Towards a Robust and Flexible Numerical Framework for Integrated Urban Water System Modeling

Abstract

Urban water system modeling offers an integrated approach to propose adapted solutions to the multiple issues faced by evolving urban water systems including hydrological variability, climate change and socio-economic factors. We design a generic and flexible numerical framework based on the integration of elementary infrastructure tools in a graph-based network. Additionally to centralized units, we consider decentralized elementary infrastructure tools such as Low Impact Development (LID) systems, local treatment units and temporary storage of harvested precipitations and grey water reusable for undrinkable uses. Infrastructure tools are represented by a hierarchy of classes in an object oriented language to integrate both common and more specifically designed systems. The integration of several instances of such objects in a connected network should eventually optimize urban and environmental planning, thanks to adaptation capacities through the infrastructure tools diversity. We propose an automatic derivation of equations from the elementary systems and their interconnection. We show how well-chosen system behaviors lead to robust and reliable differential algebraic equations, which can be integrated with advanced numerical schemes. As an illustration of our approach, we show how temporary storage may efficiently complement grey water and harvested precipitations through a connected and smart urban water network.