The Importance of Rainfall Distribution in Urban Drainage Operation

Abstract

A control simulation model, called LOCUS, is used to investigate the effects of spatially distributed rain and the possibilities to benefit from this phenomenon by means of real time control. The study is undertaken for a catchment in Copenhagen, where rainfall is measured with a network of 8 rain gauges. Simulation of a single rain event, which is assumed to be homogeneous, i.e. using one rain gauge for the whole catchment, leads to large over- and underestimates of the systems output variables. Therefore, when analyzing a single event the highest possible degree of rainfall information may be desired. Time-series simulations are performed for both an uncontrolled and a controlled system. It is shown that from a statistical point of view, rainfall distribution is NOT significant concerning the probability of occurrence of an overflow. The main contributing factor to the potential of real time control, concerning minimizing overflows, is to be found in the system itself, i.e. the distribution of available storage and discharge capacity. When other operational objectives are involved, e.g., to minimize peak flows to the treatment plant, rainfall distribution may be an important factor.

Migrating Storms and Optimal Control of Urban Sewer Networks

“…These non-uniform storms affect the runoff hydrograph and the magnitude of the peak discharge of the runoff hydrograph. The overall peak discharge of a downstream migrating storm exceeds that of an upstream migrating storm [1][2][3][4]. The safety level is low compared to a uniform rainfall event, with equal directions for both wind and flow [1], and can easily cause combined sewer overflows (CSOs).…”