Comparative analysis of reliability-surrogate measures using an ensemble of realistic water distribution networks

To evaluate the reliability of a water distribution system, surrogate measures are commonly used because of its easy computation and intuitive behavior. Some recent researchers have acknowledge the need to identify correlations between these indexes in order to select the ones that may be measuring similar properties of the networks and therefore are not adding further knowledge on how to make the systems more reliable. In order to evaluate these correlations between reliability-surrogate indexes assuring both appropriate conditions for the evaluation of flow-related indexes, and statistical significance of the found correlations, it is needed a set of realistic case studies. The research proposal is to use an ensemble of systems developed by the student in early stages of his PhD program to evaluate the correlations of the reliability-surrogate indexes using a fast and tested design algorithm as the one developed by the host supervisor. The result would be a series of correlations between indexes that may lead to recommendations on which ones are representative of the reliable behavior of a system. This results and recommendations are planned to be included in a publication for an international conference of water distribution systems (e.g. CCWI, WDSA, HIC).

Diego Paez
Faculty Supervisor: 
Yves Filion
Partner University: