Probabilistic performance prediction model for Austrian torrent control infrastructure

Due to the growing stock of torrent control structures and decreasing investments, the protection level can be reduced in the long term. Therefore, the focus of the future investments has been shifted on efficiency improvement and maintenance of existing structures. The maintenance of the existing structures is necessary to ensure safety from natural hazards, which are increasing as a consequence of the climate change. Due to the increasing number of structures, there is a strong need to develop a good maintenance management for these infrastructure works. This study is based on data from the condition rating programme of the Austrian Service for Torrent and Avalanche Control. For the structures relevant in this work, 75,343 records were available. The data were examined for different construction type and construction materials. The reliability of the structures was determined by means of the Markov chain. The study describes a homogeneous Markov chain model formed by the data of the condition rating. The result of this study is a proposal for a time to failure (TTF) and a time to repair (TTR) for several construction types with different construction materials as well as the time-dependent use of construction materials in the sector of the Austrian torrent control infrastructure.