Optimal design of virtual DMA's with gondwana
“Dutch drinking water companies are interested in improving their knowledge about their distribution systems, including background leakage, valve statuses, demand patterns, among others. District metering is a technique that can be useful in achieving this goal. Dutch water companies do, however, not wish to alter the hydraulic performance of their systems, which leads to a preference for virtual district metered areas (DMAs).
In this contribution, the performance of virtual DMAs in detecting system changes in relation to DMA size is explored. The design of virtual DMAs is formulated as a constrained single-objective optimization problem. The objective consists of minimizing the DMA boundaries (as a surrogate for number of flow meters that are installed) that are necessary to identify changes of a certain magnitude in a single DMA.
The optimization problem is solved in Gondwana, a generic software platform for the optimization of drinking water distribution networks. Gondwana has been extended with new functionality in order to be able to design virtual DMAs. This new functionality is tested on a real-life network, with good results. A trade-off between the number of required flow meters and the DMA sensitivity to detect anomalies is characterized.”
(Citaat: van Laarhoven, K., Vertommen, I., van Thienen, P., de Kater, H., Schaap, P.G., Gardien, D., van den Boom, M., de Groot, R. – Optimal design of virtual DMA’s with Gondwana. In Leading Edge Sustainable Asset Management of Water and Wastewater Infrastructure Conference (LESAM) 2017, Trondheim, Norway, 20-22 June 2017)