No substantial differences in isopleth shape or magnitude were observed over this range. Hence, a 3-dimensional network of 144 points was utilized. Two air inlets types were studied: a wall jet (WJ) with its center 2.12 m above the floor and a vaned diffuser in the center of the ceiling (CD), just above the tracer gas source. To simulate temperature variability within a workroom, one wall of the experimental room was heated or cooled to represent a building's external wall. A heated mannequin was used to investigate the impact of a stationary worker's presence, and a human participant was used to simulate a moving worker. The contaminant concentration outside the facepiece of an air-supplied respirator was measured at various locations and orientations, and for various activities. Results. Prior to studying the impact of a worker's presence on concentration fields, it was necessary to assess the effects of physical factors. Experiments were performed at two air flow rates (5.5 and 3.3 m3/min) and six thermal conditions: isothermal, three summer conditions and two winter conditions. For comparing rooms with different sizes and flow rates but similar physical configurations, the dimensionless room Reynolds number (Re) corresponding to the two flow rates was used. The Re corresponding to the two air flow rates used here are 2,100 and 1,200. The variability of contaminant concentration at the higher flow rate was not affected by thermal conditions; but, at the lower flow rate, winter conditions produced greater variability (coefficient of variability, CV = 0.72 and 1.10) than isothermal and summer conditions (CV = 0.29 to 0.34). Tests simulating winter conditions suggested that the resulting stable temperature structure inhibited the dilution of the tracer and enhanced its segregation in the lower portion of the room, especially for the lower flow rate. A worker was located at four different positions near the source, and experiments were performed to study the effect of the worker's presence on contaminant dispersion throughout the experimental room for two air flow rates and two air inlet types. Only small differences in the overall room mean concentrations were observed when the worker was absent versus when the worker was present. However, nine out of ten experimental factor combinations showed that the pollutant dispersion patterns in an occupied room depended upon the location of the worker. For these experiments, the ceiling diffuser inlet was found to be more efficient than the wall jet in diluting contaminant, resulting in a mean reduction of 11% reduction in the overall room contaminant concentration, calculated by averaging the contaminant at all sampling points for each experiment. Very high concentrations were limited to a small volume immediately above the source when the ceiling diffuser was used, and the rest of the room was virtually well mixed. Also, the concentration at one fixed monitoring location was generally higher (on average 8% and 44% higher for flow rates 5.5 and 3.3 m3/min, respectively) when the worker was stationary or absent than when the worker moved along a fixed path. The effects of location and orientation of a stationary worker on the worker's exposure were assessed. For three of four flow rate-location combinations, a stationary worker near the tracer gas source (breathing zone 0.40 m horizontally from the source) was exposed to higher concentrations than the concentrations observed at that location when no worker was present. Average exposures were higher when the worker was facing the source. This finding under mixing ventilation was similar to the effect of worker orientation reported from wind tunnel experiments or at hood faces, but smaller in magnitude. Also, the tracer concentration encountered by a worker moving along a fixed path and the concentration along that path when no worker was present differed by less than 5%. The effects of inlet, exhaust and source locations and of room dimensions on the flow field and contaminant distribution were studied by CFD simulation. Results were used to explore the optimal values for the size of the near-field zone and the air exchange rate between the zones for a simple two-zone mathematical model. The analysis showed that the optimum near-field zone size varied with room configuration and was in the range of 8.5% to 20% of the room volume for the conditions of these simulations. Coarse-grid CFD (CFD with a very small number of cells) and a new multi-zone model were also tested. Noting that accuracy depended upon numerous physical factors and their interactions, we focused on zonal models which recently had been adapted for use within single enclosed spaces. Zonal models, like CFD, divide a room into separate zones, and simultaneously solve a set of linear equations for conservation of mass and energy for all zones. Unlike CFD, zonal models do not incorporate the equations for conservation of momentum, but compensate by adding empirical terms to describe the penetration of air jets entering a room. Empirical jet equations have been validated for a limited number of room configurations and physical factors. In this research, the zonal models tested gave inaccurate results and were judged to be inappropriate for describing the details of transport within rooms; instead, other approaches such as CFD need to be used. To test CFD in an actual workplace, a capacitor manufacturing facility was surveyed extensively and the concentration of isoamyl acetate (IAA) was simulated using CFD. After careful analysis to determine the source boundary conditions for IAA emission, CFD concentration estimates agreed very well with observations at the six locations in the source near-field: a two-tailed, paired t-test found no significant difference between the CFD concentration estimates and the measured values (p= 0.92). Thus, we concluded that a very standard CFD model yielded accurate simulations of dispersions, provided that adequate efforts were made to define realistic boundary conditions. Additional research is needed to develop methods for easily and accurately obtaining boundary conditions for enclosed spaces.
Charles E. Feigley, Department of Environmental Health Sciences, Arnold School of Public Health, Room 401, University of South Carolina, 921 Assembly Street, Columbia, SC 29208