MSc Thesis Project: Improvement in Dam Leakage Monitoring by Machine Learning Technique

Company Description

For over 100 years, we have been electrifying industries, powering homes and transforming life through innovation and cooperation. Guided by our goal, we are committed to enabling a fossil-free life within one generation. As one of Europe's leading energy corporations, we must ourselves be free of fossil fuels to achieve this.

But that's not enough. That's why we are looking beyond our own business, to see where we can make the biggest contribution.

Join us on our journey towards a fossil-free life within one generation.

Want to know more about what we do? Visit .

Being a part of Sweden's critical infrastructure, many of our services are security rated. If this position is security rated, the final candidates may be subject to a security vetting process as per Swedish law.

Job Description

Power Climate Smarter living - this is our purpose. We're looking for bright students to help make this a reality. Together, you'll contribute to a fossil-free future. You'll have a unique opportunity to contribute to our purpose. All we ask is that you use your unique skills and share your energy to help us achieve our goal.

About the thesis

• The project will be conducted at Vattenfall R&D.
• Thesis subject / assignment - leakage through a dam, either concrete or embankment, is a vital indicator of the dam's status. The rate of leakage is often monitored with triangular weirs due to their high sensitivity to changes in flow (compared to, for instance, rectangular ones). The relationship used between water level and flow rate is typically borrowed or extracted via a simple curve-fitting procedure, which is not very accurate, particularly with high leakage. Machine learning offers a fresh way of creating a more reliable relationship.
• Thesis goal / purpose - accurate leakage prediction offers a reliable foundation for dam monitoring and decision-making. A machine-learning-based relationship applies to a broad range of leakages; even abnormal measurement data can be flagged with an alert (for example, damages in the weir or blockage by fallen trees or debris). Machine learning procedures are created to achieve a more accurate flow relationship for dam leakage monitoring, based on extensive experimental data.
• This task is suitable for either one or two students.

Execution

For a triangular weir with varying opening angles (usually between 30⁰ and 120⁰), a dimensional analysis of its governing parameters is initially conducted. Flow data from various hydraulic labs are gathered and normalized. Based on literature, two to three machine learning algorithms are used to derive generalized relationships between water level and flow rate. The most accurate one (based on several criteria) is recommended for dam leakage monitoring.

For the students

• To become acquainted with the ongoing dam safety and dam upgrade projects in the country
• To learn about hydraulic modeling tests
• To understand the basic principles of machine learning techniques
• To develop a useful monitoring procedure for leakage monitoring.

The project encompasses the following aspects.

1. Background and literature review
2. Collection of weir flow data from the laboratory
3. Understanding general machine learning methods
4. Method development
5. Comparison with former results
6. Report writing

Qualifications

We're searching for a student to join us on our journey towards a fossil free future. You are just about to graduate from your academic studies. You also align with our principles: being active, open, positive and safe.

• Education - an appropriate background is university education in civil (hydraulic) engineering, engineering physics or similar disciplines.
• Scope - the project corresponds to a 30 hp diploma work.
• Fluency in written English is essential.
• Strong interest in hydropower, dam safety and hydraulic engineering.
• The time frame for execution should be within four (4) months, to be completed in 2024.

Additional Information

• Start of assignment: January 17, 2024 or as agreed
• Location: Älvkarleby, Stockholm, or as agreed

• Application – submit a combined document with your CV, cover letter, and university grades in the application.
• The deadline for application is December 17, 2023. Incoming applications are processed continuously, and the position will be filled when a suitable candidate is found. For this reason, it's recommended not to wait until the deadline to submit your application, complete with contact information.
• Contact person and supervisor at Vattenfall: James Yang, [email protected], phone 070-2723 200.

Diversity & inclusion - in all we do

At Vattenfall, we value the idea that diverse teams outperform homogeneous ones. However, we can only harness and utilize the potential of diversity if everybody feels included. For more information about our work on diversity and inclusion, visit