Evaluation of the Structure of Urban Stormwater Pipe Network Using Drainage Density

In mega cities such as Seoul in South Korea, it is very important to protect the cities from surface flooding even for a short time period due to the enormous economic damage. That is why stormwater pipe networks are commonly applied to mega cities with large impervious areas to drain runoff from the city. Therefore, the stormwater pipe networks in urban catchments should be carefully designed for quick and efficient runoff removal. In this study, the structures of different stormwater pipe networks were evaluated based on the relationship between peak rainfall and runoff in the urban catchments in South Korea. More than 400 historical rainfall events from five urban catchments were used to develop respective linear regression models for estimating peak runoff for different pipe network structures. The developed regression models exhibited greater than 0.9 in determination coefficients and demonstrated overall the broader ranges in peak runoff with the greater rainfall amount, especially when the pipe networks were branched. This implies that the effect of pipe network structures on runoff is more profound in the branched networks whose runoff water flow is one-directional and thus tends to concentrate to the catchment outlet. In the case of the looped networks in which runoff paths are multiple, rainfall runoff can be routed to several alternative water paths depending on rainfall events resulting in the reduced peak runoff. The structures of pipe networks can be measured in drainage density which is defined as the ratio of total pipe length to catchment area. As a result, the range of the estimated runoff at the 95% confidence level increased as the drainage density increased, which implies increased uncertainty with the looped networks which commonly involve more pipe installation for unit area as compared to the branched. However, the looped networks with multiple water paths can reduce the time to drain rainfall from the catchments and thus the 95% confidence interval becomes narrow, which means greater reliability in peak runoff estimation. It would therefore be favorable to adopt looped stormwater pipe networks within an affordable budget and the complexity of pipe networks needs to be counted to reduce urban flood risk