Design and Optimize Advanced Oxidation Processes: The AMOZONE Model for Drinking Water Production

Abstract: Ensuring safe and stable water quality already at the design stage, and having possible further improvements in performance during operation, is nowadays gaining importance for utilities having to deal with increasingly stringent legislation's requirements. This work reports how AMOZONE kinetic model supported the design improvement of an ozone installation in France (integrating the kinetics with CFD), and the decision making in the operation of a peroxone-UV/H2O2 plant in The Netherlands for long term simulations. Keywords: Ozone; Disinfection; UV Drinking water production is a pressing and delicate concern, especially with the increasing concentrations of micropollutants resulting from the different increasing human activities (e.g. agriculture, pharma, cosmetics, ...). At the same time, both energy requirements and systems complexity increase with the stringent requests around the produced water quality and safety. Advanced Oxidation Processes can be combined to achieve required targets based on the water composition. However, the increasing complexity of chemical reactions involved and the risk of more toxic by-products formation from a non-ideal AOP application, require modelling tools as essential means to answer key questions for solid design and operational decisions. This work presents two applications of the AMOZONE model (Audenaert et al., 2019) for i) re-designing the ozonation reactors of an existing drinking water installation in France for minimal bromate formation and maximal disinfection, and ii) optimizing the operation of a full-scale peroxone UV-AOP installation (2.200 m3/h) in The Netherlands to minimize bromate formation and energy use for drinking water production after dune injection for groundwater well recharge. In first case (France), the AMOZONE kinetic model was integrated in CFD in the revamping process of an inter-ozonation tank for disinfection of drinking water in France. In the second case (The Netherlands), the same model was used in flow-sheet mode to optimize long-term operation of both the peroxone and UV/H2O2 reactors for targeting the largest micropollutants removal with minimal bromate formation. The AMOZONE model was calibrated and validated with real data from both plants reported in this study in terms of ozone residual concentration and bromate formation. The ozonation reactor (France) was designed with increased performance in terms of Contact Time (CT), Mass Transfer Efficiency (MTE) and Bromate (Figure 1, top). The peroxone-UV/H2O2 reactor (The Netherlands) was represented in a flowsheet model to run long-term simulations (Figure 1, bottom). Thanks to the better hydrodynamic behavior, the CFD-AMOZONE allowed significant improvements of the CT (at the same ozone dose) and even improved reducing the bromate risk (Figure 2). With the AMOZONE flowsheet model, long-term simulations were run to optimize the peroxone-UV/H2O2 reactor for targeting the largest micropollutants removal within the peroxone reactor while minimizing bromate formation and UV lamps usage. Figure 3, shows the ideal combination of usage of both AOP technologies as compared to the long trial and error. The AMOZONE model is nowadays largely used to answer design and operational questions on AOPs. The integration of the AMOZONE kinetic model with CFD allowed to refine the detailed design of an ozone-peroxide reactor making disinfection effective at all different expected flows. The long-term dynamic simulations of a peroxone- UV/H2O2 installation allowed to understand the possible improvement in the operation to minimize Opex while maintaining safe and stable water quality.