Ivan Stoianov

In the last decade, the increasing implementation of high frequency monitoring has shown that water distribution networks are frequently affected by pressure transients. In some instances, even thousands of cyclic loadings can be exerted on pipes posing a risk of fatigue damage. High frequency monitoring, which can be as high as 128 samples per second, allows the capturing of extremely rapid transient events, otherwise undetectable using normal 15 min telemetry monitoring. Thus, providing a precise picture and awareness of the full dynamic activity in a pipe. However, despite this new awareness of a possible fatigue risk on pipes, current research has not be able to show to the extent of the relevance of this risk. Main limiting factors of previous published research have been short high frequency monitoring periods, limiting the generalisation of previous findings, and few mechanistic analysis utilising measured data. To address these issues, this work presents the analysis and evaluation of several months to years of high frequency monitoring water pressure data at various locations in operating water distribution networks. We have obtained extensive time series datasets with highly detailed information of transient events, and utilised this information together with knowledge of fracture mechanics to assess the extent of fatigue damage on pipes. In order to perform the analyses, our time series data were converted into tuples of number of cycles and magnitude by utilising a rainflow cycle counting algorithm. This algorithm allows the counting of full cyclic loadings and their magnitude in an spectral loading history. Such information can then be utilised for fatigue analysis. In particular, we have used theory of fatigue crack propagation to assess number of cycles to failure. Results from this research suggest evidence that at some logging locations, and depending on the pipe material, and current deterioration state, fatigue damage due constant cyclic loadings can be a main trigger causing pipe breaks. In particular we have estimated examples that show that up to tens of years of pipe life could be extended if pressure transient activity were reduced. Therefore, the need for further monitoring, analysis and reduction of water pressure transients is critical to increase the life of our water supply infrastructure.