PROBABILISTIC RESILIENCE MODEL FOR MANAGEMENT OF WATER SUPPLY SYSTEMS
Author
Andersen, Tom
Term
4. term
Education
Publication year
2023
Submitted on
2023-06-01
Pages
67
Abstract
This thesis examines economic approaches to strengthen the resilience of TREFOR’s water supply system in Middelfart, focusing on critical components. It evaluates two main models: a probabilistic resilience model and a renewal model. The probabilistic resilience model uses probabilities to map potential failure scenarios in utility infrastructure and focuses on internal degradation processes in components such as pumps and valves. Its usefulness could be increased by adding external hazards in future work to capture more sources of risk. The renewal model is well suited to compare different renewal strategies and their timing and economic impacts. It provides forecasts of the future state of the water system and the funding needed to maintain resilience. Despite year-to-year cost differences between strategies, the results indicate practical economic thresholds that can guide TREFOR’s decisions. At the same time, the model may introduce inaccuracies when the lengths of pipe segments selected for renewal vary. The study recommends combining both models to achieve the best overall improvement in system resilience. Further development is needed, including integrating external hazards and examining model assumptions more closely to improve predictive accuracy. The work was partly constrained by incomplete data on economic consequences. However, assumptions informed by discussions with TREFOR, together with knowledge of pipe degradation rates and burst data, offered valuable insight into system risks. Overall, the study shows that both models can help build a more resilient water supply in Middelfart and provides insights into degradation, risk types, and response times to failures that support strategic choices. It outlines a path toward more resilient utility infrastructure grounded in economic reasoning and highlights the importance of ongoing research and model refinement.
Denne afhandling undersøger økonomiske tilgange til at gøre TREFORs vandforsyningssystem i Middelfart mere robust, med fokus på kritiske komponenter. To hovedmodeller vurderes: en probabilistisk robusthedsmodel og en fornyelsesmodel. Den probabilistiske robusthedsmodel bruger sandsynligheder til at kortlægge mulige fejlsituationer i infrastrukturen og beskriver især indre nedbrydningsprocesser i komponenter som pumper og ventiler. Modellen kan blive endnu mere nyttig, hvis den senere udvides med hensyn til eksterne trusler, så den dækker flere risikokilder. Fornyelsesmodellen er stærk til at sammenligne forskellige fornyelsesstrategier og deres tidsmæssige økonomiske konsekvenser. Den giver et bud på den fremtidige tilstand af vandforsyningen og hvilke midler der kræves for at bevare robustheden. På trods af årlige forskelle i omkostninger mellem strategierne peger resultaterne på praktiske økonomiske tærskler, der kan støtte TREFORs beslutninger. Samtidig kan modellen give upræcise resultater, når længderne af de rørstrækninger, der udvælges til fornyelse, varierer. Afhandlingen anbefaler, at TREFOR kombinerer de to modeller for bedst muligt at styrke systemets robusthed. Samtidig er der behov for videreudvikling, herunder integration af eksterne trusler og en nærmere granskning af modelantagelser for at forbedre forudsigelserne. Arbejdet var til dels udfordret af ufuldstændige data om økonomiske konsekvenser. Antagelser baseret på drøftelser med TREFOR samt viden om rørnedbrydning og bruddata gav dog nyttig indsigt i systemets risici. Samlet viser studiet, at begge modeller kan bidrage til en mere robust vandforsyning i Middelfart og giver viden om nedbrydning, risikotyper og reaktionstider ved fejl, som kan understøtte strategiske valg. Det peger på en vej mod mere modstandsdygtig infrastruktur med solid økonomisk begrundelse og understreger behovet for løbende forskning og modelforfinelse.
[This apstract has been rewritten with the help of AI based on the project's original abstract]
Keywords