Urban Water Supply Systems (UWSS) are critical for supporting human life in the urban environment. Therefore detecting faults in UWSS before and after they happen has become a key global goal. The objective of this thesis is therefore to develop tools and models in support of detecting leaks and blockages in water pipelines. The thesis makes two important contributions. The first contribution is the development of a high precision measurement system for UWSS related applications. This includes designing a robust Noise Measurement System, a Low Frequency Vector Network Analyzer for water pipeline channel measurements and a flexible signal generator and acquisition system for performing defect detection experiments. The second contribution of this thesis is the development of a com...[ Read more ]

Urban Water Supply Systems (UWSS) are critical for supporting human life in the urban environment. Therefore detecting faults in UWSS before and after they happen has become a key global goal. The objective of this thesis is therefore to develop tools and models in support of detecting leaks and blockages in water pipelines. The thesis makes two important contributions. The first contribution is the development of a high precision measurement system for UWSS related applications. This includes designing a robust Noise Measurement System, a Low Frequency Vector Network Analyzer for water pipeline channel measurements and a flexible signal generator and acquisition system for performing defect detection experiments. The second contribution of this thesis is the development of a comprehensive acoustic noise model for water pipeline channels. This involves analyzing experimental data of acoustic noise from a wide range of experimental conditions in laboratory test beds and operational UWSS systems. Using these results, characteristics such as power spectral density, amplitude probability distributions and stationarity properties of acoustic noise in pipelines have been modeled. The analysis of temporal variations in properties of noise also helped in identification and classification of different types of noise sources and their influence on the channel properties. Using these results, it is shown that acoustic pipeline noise follows an α-stable distribution and a Wenz like framework for colored power spectral density. Furthermore, an acoustic noise model for pipeline channels is also delivered that can be used for simulation and verification of leak and blockage detection techniques.