An integrated approach for modeling uncertainty in human health risk assessment.

Abstract

A human health risk assessment (HHRA) will not remain simple and straightforward when it involves multiple uncertain input variables. Uncertainties in HHRA result from the unavailability and subjectivity of input variables. Though several studies have performed HHRA, the quantification of uncertainty in HHRA under a situation of data scarcity and the simultaneous application of random and non-random input variables have rarely been reported. The present study proposes an integrated hybrid health risk modeling framework involving the concurrent treatment of random and non-random input variables and estimating the uncertainties linked to the input variables in HHRA. The proposed framework presents the flexibility to classify the input variables into fuzzy and probabilistic categories, based on their data availability and provenience nature. The framework is demonstrated over the Turbhe sanitary landfill in Navi Mumbai, India, where the fate and transport of heavy metals in leachate are investigated through LandSim modeling. The present study considers the LandSim-simulated heavy metal concentration and body weight as a random variable and water intake, exposure duration, frequency, bioavailability, and average time as fuzzy variables. Further, the uncertainties in the non-carcinogenic human health risk have been quantified using Monte Carlo simulations, followed by a comprehensive multivariate sensitivity analysis of the proposed framework. High health risk at Turbhe is estimated for the male and female population. This study presents the first effort to quantify the non-carcinogenic human health risks from leachate-contaminated groundwater considering the health risk input variables as non-deterministic. The proposed framework is generic and applicable to any landfill site and will remain unaltered when integrated health risk assessment and uncertainty assessment are performed for the landfill.