A Complex System Approach to Reliability Analysis of Water Distribution Networks

Abstract

Water Distribution Networks (WDNs) are confronted with numerous operational threats that lead to disruption and dysfunction of their performance. As a response to the growing operational dysfunctions, researchers have recognised the importance of using reliability theory to examine the ability of WDNs to provide continuity in operation. However, the current approaches to reliability analysis of these networks mainly focus on one aspect of the reliability problem and fail to provide a complete representation of all factors involved in reliability analysis. These methods are embedded in capturing either the topological properties or the hydraulic attributes of WDNs. On one hand, the hydraulic-based approaches yield insufficient information as to the structural complexity and the level of interaction among components. On the other hand, the existing topological-based approaches just capture very generic topological properties and ignore various hydraulic attributes of WDNs such as demand and pressure head. Furthermore, the conventional reliability analysis methods are only effective for demonstrating a snapshot of these networks at a given point in time and ignore the variation in the parameters involved in the reliability analysis. This thesis attempts to fill these gaps by generating new knowledge in the area of reliability analysis of WDNs through using a combination of scientific approaches. This includes reliability engineering, system thinking, network theory, probabilistic analysis and hydraulic engineering. It is in this spirit that this research introduces a three-tiered approach. Tier 1 is explicitly tied to evaluate the topological reliability of WDNs. Tier 2 will be developed based on the results of Tier 1, aimed at establishing an integrated framework for reliability analysis. Tier 3 will use the outputs generated by tier 2 and will attempt to capture the dynamic nature of WDNs. In attempting to develop a comprehensive reliability assessment model, the present thesis proposes a number of novel reliability analysis methods for WDNs. Using three case studies from the literature as well as four real-world WDNs of Australian towns, this thesis demonstrates the effectiveness of the proposed methods. This research provides two types of implications. For theory development, it offers new insight and interpretation into the reliability analysis of WDNs by integrating a broad spectrum of various approaches. For water engineering management, the predictive maintenance strategy based on the reliability assessment model proposed here will provide an expert facilitator that helps water service providers to establish and implement a cost-effective maintenance strategy, which relies on identifying and prioritising the vulnerabilities, thereby reducing expenditures on the maintenance activities.