Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of Graph Search.
This paper presents a study of similarities between electrical and hydraulic pressurized networks. The primary objective is to examine whether or not it is possible to use electrical laboratory networks measuring voltage to study leak-region detection strategies measuring flow in water-distribution networks. In this paper, the strategy used to compare the networks is error-domain model falsification, a previously developed methodology for data interpretation that combines engineering knowledge with models and data to enhance decision making. Simulation results obtained for a part of the water-supply network from the city of Lausanne are compared with an analogous electric network. The electrical network is simulated using resistors to mimic the pipes. The consequence is that the electrical model is linear. The resistance values are obtained by computing the hydraulic resistance for each pipe, given by the Hazen-Williams equation. The compatibility of the two networks is evaluated through simulations in three ways: (1) comparing flow predictions obtained by simulating several leak scenarios; (2) comparing the expected identifiability (performance) of the two networks; and (3) comparing sensor placement configurations. The analyses show that even though the models have varying characteristics of underlying physical principles (the electrical model is linear while the hydraulic model is non-linear), the results are within generally accepted engineering limits of similarity (10%). This indicates that measurements on electrical laboratory networks have the potential to illustrate the efficiency and adaptability of leak-detection methodologies for full-scale water-supply and other pressurized hydraulic networks. Finally, two electrical laboratory physical networks, including an electrical model of part of the water network in Lausanne, were constructed and used in case studies to illustrate this adaptability.
Karim Achouri, Shourya Dutta Gupta, Ville Tuovi Tiukuvaara
Jürgen Brugger, Demetri Psaltis, Giulia Panusa, Pierrick Guy Robert Clément, Eirini Kakkava, Jongmoon Jang
François Avellan, Andres Müller, Keita Yamamoto