Water quality indices for water distribution network optimisation

Abstract

Safe drinking water is the basis for human health, survival, growth and development. The provision of safe drinking water is additionally recognised as a basic human right and even though numerous global initiatives exist to this end; it is reported that at least 1.9 billion people rely on an unimproved source or an improved source that is faecal contaminated for drinking water. The provision of safe drinking water is protected by standards for drinking water quality where water supply agencies must ensure when planning, designing and running water treatment and supply infrastructure. These standards are not always upheld as reports of higher levels of salinity, microbial contamination, eutrophication and endocrine disrupting compounds are being reported at treatment plants in South Africa and worldwide. The design and optimisation of water distribution/reticulation networks using computational methods is a widely researched area in water design and management and is gaining more popularity as technological advances increase. Water distribution network design and optimisation typically includes selecting pipe elements, determining nodes, selecting tank locations and sizes, selecting pumps and pumping schedules while considering cost efficiency; however, inclusion of water quality has generally been excluded due to the increase in complexity resulting in increased computational time and space. However, to ensure efficient and safe provision of clean water; water quality has to be included in the design of water distribution networks to minimise potentially harmful effects. This research dissertation developed and tested the use of water quality indices that were historically used for studies on surface water bodies. The water quality indices are capable of transforming and combining multiple water quality constituents into a single aggregated value that can be incorporated into optimisation algorithms. Five water quality indices were developed and used here to assess the incorporation of water quality into the optimization of service reservoir designs. A set of sub-indices were developed for the four chosen water quality constituents: water age, chlorine, tri-halomethanes and halo-acetic acids and were then aggregated into indices that quantified the water quality for each reservoir. This formulation provided an estimate of the overall water quality in a reservoir where multiple water quality constituents with various units of measurement are involved. The formulation was intended to reduce computational effort for the optimisation of water distribution systems which include water quality as an objective.