Optimal design of robust and future-proof network blueprints - Executive Summary - New approach produces network blueprints that are robust to alternative water demand scenarios

Water demand is an essential factor in the design of network blueprints. The current design process usually follows a deterministic approach. This produces a single design that satisfies the chosen water demand, but it can under-perform if the water demand turns out to be different. This research elaborated a new approach in which different water demand scenarios are taken into account in the design process. This provides insight in the trade-off between costs and performance of of different designs. The proposed approach allows designers to choose their desirable level of risk aversion, and the performance of the design can be assessed on the water demand effectively supplied under different scenarios. In this research the water demand scenarios were elaborated on the basis of historical water demand measurements and on future scenarios. These were used for the design of a network blueprint for a Dutch city. The results clearly show the relationship between the costs and performance under the different scenarios: a more robust design involves higher costs. The application of optimisation techniques helps finding better design solutions when compared to manual approaches. Thanks to the approach developed, drinking water utilities can make informed choices about how much to invest in a network blueprint and how to design it in order to achieve a certain level of robustness.

Interest: an ideal opportunity to install future- proof networks. Drinking water utilities are busy with the (proactive) replacement of their water distribution networks. This offers an ideal opportunity to improve these networks with an optimally designed, robust and future-proof network blueprint. The current design process does not take sufficient account of uncertainty in water demand, one of the greatest and most influential uncertainties in the design of distribution networks. In view of the long lifespan of these systems, it is extremely important that the newly-installed network be robust and future-proof: a network that can handle changes in water demand and continues performing as well as possible. This means that not one, but several water demand scenarios are taken into consideration in the design phase. Such an approach requires the calculation of several scenarios and designs: a task which quickly becomes unmanageable when done manually. It is here that numerical optimisation techniques offer a helping hand.

Report – This research is described in the report Optimaal ontwerp van robuuste en toekomstbestendige streefstructuren (BTO 2020.002).