THESIS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY Aspects of historical data and health criteria for drinking water network replacement strategies

The drinking water distribution network represents a major proportion of the investments and capital assets of a water utility. Consequently, qualified insight into future replacement needs provides water utilities with a foundation for financial planning. This insight would allow responsible engineers to choose the right projects (pipes and pipe systems) for replacement. Currently, support for the correct project choice is through sophisticated methods and models. However, utilities (especially smaller) need simpler procedures as they often lack both input data and competence for advanced infrastructure asset management models, as well as the experience of using such models. The aim of this thesis has been to provide new knowledge and useful, simple and transparent tools for the assessment and evaluation of long-term needs and prioritization of drinking water pipe replacement. An assessment of the future long-term replacement needs for drinking water distribution networks can be made through a combination of lifetime distribution functions and current network age data. Reliable lifetime predictions are limited by a lack of understanding of deterioration processes for different pipe materials under varying conditions. However, in this thesis a method was applied to calculate national investment needs and the results provided a basis for estimates for Swedish utilities where there is a scarcity of data. An alternative approach, employed successfully in this thesis, was the use of real historical data for replacement over an extended time series. The verified data provided a good fit to commonly used lifetime distribution curves. Further, reasonable projections of replacement needs into an uncertain future could be made. CBA (Cost-benefit analysis) can be used to evaluate the replacement strategy for utilities’ water distribution networks. CBA was applied to evaluate how first, pipe failure data and second, leakage strategies, might be used in pipe prioritization strategies. CBA was applied to pipe failure data replacement priorities, and here the cost of replacement was compared to the benefits of fewer pipe failures. The method enabled the selection of prioritised pipe sections for replacement without the need for a range of parameters and advanced methods that are difficult to interpret. Scenario analysis showed that health aspects have a significant impact on the result, and a method for evaluating the health risk was developed. For the CBA application to leakage management, the benefits of leakage reduction were compared to the cost of alternative management options to determine which was the most cost-effective. In the case study distribution system it was demonstrated that it is significantly more cost-effective to reduce leakage volumes by reactively repairing broken pipes than to proactively replace them, despite large leakage losses.