A New Model Algorithm for Estimating the Inhalation Exposure Resulting from the Spraying of (Semi)-Volatile Binary Liquid Mixtures (SprayEva)

Abstract

The spraying of liquid multicomponent mixtures is common in many professional and industrial settings. Typical examples are cleaning agents, additives, coatings, and biocidal products. In all of these examples, hazardous substances can be released in the form of aerosols or vapours. For occupational and consumer risk assessment in regulatory contexts, it is therefore important to know the exposure which results from the amount of chemicals in the surrounding air. In this research, a mechanistic mass balance model has been developed that covers the spraying of (semi)-volatile substances, taking into account combined exposure to spray mist, evaporation from droplets, and evaporation from surfaces as well as the nonideal behaviour of components in liquids and backpressure effects. For wall-spraying scenarios, an impaction module has been developed that quantifies the amount of overspray and the amount of material that lands on the wall. Mechanistically, the model is based on the assumption that continuous spraying can be approximated by a number of sequentially released spray pulses, each characterized by a certain droplet size, where the total aerosol exposure is obtained by summation over all release pulses. The corresponding system of differential equations is solved numerically using an extended Euler algorithm that is based on a discretisation of time and space. Since workers typically apply the product continuously, the treated area and the corresponding evaporating surface area grows over time. Time-dependent concentration gradients within the sprayed liquid films that may result from different volatilities of the components are therefore addressed by the proposed model. A worked example is presented to illustrate the calculated exposure for a scenario where aqueous solutions of H2O2 are sprayed onto surfaces as a biocidal product. The results reveal that exposure to H2O2 aerosol reaches relevant concentrations only during the spraying phase. Evaporation from sprayed surfaces takes place over much longer time periods, where backpressure effects caused by large emission sources can influence the shape of the concentration time curves significantly. The influence of the activity coefficients is not so pronounced. To test the plausibility of the developed model algorithm, a comparison of model estimates of SprayExpo, SprayEva, and ConsExpo with measured data is performed. Although the comparison is based on a limited number (N = 19) of measurement data, the results are nevertheless regarded as supportive and acceptable for the plausibility and predictive power of SprayEva.