Study of Water Losses’ Determinants in order to introduce Pressure and Number of Connections into Water Loss Indicators for French drinking water supply networks | Etude des déterminants de pertes des réseaux d'eau dans l'objectif d'introduire le nombre de branchements et la pression dans les indicateurs de pertes

[Departement_IRSTEA]Eaux [TR1_IRSTEA]GEUSI

International audience

Английский. Water loss performance indicators are used to target the most critical zones on networks on which actions have to be done to reduce the leakage. They are a first diagnosis tool to identify the origin of water losses. However the performance indicators commonly used in France (that are the network efficiency and the linear loss index) do not seem relevant since they do not take into account some parameters of the networks –such as pressure or the number of connections- that influence the water losses. In this context, Veolia Water and Irstea (French research institute) have studied the introduction of pressure and the number of connections in performance indicators. These parameters are already used in IWA indicators but previous studies have demonstrated that these indicators are not adapted to French networks (due to their sizes). Thus, a multi-variable model of the water losses has been studied in order to identify the properties of the networks that explain the water losses at best. The objectives of this work is to create relevant performance indicators to benchmark the technical performance of different water networks and above all, to settle adapted water loss targets regarding the properties of each network. Two new indicators have been investigated: -One related to the infrastructure as a ratio between the current annual water losses and the product of the number of connections and the average pressure -The other related to the pressure as a ratio between the current average pressure and the minimum average pressure. The first part of the study consists in working at the DMA level of different networks. The yearly data from one hundred DMAs (issued from five water networks) were collected and analysed. The basic data used are the distributed and consumed volumes, the number of connections and the average pressure (calculated with hydraulic model). Others have been collected such as the number of failures on pipes and connections, the pipe ages, the pipe diameter and the pipe material. This analysis at the DMA level highlights a strong relationship between the water losses and the number of connections on one hand, and, on the other hand, the lack of significant relationship between water losses and the average pressure. Besides, this analysis underlines the difficulties to collect some information at the DMA level such as the average pressure or sometimes the number of connections and then the volume of water consumed. A second part of the study at the level of the whole water networks is in progress. Indeed, this allows to increase significantly the panel of data. Besides, the difficulties of gathering data at the DMA level and the associated uncertainties are deleted so the results will be more robust. This part of work required the development of an EPAnet module to calculate the average pressure of the network based on hydraulic simulation. Nowadays, the results at this scale are similar to the ones obtained at the DMA level. Other variables such as the intensity of active leakage control could be introduced in the model of water loss volume. This study shows that the criteria “average pressure” is far from explaining the level of water losses on a network, despite the well-known impact of pressure management on water losses. However, the new water losses performance indicators defined in this study provide new tools to identify whether the water losses are due to an excess of pressure or to poor infrastructure. Besides, since few data are necessary to calculate these indicators, they can be evaluated on all the networks whatever their sizes (length of pipes or number of connections) and can be used as benchmark indicators. The next step of the study will be to define reference values for these indicators.