In recent years, sewer overflow related to stormwater has become a problem in many cities around the world. Therefore, an alternative to traditional sewer networks has been developed; the so-called ‘Sustainable Urban Drainage System’ (SUDS), which includes infiltration trenches, detention beds, etc. However, traditional urban drainage models do not so far include SUDS, making design of SUDS a challenge. In this study modeling and design possibilities of SUDS has been investigated by a combination of modeling and field experiments.
Investigation of modeling and design possibilities of Sustainable Urban Drainage Systems (SUDS)
In recent years, sewer overflow related to stormwater has become a problem in many cities around the world. This is a consequence of more frequent high-intensity rainfall events, resulting from climate change, combined with increasing impermeable surface area in urban areas. Therefore, new methods, denoted ‘Sustainable Urban Drainage Systems’ (SUDS), have been developed, including filter and infiltration trenches, detention beds and basins, ponds, etc. which delay and decrease the amount of stormwater entering the sewer system. Several models for traditional hydraulic infrastructure exist. However, since these models do not, so far, include SUDS, modeling and design of such systems remains a challenge. Eventually, better model tools for SUDS-modeling may be needed.
In relation to a SUDS project in Baunebakken, a residential area in Hvidovre, Denmark, an attempt was made to model a detention bed with limited infiltration through the bottom and walls, using two different modeling tools: 1) MIKE SHE, which deals with the processes in the hydrological cycle, and 2) MIKE Urban, which deals with traditional urban drainage systems. Simulations were run in MIKE Urban with various rainfall input. Model results indicated a fast response to rainfall. The bed was emptied almost immediately after the events and no overflow occurred for rainfall events with 5 and 10-years return periods lasting 10 and 20 minutes, respectively.
Also, four test detention beds of various designs connected to an infiltration trench were constructed in Baunebakken to test their hydraulic response to rainfall. The water levels in the bottom of the detention beds as well as ponding at the surface of the beds were measured, while water was led to the beds with a flow corresponding to the rainfall events simulated in MIKE Urban. One detention bed showed excellent hydraulic performance with no overflow, while the three other beds flooded. For two beds, all water had disappeared from the beds in 0,5-2 hours, while 2-8 cm of water stayed in the two remaining detention beds throughout the entire measuring period of several hours.