Do EASE Scenarios Fit Workplace Reality? A Validation Study of the EASE Model

Abstract

Within the framework of European risk assessment of new and existing substances, the EASE model (Estimation and Assessment of Substance Exposure) is often applied to assess inhalative exposure at workplaces. To contribute to the validation of this model, single EASE scenarios were compared with independent measurement data on inhalative exposure to vapors and dusts. For this purpose, workplace measurements obtained in the areas of production, textile printing and coating, screen printing and offset printing, the rubber industry, and the plastics processing industry as well as flame spraying were used. A good correspondence between model estimates and measurement data was found in the case of exposure to vapors for the production of chemical substances in closed systems (EASE scenario: closed system without breaching) and for procedural control works at mainly automated printing workplaces (EASE scenarios: non dispersive use, segregation, low/medium volatility). For the handling of powdery substances measurement results obtained during weighing and filling works, in part semiautomatic, were compared with the EASE estimates for the scenario: dry manipulation, with/without local exhaust ventilation. Good correspondence was estimated if approx. 1 t to 200 t powdery substances were handled per shift. For the EASE scenarios "direct handling with natural ventilation/with local exhaust ventilation," high discrepancies exist between model prediction and measurement results obtained during manual screen printing or the preparation of printing inks.

Chemical Exposure and Risk Assessment at Workplaces--Modeling Approach

“…The level of concern and the technical control action needed are probably the same whether the dust exposure is, for example, 30 or 100 mg/m 3 . Bredendiek-Kämper (13) noticed that the amount of substance handled is an important determinant for exposure to dusts as well as to vapors, but the EASE model does not take that into account. If the amount handled could be added to the model parameters, it would be possible to increase the exposure categories and enhance the relevance of the estimation.…”

“…The further validation of EASE also requires more well-documented, highquality measurement data. (13) When the Risk-EASE model (combination of risk assessment model BS 8800 and quantitative data on concentrations [EASE]) was used, the risk was, in most cases, categorized higher than when the original BS 8800 model was used alone. Expert judgment gave results similar to Risk-EASE.…”

Evaluation of the Predictive Abilities of a Qualitative Exposure Assessment Model

“…These results were also reported in the evaluation of other models. (5,10,11) Other authors have reported that the performance of predictive models may be affected by the variability of the data used for the comparison. (1,3) To test if this may have been the case in this study, the correlation and Fisher exact test were repeated for the mean and 90th percentile of the exposure data.…”

“…The percentage classification ranged from 0% to 100%; if worker groups with four or more chemicals are considered, correct classifications ranged from (5,10,11) George E.P. Box once stated that, "All models are wrong, but some are useful."…”

“…In a semiqualitative comparison of 11 exposure scenarios, Bredendick-Kämper (10) stated that there was agreement between the EASE prediction and measurement data if the 25th percentile and 75th percentile of the data for an exposure scenario overlapped the EASE endpoint range. Using this criterion, the author concluded that the EASE predictions were in good agreement in four of six industry groups, whereas they were poor predictors in two others.…”