Development of methods to estimate microcystins removal and water treatment resiliency using mechanistic risk modelling.

Drinking water treatment processes are capable of removing microcystins but consistent operation of processes optimized for cyanobacterial harmful algal bloom (cHAB) conditions is not fiscally feasible. Therefore, utilities must ready themselves and start the cHAB processes as a reactionary response. Predictive analytics and modelling are impactful tools to prepare water systems for cHABs, but are still in early stages of development. Until those prospective models are completed, a method to determine best actions in advance of a bloom event thus improving system resiliency is needed. In this study, an adaptation of the quantitative microbial risk analysis (QMRA) methodology was applied to develop this method. This method and resulting models were developed around the Toledo (Ohio, USA) water crisis of 2014, but being mechanistic, they are easily adaptable to other systems' process operations data. Results from this internally validated model demonstrate how rapid action using both powdered activated carbon and measured increases in chlorine dose can mitigate health risks. Our research also demonstrates the importance of modelling the cellular status of the toxins (toxins either in an intact cell or in the water from a lysed cell). Risks were characterized using hazard quotients (HQ) and at the peak of the crisis ranged from a minimum of 0.00244 to a maximum of 2.84 for adults. In simulations where cHAB-specific treatment was used this decreased to 0.00057 and 0.236 respectively. We further outline how this methodology can be used to simulate water system resiliency to likely and aberrant microbial hazard events to plan for the best interventions to protect public health. This method can be used for other hazards expected to be variable in the future, where system prepatory planning is critical to continued public health protection. Considering the water quantity and quality fluctuations occurring and likely to intensify under climate change, this type of computationally supported preparedness is vital to maintaining robust water system resiliency.