Sensitivity analysis of grate inlet representation and a comparison of two coupled hydraulic models for urban flood simulation

• Main Author: Lundqvist, Daniel
• Format: Others
• Language: English
• Published: Luleå tekniska universitet, Geovetenskap och miljöteknik 2020
• Subjects: Urban flood models; 1D-2D coupled models; MIKE FLOOD; FLO-2D; inlet grates; manholes; Water Engineering; Vattenteknik
• Online Access: http://urn.kb.se/resolve?urn=urn:nbn:se:ltu:diva-78224

Urban flood models are an important tool in designing and analyzing municipalities sewer drainage systems and predicting the effect of potential extent and depth of future floods. In urban areas, coupled 1D-2D flood models are particularly useful as they can represent the surface- and sewer system and their interactions. But it iss common practice to simplify the sewer system by only simulating water exchange between both systems at manholes while neglecting the effect of grate inlets. To investigate the effect grate inlet representations have in flood models, the simulation results of different models created in the software MIKE FLOOD with the number of nodes and inlet sizes adjusted according to the location of actual grate inlets were compared. In addition, a comparison between the flood modeling softwares MIKE FLOOD and FLO-2D was performed, based on a case study in Motala.. It was found that both MIKE FLOOD and FLO-2D can predict similar flood propagation and maximum water depths. The MIKE FLOOD models predicted larger amounts of drained water via the sewer system. This was likely caused by the extra water added through water level correction in the MIKE FLOOD models combined with numerical instabilities in the FLO-2D sewer models. Adjusting the number and dimensions of nodes according to actual grate inlets proved to have little effect on the predicted maximum surface water depths. But it did result in decreased drainage capacity together with less sewer inflow compared to the models neglecting grate inlets. The inlet representation did have a significant effect on predicted flood durations, with the models neglecting grate inlets having shorter flood durations in downstream areas and longer flood durations in upstream areas compared to the other models. It was also found that that the effect inlet node representations has on flood durations heavily depends on their locations with nodes located in water gathering areas such as depressions with ponding water having the most effect.