|dc.description.abstract||A reliable and clean supply of drinking water is essential to the health and quality of life. The treatment of contaminants has been a major water quality concern for drinking water supply system providers. Manganese, one of the main regulated inorganic contaminants, is not commonly identified as a health concern in drinking water, but it can cause detrimental aesthetic water issues such as poor taste, odour and colour. Specifically, total manganese concentrations at or more than 0.05 mg/L can cause issues for drinking water providers by causing customer dissatisfaction, manifesting as complaints about staining of plumbing fixtures and of laundry, discoloured water, and taste and odour issues. In addition, out of consideration for the protection of the water distribution system, manganese in drinking water must be maintained at less than 0.02 mg/L so that it does not cause coating of manganese oxides on the internal surface of pipes.
Effective and timely treatment of manganese first requires a comprehensive understanding and accurate prediction of the manganese cycle in reservoirs since spikes in raw water manganese can challenge the capacity of water treatment plant (WTP) operators to promptly adjust the treatment procedures accordingly. In the present study, the Tarago Reservoir, Melbourne, Australia (a monomictic water supply reservoir), may experience manganese increases during reservoir turnover events, which can lead to elevated manganese concentrations in treated water. The vertical profiling system (VPS) installed in the reservoir and weekly field water sampling tests in both the reservoir and the WTP collect key parameters such as water temperature, dissolved oxygen and total and soluble manganese concentrations, which provide a basis for the data analysis and modelling study.
One of the main processes in a reservoir’s manganese cycle consists of the exchanges of soluble manganese and insoluble manganese between the bottom sediments and the epilimnion, and is highly susceptible to the hydrodynamic processes. Therefore, the hydrodynamic characteristics of water-supply reservoirs is fundamental to simulate the manganese cycle. Numerical deterministic models have been widely applied for simulating the hydrodynamic characteristics of various waterbodies. The developed three-dimensional (3D) numerical model of the Tarago Reservoir can quantify the reservoir thermal structure and facilitate investigation of the effects of seasonal and longitudinal factors. The model revealed that wind forces and river plumes play a vital role in the formation of stratification and turnover. Then, the developed 3D hydrodynamic model was coupled with an innovative manganese cycle model and applied to the Tarago Reservoir and was able to simulate the distribution of manganese in the reservoir. The 3D manganese model was developed based on an existing one-dimensional manganese cycle model. The stratification and turnover occurring in the reservoir significantly influenced the seasonal change in manganese levels, and the horizontal advection dominated by the wind-driven currents considerably varies the levels of soluble manganese near the dam wall.
A data-driven model based on correlations between manganese in the reservoir and the raw water of the WTP enabled the integration of the manganese transport simulations between the reservoir and the WTP. A decision support system (DSS) to improve the treatment of manganese was developed based on the data analysis of the manganese levels at the different water treatment stages. Situations where relatively high concentrations of Mn raw water entering the WTP were detrimentally compounded by processes in the sludge-thickener tank through the supernatant return were revealed. Through data analysis/model development between manganese in the raw water and in the outlet water, a prediction model of the highest expected manganese level (i.e. the worst-case scenario) in the treated water was established. The catchment and WTP model were integrated to enable the exploration of scenarios with different weather/environmental conditions, in order to predict expected manganese concentrations in the treated water and aid the formulation of potential mitigation strategies.
The development of the 3D manganese cycle model and its integration with the DSS for the WTP addresses a significant research gap in the literature. The outcomes of the research provide a comprehensive modelling study of manganese in water supply systems (catchment to treated water) and have beneficial implications for water utilities. Furthermore, this modelling study advances current knowledge and understanding of the manganese cycle for an entire waterbody through accurate 3D simulations, which in turn can lead to better manganese treatment management.||