Abstract
While the presence of microcystin-LR (MC-LR) in raw water from eutrophic reservoirs poses human health concerns, the risks associated with the ingestion of MC-LR in drinking water are not fully elucidated. We used a time series of MC-LR in raw water from tropical urban reservoirs in Brazil to estimate the hazard quotients (HQs) for non-carcinogenic health effects and the potential ingestion of MC-LR through drinking water. We considered scenarios of MC-LR removal in the drinking water treatment plants (DWTPs) of two supply systems (Cascata and Guarapiranga). The former uses coagulation/flocculation/sedimentation/filtration/disinfection, while the latter has an additional step of membrane ultrafiltration, with contrasting expected MC-LR removal efficiencies. We considered reference values for infants (0.30 μg L
−1
), children/adults (1.60 μg L
−1
), or the population in general (1.0 μg L
−1
). For most scenarios for Cascata, the 95% upper confidence level of the HQ indicated high risks of exposure for the population (HQ > 1), particularly for infants (HQ = 30.910). The water treatment in Cascata was associated to the potential exposure to MC-LR due to its limited removal capacity, with up to 263 days/year with MC-LR above threshold values. The Guarapiranga system had the lowest MC-LR in the raw water as well as higher expected removal efficiencies in the DWTP, resulting in negligible risks. We reinforce the importance of integrating raw water quality characteristics and treatment technologies to reduce the risks of exposure to MC-LR, especially for vulnerable population groups. Our results can serve as a starting point for risk management strategies to minimize cases of MC-LR intoxication in Brazil and other developing countries.