Sewer model development under minimum data requirements

Abstract

Planning, design and operation of urban drainage systems is often based on hydraulic sewer modelling. Sewer models are also increasingly used to quantify pollution loads discharged to aquatic ecosystems (e.g. via combined sewer overflows), which ultimately allows an estimation of the ecological impact emanating from urban drainage systems. The establishment of such network models, however, requires detailed and accurate information about the sewer network structure and the connected surface area. This infrastructure data is often unavailable, confidential or available in ‘paper’ format only. The present paper outlines a novel approach to develop a hydraulic sewer model constrained by a minimum amount of data. The approach combines the application of a surface flow accumulation algorithm to a selectively manipulated Digital Elevation Model (DEM) with a routine for hydraulic network dimensioning to generate a close-to-reality sewer network ready to be implemented in a hydraulic modelling platform. The method is tested for three real-life catchments of which characteristics vary in scale, topography, state of development and network complexity. For all cases the generated network is implemented on the EPA-SWMM platform to allow hydrodynamic simulations. Model performance is assessed by (1) evaluating the spatial match of existing and generated network layout, (2) comparing the estimated hydraulic dimension with real-life infrastructure data and (3) benchmarking simulated runoff with measured data for a defined validation period. The analysis shows that the presented method is capable of reproducing the original network layout, network length and corresponding discharge rates based on little, freely available information. Further research potential is identified to improve the hydraulic dimensioning and the application to complex systems that include control structures. The presented approach is useful to estimate the scope of drainage networks including layout and design (e.g. for preliminary planning in emerging areas) to screen existing networks and to identify critical spots where more precise information is required.