Addressing high dimensionality in water quality modelling in water distribution networks

Journal Title
Journal ISSN
Volume Title
University of the Witwatersrand, Johannesburg
Water quality models are the most effective tools for characterizing water quality conditions, assessing the effects of water pollution, and supporting decision-makers with water quality management. They can be utilised for detecting the variations in the water quality parameters. Despite the usefulness of water quality models, an appropriate and simple water quality descriptor for a particular application, considering the high dimensionality of various water quality parameters, remains a challenge (Chapman, 1992). To address this high dimensionality, a single dimensionless index is commonly used to describe water quality for a particular application. While pollution loads at various points in a river reach have been widely assessed by studies using water quality indices, little research has been done on water distribution networks with service reservoirs and a variation of loading conditions. In a water distribution network, service reservoirs function similarly to rivers in that they have complicated mixing mechanisms, are subject to a variety of water quality factors, and are sized and located differently. The most common water quality indices require the formation of subindices and weights to avoid ambiguity, eclipsing and rigidity. The Canadian Council of Ministers of the Environment Water Quality Index (CCME WQI) does not require the formation of sub-indices and weights, thus providing a simplified way of describing water quality. This study investigates the use of the CCME WQI to address high dimensionality in water quality modelling of water distribution networks, taking into consideration the locations of multiple service reservoirs. This study was carried out primarily for decision-making and design optimization purposes only. Using EPANET 2.2, four hydraulically optimised solutions (which satisfied minimum pressure requirements) were further analysed for water quality performance. This was achieved by incorporating simulated data on three water quality variables (chlorine residual, water age and THM concentration) into the CCME WQI for a hypothetical water distribution network, Anytown. The results indicate that two of the four hydraulically optimised solutions achieved excellent water quality levels. This study has demonstrated the usefulness of a dimensionless index as a proxy for multiple water quality variables of a water distribution system in facilitating decision-making.
A Research Report submitted to the School of Civil and Environmental Engineering in partial fulfilment of the requirements for the Degree of Master of Science in Engineering, in the Faculty of Engineering and Built Environment, School of Civil and Environmental Engineering, at the University of the Witwatersrand, Johannesburg, in 2024.
Water quality, Water quality modelling, Water quality index, Design optimisation, Wwater distribution network, Canadian Council of Ministers of the Environment Water Quality Index (CCME WQI), UCTD
Machweu, Morongwa Ednah. (2024). Addressing high dimensionality in water quality modelling in water distribution networks. [Master's dissertation, University of the Witwatersrand, Johannesburg]. WIReDSpace.