Pipelines are essential elements in the daily operation of modern cities, but significant economic loss and environmental damage commonly occur because of pipeline leakage. The acoustic method has proven to be effective for leak detection and location in gas pipelines. However, very little is currently known about the characteristics of leak noise generated by different leak shapes, sizes, and pressures. Moreover, the feasibility of the acoustic method in the presence of complexities has not been closely examined. This study attempted to address these research gaps through an experimental investigation. A pipeline test bed with a length of approximately 100 m and an internal diameter of 150 mm was established, and was structured to permit different leak shapes (circle and rectangle), sizes ($\phi 1$ , $\phi 3$ , and $\phi 5$ for circular leaks; $2\times 1$ , $2\times 3$ , and $2\times 10\mathrm{mm}$ for rectangular leaks) and pressures (0.1 and 0.3 MPa). The cross-correlation method was used to locate the leakage. The results suggest that leak noise is mainly concentrated at low frequencies and was capable of being transmitted into the branch and across bends. It is likely to be the leak area that affects leak noise and the effect of leak shapes can be neglected. The leak was accurately detected and located with a sensor distance of up to 70.15 m when the bandpass filter was appropriately selected. The feasibility of the acoustic method to detect and locate leakages in complex pipelines was demonstrated.

The reliability of urban water distribution networks (WDNs) is critical for ensuring sustainable infrastructure management. However, traditional failure prediction models often overlook the complex interdependencies between water pipelines and road networks, leading to suboptimal predictive accuracy. This study introduces a novel pipeline failure prediction framework that leverages Graph Neural Networks (GNNs) to incorporate coupled road–pipeline network features. By integrating traffic-related indicators, such as intersection proximity, pipeline–road angles, and network topology, this approach systematically assesses their impact on failure risk. A comparative evaluation of various GNN architectures, including Graph Convolutional Networks (GCNs), Graph Attention Networks (GATs), and GraphSAGE, demonstrates that GraphSAGE achieves the highest predictive performance, significantly surpassing traditional machine learning methods. The findings underscore the necessity of incorporating network topology into predictive models, validating the role of spatial dependencies in accurately assessing pipeline failure risks. This study contributes to advancing infrastructure resilience modeling by providing a robust predictive framework that supports proactive maintenance strategies and enhances risk mitigation in urban water distribution systems.

Engineering risk management in China is being paid increasing attention. More than ten years ago, risk assessment was only used for very large-scale infrastructure projects and focused on the project's economic risk. After 2002, engineering risk management was recommended for China's large-scale infrastructure construction. Project accidents were happening continuously, which lead to a lot of economic loss and severe injuries. All of the accidents on big projects provoked an urgency to take measures for dealing with risks. The government recognized that project risk management (RM) could manage the engineering construction risks and published some new guidelines/standards for large scale projects. The project RM guidelines/standards are focused chiefly on large infrastructure projects (tunnels, roads, railways, bridges, and harbor works) to manage the technical risks during project feasibility study, design and construction phases. The present article addresses the state-of-the-art risk management on tunnels and underground works in China. The features and functionality of the Project PM application (as illustrated from national data on accidents and casualties from 2003 -2011) can reduce the hazards and control the economic loss significantly.