This project’s objective is to determine, with the assistance of Gondwana, the optimal number and placement of pressure sensors in the Seppe area, with a view to maximising detection probability and the localisation of possible leaks. The pressure measurements will be processed by Brabant Water using BurstAlert. This will meet an additional objective, namely: to acquire insight into the value of numerical optimisation methods in the design process of measurement networks in practice.
Brabant Water will invest in pressure sensors for the Seppe area so as to detect and locate leaks more quickly. To benefit as much as possible from this investment, it is important that the number of pressure sensors as well as their placement be carefully considered. KWR’s Gondwana optimisation platform offers a suitable tool to support this process.
Gondwana is a software tool that is capable of structurally incorporating modifications in the design or the operation of a drinking water distribution system model. It can then assess the performance of the modified design according to one or several objectives chosen by the user and, on this basis, introduce further modifications. Thus, in an iterative process, an optimally performing design or operation is arrived at. Because of the huge number of possibilities, it is simply impossible to calculate all the conceivable combinations – an example: 10 different possible diameters for 100 pipe segments demands calculations for 10,100 network designs. Given the complexity of networks and the large number of prerequisites, even an informed and experienced professional can no longer determine which design is the best. In this context, numerical optimisation techniques, as implemented in Gondwana, offer a helping hand.