This paper introduces a novel decision-making framework for the optimization of water treatment plant operations. Managers at water utilities face increasing tensions between cost, public health risk, public perception, and regulatory compliance. Multi-objective optimization techniques have been developed to generate innovative solutions to environmental problems with competing objectives. By integrating these optimization techniques with water quality scenarios, water treatment modeling, and interactive visualization, our framework enables water managers to choose among an ensemble of optimal treatment operations. By automating the generation of treatment options, this paradigm represents a shift toward exploration and insight discovery in drinking water decision making. To illustrate this framework, we create a disinfection byproduct (DBP) management problem that incorporates the influence of competing risks and cost objectives on decision making. Using data from the Cache la Poudre River—a source water in Colorado with seasonally-varying water quality—and a hypothetical conventional treatment plant, we evaluate the impact of organic carbon increases on the performance of optimal treatment operations. These results suggest that the hypothetical utility should consider infrastructural improvements if organic carbon concentrations increase more than approximately 25% of maximum historical levels. An interactive exploration of the optimization results reveal to what extent there are tradeoffs between solids handling costs, chemical costs, and DBP exposure. A k-means clustering of these data illustrates that the utility can achieve compliance through a variety of treatment strategies depending on decision maker preferences for cost and risk.