Jonatan Zischg

Smart Rain Barrels: Advanced LID Management Through Measurement and Control

Rain barrels are micro-scale applications which are used as temporary storage and for rainwater harvesting. They can be easily implemented into existing stormwater infrastructure. Recent advances in the field of Internet of Things (IoT) have opened up new possibilities for real-time monitoring and control of such structures, that enable the reduction of urban flooding or combined sewer overflows. The special feature of our smart rain barrel is its integration into a pilot project for smart cities, where every water inflow and outflow of the university campus in Innsbruck (Austria) is measured. Weather forecasts and time-controlled filling levels of different Low Impact Developments (LID) structures and the connected sewer system are used for real-time control (RTC). In a first step, the smart rain barrels are implemented into a SWMM-model with the objective of reducing the peak runoff rate by using the filling level in the main conduit as the control variable for real-time control. Results show that depending on the installation site and the storage volume of the rain barrel, a flood volume reduction of 18–40% can be achieved although only a simplified automatic control system has been implemented.

Parameter Sensitivity of a Microscale Hydrodynamic Model

Here we present the results of a global sensitivity analysis (GSA) applied for a microscale hydrodynamic model, which combines pipe infrastructure and small scale source treatments in terms of raingardens (RGs). The aim is to identify the most influential model parameters to support the decision for future measurement installation sites and smart water control. For the model creation and simulation, the Storm Water Management Model (SWMM) is used. For the GSA method the Elementary Effect Test (EET) is applied, were uncertainties to 18 model input parameters, comprising 10 subcatchment and 8 Low Impact Development (LID) parameters, are assigned and analysed by 1,900 simulations. The model’s responses are evaluated at four main RGs and for two model outputs: Inflow and Surface runoff at the RGs. First results show that the most sensitive factors are the Depression Storage Impervious and the Soil Hydraulic Conductivity for the Inflow and Surface Runoff at RGs, respectively.

Morphogenesis of Urban Water Distribution Networks: A Spatiotemporal Planning Approach for Cost-Efficient and Reliable Supply

Cities and their infrastructure networks are always in motion and permanently changing in structure and function. This paper presents a methodology for automatically creating future water distribution networks (WDNs) that are stressed step-by-step by disconnection and connection of WDN parts. The associated effects of demand shifting and flow rearrangements are simulated and assessed with hydraulic performances. With the methodology, it is possible to test various planning and adaptation options of the future WDN, where the unknown (future) network is approximated via the co-located and known (future) road network, and hence different topological characteristics (branched vs. strongly looped layout) can be investigated. The reliability of the planning options is evaluated with the flow entropy, a measure based on Shannon’s informational entropy. Uncertainties regarding future water consumption and water loss management are included in a scenario analysis. To avoid insufficient water supply to customers during the transition process from an initial to a final WDN state, an adaptation concept is proposed where critical WDN components are replaced over time. Finally, the method is applied to the drastic urban transition of Kiruna, Sweden. Results show that without adaptation measures severe performance drops will occur after the WDN state 2023, mainly caused by the disconnection of WDN parts. However, with low adaptation efforts that consider 2–3% pipe replacement, sufficient pressure performances are achieved. Furthermore, by using an entropy-cost comparison, the best planning options are determined.