Methods for evaluating the impact of pressure transients on water distribution pipe breaks
Author(s)
Jara Arriagada, Carlos Ingnacio
Type
Thesis
Abstract
Water distribution networks worldwide are deteriorating, posing significant challenges for maintenance and repair due to their vast extent and the increasing frequency of pipe breaks. Effective management of these networks requires understanding the factors driving pipe breaks. Among these factors, water pressure emerges as a critical variable that can be regulated to reduce pipe deterioration. The concept of cumulative pressure-induced stress (CPIS) was recently introduced to capture the combined impacts of both mean pressure and pressure dynamics on pipe integrity. However, understanding of water pressure as a cause of pipe breaks and the development and testing of CPIS remain areas needing further investigation.
This thesis advances the understanding of CPIS through three key components: motivation and theoretical foundations, interpretation and analysis, and practical application to operational networks. First, the impact of mean pressure and pressure range metrics on pipe breaks is analysed using a large dataset, providing empirical evidence for the relationships of quasi-steady state pressures and pipe breaks. This analysis motivated the development of a fracture mechanics-based theoretical study to explain how pressure dynamics influence pipe deterioration, forming the foundation for CPIS. Second, the predictive capability of CPIS on pipe breaks is examined using machine learning models, which demonstrate its potential to enhance pipe break predictions and identify areas in need of pressure management. Finally, the practical challenges of implementing CPIS in operational networks are addressed. This includes investigating sampling resolution requirements, evaluating an interpolation method for CPIS, and formulating and testing a transient source localisation method to guide CPIS interpolation.
The insights and methods presented in this thesis enable the application of CPIS in both large-scale experimental and operational settings. These contributions will improve the resilience and efficiency of water distribution systems, supporting the development of sustainable and reliable water distribution networks.
This thesis advances the understanding of CPIS through three key components: motivation and theoretical foundations, interpretation and analysis, and practical application to operational networks. First, the impact of mean pressure and pressure range metrics on pipe breaks is analysed using a large dataset, providing empirical evidence for the relationships of quasi-steady state pressures and pipe breaks. This analysis motivated the development of a fracture mechanics-based theoretical study to explain how pressure dynamics influence pipe deterioration, forming the foundation for CPIS. Second, the predictive capability of CPIS on pipe breaks is examined using machine learning models, which demonstrate its potential to enhance pipe break predictions and identify areas in need of pressure management. Finally, the practical challenges of implementing CPIS in operational networks are addressed. This includes investigating sampling resolution requirements, evaluating an interpolation method for CPIS, and formulating and testing a transient source localisation method to guide CPIS interpolation.
The insights and methods presented in this thesis enable the application of CPIS in both large-scale experimental and operational settings. These contributions will improve the resilience and efficiency of water distribution systems, supporting the development of sustainable and reliable water distribution networks.
Version
Open Access
Date Issued
2025-01-20
Date Awarded
01/04/2025
License URL
Advisor
Stoianov, Ivan
Sponsor
ANID (Agencia Nacional de Investigación y Desarrollo, Chile)
Grant Number
72210314
Publisher Department
Department of Civil and Environmental Engineering
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)