In the last years, the relieving activity of overflow structures in combined sewer systems was mainly determined by statistical analysis of rainfall data. Due to the rain lost on the surface of a drainage area and to the flow rétardation through the sewer network, the resulting runoff variation differs a lot from the rainfall variation. Therefore, the rainfall measurements of 4 rain gaging stations in Hamburg with a total of 67 recorded years and a rain gaging station in Berlin with 22 recorded years were used to carry out runoff simulations.
First of all, a rainfall statistical analysis was performed with the measured rainfall data in order to determine the annual rainfall durations, depths and frequencies in relation to the average rain intensity.
The annual overflow data from overflow structures (overflow weirs and overflow basins) of not prerelieved drainage areas were computed by means of a simplified runoff model. The so determined overflow data is considerably lower than the data of rainfall analysis. Hence, overflow data from overflow structures in combined sewer systems wouldn't be satisfying if they are derived from only statistical rainfall analysis. Recorded rainfall events should rather be transformed into runoff events and accordingly analysed.
As there were only 5 rain gaging stations available, a generalization of the computed overflow data is still impossible. However, these overflow data seem to be on the whole plausible if compared with rainfall data which were performed in the past only during the summer months.