An Integrated Modeling Approach to Optimize the Management of a Water Distribution System

Water distribution systems are essential infrastructures for any modern society and key elements for growth and human development. A reliable water service should be guaranteed to customers at all the aspects of the urban water cycle. However, an integral approach is often limited by the fact that responsibility of the different components of the water cycle is attributed to different entities. To fulfil the objective of providing acceptable level of service to customers, with the use of an asset in satisfactory performance conditions, the water managers have to plan how to operate, maintain and rehabilitate the system under budget constraints. The definition of rehabilitation plans requires the support of methodologies to select how to distribute the available resources. The goal of this research has been to develop a methodology to compute the energy embedded in the operation of a WDS and therefore the corresponding CO2 emissions. The innovative part of the work is the integration of two existing models: a hydraulic model and a metabolism model. The metabolism model, applying the urban metabolism concept to the water system, allows the evaluation of the energy consumptions and environmental impacts of a city analysing how the input fluxes of resources are transformed in useful energy, physical structure and waste like in a real ecosystem.  The hydraulic model, pressure driven, allows the estimation of the water flows, including background leakages, and functioning conditions of hydraulic machines (pumps or PAT or turbines if present) and therefore it provides inputs to the metabolism model to estimate the energy required to produce the water supplied or lost and to operate the network. The methodology is then applied as decision support system for rehabilitation and or repair plans, by ranking pipes in terms of risk. Risk is defined in the traditional way as combination of probability of a failure and its impact. Here the probability of pipes to break has been assessed, as simply exponential probability distribution function, for the aim of the demonstration the consequence as unsupplied water demand and pressure deficit at node level in case of failure. The methodology was applied to Laives network case study, a town in the province of Bolzano in Italy.