Evaluation of Recommended REACH Exposure Modeling Tools and Near-Field, Far-Field Model in Assessing Occupational Exposure to Toluene from Spray Paint

Abstract

Predictive modeling is an available tool to assess worker exposures to a variety of chemicals in different industries and product-use scenarios. The European Chemical Agency (ECHA)'s guidelines for manufacturers to fulfill the European Union's legal requirements pursuant to the Registration, Evaluation, Authorization, and Restriction of Chemicals (REACH) initiative include recommendations for the use of modeling to predict worker exposures. ECHA recommends different models for different target populations (i.e. workers, consumers, environment) and routes of exposure (i.e. skin absorption, ingestion, inhalation), and presents them hierarchically, with Tier 1 models presented as the most simplistic, conservative models and Tier 2 models recommended for further intensive evaluation of substances or preparations. In order to assess these models for one exposure (product-use) scenario, a simulation of the scenario was completed in a controlled environment and the measured results were compared with the modeling outputs. The authors predicted, based on the design of the modeling tools, that all models would overestimate worker exposures under the simulated product-use scenario, with the lower-tiered model producing the most conservative estimate of exposure. In this study, a Tier 1 model and a Tier 2 model were evaluated for comparison with the near-field, far-field (NF-FF) deterministic model and measured experimental results in a real-time worker inhalation exposure assessment. Modeling was conducted prior to actual air monitoring. The exposure scenario that was evaluated involved the application of a toluene-containing spray paint to a work surface. Air samples were collected to evaluate short-term (15-min) and long-term (240-min) exposures. Eight-hour time-weighted averages (8-h TWAs) were calculated and compared with the modeling outputs from the recommended REACH modeling tools and the NF-FF model. A comparison of each of the modeling tools with measured experimental results was generated. The Tier 1 Targeted Risk Assessment tool overestimated the 8-h TWA airborne concentration of toluene in the spray scenario by a factor of 3.61. The higher tiered Advanced REACH Tool and NF-FF models showed greater concordance with experimental results, overestimating the TWA exposure by a factor of 2.92 and 1.96, respectively. In conclusion, the Tier 1 and 2 exposure modeling tools performed as expected for the simulated exposure scenario, providing relatively accurate, though conservative, estimates according to the level of detail and precision accounted for in each model.

Chemical Risk Assessment Screening Tool of a Global Chemical Company

“…An evaluation in 2011 [17] showed that ART could estimate with 90% confidence geometric mean exposure levels within a factor between two and six of the measured geometric mean exposure levels for levels for dusts, mists, and vapors. Two validation studies using independent measurement series indicated that ART estimates were within the uncertainty ranges found in the calibration [18] , [19] .…”

How to Obtain a Reliable Estimate of Occupational Exposure? Review and Discussion of Models’ Reliability

“…Liquid scenarios were considered [26,32,44], as were vapors and mists [30,45,46] and volatile substances [7,22,27,30,43]. Other chemicals were evaluated by Landberg et al [14], and others evaluated benzene [24], ethylbenzene [33], toluene [3,37,42], ethyl acetate [36,39], and acetone [36]. Petroleum substances were investigated by Hesse et al [12], including kerosene, heavy fuel oils, the naphtha-2 group, gas oils, and other lubricant base oils.…”

A Near Field–Far Field model for assessing Oxygen Deficiency Hazard

“…Note that these two fields are a subdivision of the working environment, and the sum of their volumes is the overall volume of the working environment. In the literature there are several examples of application of this model to estimate the concentration of different chemicals: isoflurane (Sakhvidi et al, 2013), benzene (Nicas et al, 2006), solvent (Spencer and Plisko, 2007), methanol vapours (Gaffney et al, 2008), dusts (Jones et al, 2011), sulfur hexafluoride (Furtaw Jr. et al, 1996), laser-generated particulate matter (Lopez et al, 2015), cleaning products (Earnest and Corsi, 2013), toluene (Hofstetter et al, 2013), and unspecified substance (Feigley et al, 2002). One of these studies (Nicas et al, 2006) predicts concentrations using a non-constant emission rate, as done also in other papers: for example, Nicas and Armstrong (2003b) (a spreadsheet to compute a sine function emission rate), Nicas and Neuhaus (2008) (a formulation valid in the case of a variable emission rate), Nicas and Armstrong (2003a) (Excel spreadsheets and a Matlab code for studying the two-zone model with a constant emission rate and an exponentially decreasing contaminant emission rate), and Nicas (2016) (a revisited study of Nicas and Neuhaus (2008) with constant application of chemical mass and exponentially decreasing emission of the mass applied).…”