Advancing microbial risk management through strategic integration of risk assessment and water supply operations

Abstract

Controlling risks posed by microbial contaminants is a key focus for the provision of safe drinking water. Increasing threats to water supply systems—such as rapid water quality changes associated with climate-shocks (e.g., wildfires, heavy precipitation, etc) as well as infrastructure failures and treatment deficiencies—increasingly challenge microbial risk management and underscore the need for resilience in safe water supply. The goal of this research was to advance drinking water microbial risk management through integration of risk assessment approaches and real-world water supply treatment operations scenarios. Quantitative and qualitative risk-based frameworks were developed to address gaps in knowledge and practice related to monitoring, treatment decision-making, risk assessment and treatment operations. Routine pathogen monitoring programs and the use of monitoring data in risk assessment for treatment decision making were addressed using data from a full-scale water treatment plant. Practical guidance was provided about monitoring and associated risk-based treatment requirements including methods for choosing sampling locations and frequency. Policy contributions involved definition of treatment sufficiency compliance rules and evidence for inclusion of Giardia in protozoa monitoring programs. This work underscored that data collection should be tailored to best represent water entering treatment plants so their use in risk assessment can best inform treatment decisions and maximize return on investment. While this may seem obvious, how to best represent system-specific attributes in monitoring design and implementation is often far from obvious. To shift from routine operations to extreme events, a scenario-based quantitative microbial risk assessment (QMRA) framework was applied in a case study involving a wastewater spill upstream of a drinking water intake, in which urgent risk-informed treatment decisions were needed. Conducting comprehensive and mathematically complex QMRA is not practical when time-sensitive decisions are needed. Thus, the simplified framework developed here demonstrated how preliminary risk assessment can be conducted for short-term, rapid source water quality degradation events and used to inform operational decisions and improve drinking water system response to adverse events. Filtration plays a central role in microbial risk management during drinking water treatment. Although filtration plants include multiple filters, risk is typically evaluated based on individual filter performance. A model was developed to evaluate how dynamic, concurrent operation of multiple filters affects plant-scale filtration performance and thus microbial risk. It was demonstrated that operational response strategies should prioritize improving performance when it is “low” rather than maximizing performance when it is already “high” because microbial treatment performance is typically expressed on a logarithmic scale. Practical implications of several design or operational decisions for microbial risk management such as the number of filters in operation, backwash staggering schedules and filter-to-waste operation were presented. Plant-scale risk assessment is essential to microbial risk management because risk is driven by the collective and dynamic operation of multiple treatment units. Yet, current practice and regulatory guidance do not always acknowledge plant scale considerations. Linking several aspects of water supply integrally related to risk, the effect of dependence between model inputs in drinking water QMRA was then evaluated. Hypothetical scenarios and data from a full-scale water treatment plant were used to demonstrate that ignoring dependence can over- or underestimate risk, sometimes substantially (e.g., more than an order of magnitude), and mislead decision-making. A framework was developed for characterizing the potential for biased interpretation of risk when assumed independence that is typical in QMRA practice is not valid. Finally, the themes and insights described above were connected to more broadly highlight the inextricable linkage between microbial risk assessment and resilience in safe drinking water supply. The concept of operational resilience was defined as a water supply system’s capacity to limit the risk consequences of insufficient treatment performance when adverse events perturb it. Then, hypothetical case studies were used to demonstrate that resilience description must explicitly reflect risk to be an actionable tool. These foundations can be used to develop tools to operationalize resilience in the future. In addition to contributions to advancing risk and resilience sciences and drinking water policies, this research provides actionable knowledge to water utilities for improving decision-making during challenging conditions, assessing microbial risks more accurately and increasing the resilience of treatment operations.