Graduation Semester and Year

Spring 2025

Language

English

Document Type

Dissertation

Degree Name

Doctor of Philosophy in Civil Engineering

Department

Civil Engineering

First Advisor

Dr. Mohammad Majafi

Abstract

Supervising Professor: Dr. Mohammad Najafi

The United States boasts more than 2 million miles of water transmission pipes that have water main breaks. According to the American Society of Civil Engineers (ASCE) 2025 Report Card, an estimated six billion gallons of drinking water are lost per day. The detection of pipeline damage must always be a top priority in the water pipeline sector. This dissertation proposes a novel data driven approach for predicting number of broken wires in Prestressed Concrete Cylinder Pipes (PCCP) over time, a critical parameter in managing the integrity of aging water mains. With a large dataset of 53 years of historical pipeline inspection records with over 1,000,000 pipe segments and Artificial Neural Networks (ANNs), this research found aside from corrosion, pipe age and slope to be the most significant predictors of wire breaks among the other studied structural factors. Various ANN models were developed and compared using performance metrics such as Mean Absolute Error (MAE), Coefficient of Determination (R²), Root Mean Squared Error (RMSE), and Mean Bias Error (MBE). While the highest explanatory power (R²) was given by a model that used all the variables, a simplified model based on age and slope had comparable predictability but reduced complexity. This model is more suitable for applications with limited data availability or computational resources. Sensitivity analysis also established slope and age dominance, with significant error increases after their elimination. This research underpins more precise prediction of wire rupture, requiring data driven models that focus on the most critical structural factors. The derived models are valuable tools for enhancing predictive maintenance planning, optimal resource allocation for pipeline surveillance, and the long-term resilience of critical water infrastructure.

Disciplines

Civil and Environmental Engineering

License

Creative Commons Attribution 4.0 International License
This work is licensed under a Creative Commons Attribution 4.0 International License.

Comments

This dissertation was completed under the supervision of Dr. Mohammad Najafi. I would like to thank my advisor, committee members, and colleagues for their invaluable guidance and support.

Available for download on Sunday, May 30, 2027

Share

COinS