An agent-based approach for generating virtual sewer systems.

The application of artificial case studies is a well established technique in urban drainage to test measures, approaches or models. However, the preparation of a virtual case study for a sewer system is a tedious task. Several algorithms have been presented in the literature for an automatic generation of virtual sewer systems. Applying the approach of generating virtual cities by means of the software VIBe (Virtual Infrastructure Benchmarking) the urban structure (including elevation map, land use and population distribution) is generated firstly and the infrastructure is designed meeting the requirements of the urban structure. The aim of this paper is the development of an agent based approach for generating virtual sewer systems. This new algorithm functions as module of the software VIBe but can of course also be applied to a real city in order to get information on possible/optimal sewer placement. Here hundred virtual VIBe cities and for each twelve virtual sewer networks are generated and calibrated based on data of an alpine region. It is revealed that with the approach presented virtual sewer networks which are comparable with real world sewer networks can be generated. The agent based method provides data sets for benchmarking and allows case independent testing of new measures.