Demonstration of a modular computational tool for human health risk assessment due to contaminated sites or emissions from facilities
Résumé
In France, risk assessment studies have to be conducted before the implementation or the enlargement of some new facilities and before the reuse of some contaminated sites. These studies have to be performed in accordance to four principles : precautionary, specificity, proportionality, transparence. Until now, no devoted software has been developed in France to assess future exposure and risks for such studies. Because of that, many problems have been reported in exposure assessment studies leading to inconsistency between risk assessment studies and a lack of confidence in results provided. To improve the practices and the transparency of the estimates obtained in these studies, INERIS develops and diffuses modeling tools in the framework of one's missions for the Ministry in charge of the Environment. A peer-reviewed handbook, entitled 'Sets of equations for modeling exposure linked to soil contamination or emissions from an industrial facility', has been published and is available on INERIS'website. This document presents the equations used at INERIS for estimating the media concentrations, the exposure and risk levels. It also describes the origin of these equations and underlines the hypotheses on which they are built and their limits. A modeling and simulation platform (MODUL'ERS) based on the equations presented in this handbook has also been developed. Our main objectives during its conception were to provide a tool (1) suited to different site conditions and tier studies, (2) transparent for any stakeholders and helpful to perform uncertainties analysis. MODUL'ERS consists in a library of preset modules enabling the users to build models in accordance with the site conceptual model (that is to say the pathways from the source to the receptor), by downloading modules and connecting them, to create an exposure matrix. Many options are also available to create a customized application. To improve transparency of studies all the equations, parameters can be viewed by the users, as well all the intermediate calculations performed. Especially, hyperlinks enable to browse among variables and their equations. The first version of the tool will be available at soon. Coupling with a GIS is forecast. Some reports will be also published to make more transparent the assignment of values to input parameters, by describing the values collected and the choices made (best-estimate, ranges, probabilistic distributions).