| Author: | Boadu, Vincent |
| Title: | Leak detection system for water mains : a Hong Kong case study |
| Advisors: | Zayed, Tarek (BRE) |
| Degree: | M.Sc. |
| Year: | 2019 |
| Subject: | Hong Kong Polytechnic University -- Dissertations Water-pipes -- Inspection Water-pipes -- China -- Hong Kong -- Maintenance and repair Water-supply -- China -- Hong Kong Water leakage -- China -- Hong Kong |
| Department: | Faculty of Construction and Environment |
| Pages: | x, 78, xiv pages : color illustrations |
| Language: | English |
| Abstract: | One of the most critical and valuable assets in the development of municipal infrastructure systems is the water distribution network (WDN). However, water mains especially older pipelines are constantly threatened by the existence of leaks. For instance, 45 million cubic meters of water is loss daily through the water distribution network in developing countries and a total of more than 32 billion cubic meters yearly on the global level. A report released by the Water Supplies Department in Hong Kong in 2018 indicates that the water leakage rate in the water network is 15.2%. Leakage in water mains do not only lead to financial losses but also raises concern on the environmental, health and safety as well as sustainability issues. Early detection of leaks in the mains using wireless noise loggers has proven to be a good technology to detect leaks in the WDN. Hence, the objectives of the present research include; (1) to identify and study the causes of water loss in water mains, (2) to identify and study the existing methods used to detect leaks in water mains and (3) to develop a leak detection model. On this bases, critical literature review was conducted to ascertain the causes of water loss and the existing methods for detecting leaks in water mains. In addition, acoustic sound information was obtained from the leak detection consultants and analysed using Fast Fourier Transform and RapidMiner for the development of the leak detection model. As a result, using semi structured questions, interviews were conducted with professionals at the Water Supplies Department (WSD) and with leak detection consultants involved in detecting leaks in water mains in Hong Kong. The findings from the research shows that powerful storms, leakages, construction activities, industrial activities and aging of pipes and fittings are the major causes of water loss in Hong Kong. Ground microphones, leak noise correlators, noise loggers and mechanical and plastic listening sticks are some of the technologies used in detecting leaks on the WDN. Also, three leak detection models were developed using Decision Tree, Naive Bayes and Deep Learning algorithms. The accuracy of the developed models was tested on randomly selected datasets. The Decision Tree, Support Vector Machine and Naive Bayes had accuracies of 80.95%, 79.29% and 98.57%, respectively. As a result, the Naive Bayes model was selected as the best model for further development. Sensitivity analysis of the Decision Tree model and the Support Vector Machine model conducted indicated that the developed models were highly sensitive to changes in changes in the overall accuracy with slight change in their thresholds. The developed model proves that automated real-time leak detection is possible and can help utility departments to save water resource and funds. Moreover, this research proves the viability of noise loggers as a standalone leak detection technology and the opens the door for further research and experimentations. The research therefore recommended that the developed model be validated on the field. Future works can consider using acoustic filtering techniques, such as white noise filtering, impulse response filtering and linear filtering, to enhance the performance of the model. |
| Rights: | All rights reserved |
| Access: | restricted access |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| 991022289507503411.pdf | For All Users (off-campus access for PolyU Staff & Students only) | 2.81 MB | Adobe PDF | View/Open |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/10269