Combined sewer overflows (CSO) can have a strong impact on the quality of surface waters. A common measure to reduce CSO is the construction of storage tanks. The objective of this study was to determine the required volume of a storage tank by means of a numerical long-term simulation and to assess uncertain input data. Particularly, the influence of the considered rain series’ length on the calculated storage volume was investigated. Engineering standards usually recommend the use of at least 10 to 15 years of rain series. Here, the hydraulic behaviour of the studied sewer system was simulated in a 30 year hydrodynamic simulation. Special effort was made to calibrate an available model by use of currently measured data. The quality of calibration was evaluated by means of the Nash-Sutcliffe model efficiency coefficient. The analysis of input data uncertainty revealed that applying a 10 year series results in tank volumes that differ between -12 % and +19 %, respectively from the dimensioning result achieved by applying the 30 year rain series.

The overall objective of MIA-CSO is to develop a model-based planning instrument for impact based CSO control. The objective of this study was to examine the potential and the drawbacks of different model coupling techniques that may be taken into account within the MIA-CSO project.

Within the project SAM-CSO it shall be tested if the Open Modelling Interface and Environment (OpenMI) can be applied to link models of the Berlin sewerage (modelled in the urban drainage software InfoWorks CS,Wallingford Software) to a river water quality model. This report gives an overview on the OpenMI and its application. Chapter 1 outlines the general background of integrated water management and integrated modelling as it is aimed at by the European Water Framework Directive. The development process, which resulted in the release of the OpenMI is summarized in chapter 2. An introduction to the objectives, the concept and the technology of the OpenMI is given in chapter 3. Chapter 4 lists case studies in which the OpenMI has been applied. In Appendix B, each of the reported studies has been described in generalized form. A matrix showing all model links, which have been established within the case studies, has been developed. Finally, in chapter 5, an overview on other model linking approaches is given. This report shows that in many use cases the Open Modelling Interface could be used successfully for model linking. Even out of Europe, at a workshop of the U.S. EPA it is stated that, in terms of the ability to go between different temporal and spatial scales, a framework such as OpenMI might have the necessary flexibility. Actually, it was found that in many cases models of the InfoWorks software family have been part of the OpenMI linked systems. In cases of many interaction points between models, the OpenMI mechanism may not be applicable. In the Berlin case the impact of combined sewer overflows on the water quality of the receiving river shall be examined. With far less than a hundred interaction points between sewer model and river model it is assumed that the OpenMI could be used for a successful model linking. The difficulty within the SAM-CSO project may be to find an apropriate river quality model, which is ready to be linked to InfoWorks CS using the OpenMI. Unfortunately, there are few use cases reported in which a freely available river water quality model was involved. The water quality model QSIM of the German Institute of Hydrology (BfG) that is used within the project is currently not equipped with OpenMI. Nevertheless, using the OpenMI mechanism for model linking is assumed to be a promising approach. It is expected to become an internationally accepted standard. As the OpenMI specification is fully free, anyone may contribute to its further development. The OpenMI Association will give advice to modellers and will be open to discussions on improvement of the OpenMI. With the OpenMI linking mechanism not only models can be linked. Modules for calibration, optimization, statistical evaluation etc. can be part of an OpenMI system as well as components for generic data access or visualization. It will be tested, if the integration of such a module for statistical evaluation into the CSO impact assessment method (to be developed within the project SAM-CSO) is applicable and useful.