Vincent Wolfs

Development of a semi-automated model identification and calibration tool for conceptual modelling of sewer systems.

Applications such as real-time control, uncertainty analysis and optimization require an extensive number of model iterations. Full hydrodynamic sewer models are not sufficient for these applications due to the excessive computation time. Simplifications are consequently required. A lumped conceptual modelling approach results in a much faster calculation. The process of identifying and calibrating the conceptual model structure could, however, be time-consuming. Moreover, many conceptual models lack accuracy, or do not account for backwater effects. To overcome these problems, a modelling methodology was developed which is suited for semi-automatic calibration. The methodology is tested for the sewer system of the city of Geel in the Grote Nete river basin in Belgium, using both synthetic design storm events and long time series of rainfall input. A MATLAB/Simulink(®) tool was developed to guide the modeller through the step-wise model construction, reducing significantly the time required for the conceptual modelling process.

Modular conceptual modelling approach and software for river hydraulic simulations

Numerous applications in river management require computationally efficient models that can accurately simulate the state of a river. This paper presents a reduced complexity modelling approach that emulates the results of detailed full hydrodynamic models. Its modular design based on virtual reservoirs allows users to combine different model structures depending on the river dynamics and intended use. A semi-automatic software tool (Conceptual Model Developer, CMD) was developed to facilitate model set-up. To prevent instabilities during simulations, a highly efficient discrete calculation scheme is presented with a variable time step. To illustrate the effectiveness of the presented approach, the Marke River in Belgium was conceptualized based on simulation results of a detailed model. Results show that the derived conceptual model mimics the detailed model closely, while the calculation time is reduced by more than 2000 times. Finally, several applications are discussed that employ conceptual models built according to the presented approach. A reduced complexity modelling approach for river hydraulic simulations is proposed.Flexible modular framework that incorporates a variety of structures and techniques.A software tool (CMD) was created to speed-up and facilitate model set-up.A discrete calculation scheme was developed with time and space varying time step.A case study and discussion of applications illustrate the approachs effectiveness.

Green–blue water in the city: quantification of impact of source control versus end-of-pipe solutions on sewer and river floods

Urbanization and climate change trends put strong pressures on urban water systems. Temporal variations in rainfall, runoff and water availability increase, and need to be compensated for by innovative adaptation strategies. One of these is stormwater retention and infiltration in open and/or green spaces in the city (blue-green water integration). This study evaluated the efficiency of three adaptation strategies for the city of Turnhout in Belgium, namely source control as a result of blue-green water integration, retention basins located downstream of the stormwater sewers, and end-of-pipe solutions based on river flood control reservoirs. The efficiency of these options is quantified by the reduction in sewer and river flood frequencies and volumes, and sewer overflow volumes. This is done by means of long-term simulations (100-year rainfall simulations) using an integrated conceptual sewer-river model calibrated to full hydrodynamic sewer and river models. Results show that combining open, green zones in the city with stormwater retention and infiltration for only 1% of the total city runoff area would lead to a 30 to 50% reduction in sewer flood volumes for return periods in the range 10-100 years. This is due to the additional surface storage and infiltration and consequent reduction in urban runoff. However, the impact of this source control option on downstream river floods is limited. Stormwater retention downstream of the sewer system gives a strong reduction in peak discharges to the receiving river. However due to the difference in response time between the sewer and river systems, this does not lead to a strong reduction in river flood frequency. The paper shows the importance of improving the interface between urban design and water management, and between sewer and river flood management.

Modular Conceptual Modelling Approach and Software for Sewer Hydraulic Computations

A major challenge in urban water management is the identification of cost-effective and future-proof strategies that can cope with the rapid urbanization and changing environmental conditions. Water quantity modelling forms a key-element in the development of such strategies. Conventional detailed hydrodynamic models are not well suited for use in decision support systems due to several important drawbacks. Therefore, this paper presents a novel and computationally efficient conceptual modelling approach for sewer water quantity simulations. A modular framework is considered that combines well-established model structures with machine learning techniques. This flexible framework ensures that even complex flow dynamics can be emulated accurately. An accompanying software tool was developed to facilitate model configuration. As an example, a full hydrodynamic sewer model of a city in Belgium was transformed into a conceptual model. This model delivered precise results, while the calculation time was 106 times shorter than the detailed model.

Development and Comparison of Two Fast Surrogate Models for Urban Pluvial Flood Simulations

Detailed full hydrodynamic 1D-2D dual drainage models are a well-established approach to simulate urban pluvial floods. However, despite modelling advances and increasing computational power, this approach remains unsuitable for many real time applications. We propose and test two computationally efficient surrogate models. The first approach links a detailed 1D sewer model to a GIS-based overland flood network. For the second approach, we developed a conceptual sewer and flood model using data-driven and physically based structures, and coupled the model to pre-simulated flood maps. The city of Ghent (Belgium) is used as a test case. Both surrogate models can provide comparable results to the original model in terms of peak surface flood volumes and maximum flood extent and depth maps, with a significant reduction in computing time.