Comparing Two-Zone Models of Dust Exposure

Abstract

The selection and application of mathematical models to work tasks is challenging. Previously, we developed and evaluated a semi-empirical two-zone model that predicts time-weighted average (TWA) concentrations (C_twa) of dust emitted during the sanding of drywall joint compound. Here, we fit the emission rate and random air speed variables of a mechanistic two-zone model to testing event data and apply and evaluate the model using data from two field studies. We found that the fitted random air speed values and emission rate were sensitive to (i) the size of the near-field and (ii) the objective function used for fitting, but this did not substantially impact predicted dust C_twa. The mechanistic model predictions were lower than the semi-empirical model predictions and measured respirable dust C_twa at Site A but were within an acceptable range. At Site B, a 10.5 m³ room, the mechanistic model did not capture the observed difference between PBZ and area C_twa. The model predicted uniform mixing and predicted dust C_twa up to an order of magnitude greater than was measured. We suggest that applications of the mechanistic model be limited to contexts where the near-field volume is very small relative to the far-field volume.

[Supplementary materials are available for this article. Go to the publisher's online edition of the Journal of Occupational and Environmental Hygiene for the following free supplemental resource: a PDF file containing tables giving data for fitted emission rates and random air speeds for testing events with measured airflow rates for both near- and far-field zones and near-field zone only, and figures showing fitted random air speed varies with near-filed geometry when fitted with objective functions 1 and 2.]

Keywords:

• emission rate
• joint compound
• mathematical models
• random air speeds
• respirable dust

ACKNOWLEDGMENTS

This work was funded by Georgia-Pacific (Atlanta, Ga.). Georgia-Pacific did not direct or review the field studies, modeling work, or manuscript, or influence the decision to publish this analysis.