PhD Studentship - Prediction of the Interaction between Ocean Waves and Tidal Turbine Arrays through Machine Learning

Research theme: "NetZero", "Energy", "Decarbonisation", "AI in Engineering"

How to apply: uom.link/pgr-apply-2425

No. of positions: 1

This is a fully-funded Strategic Doctoral Landscape Award (EPSRC-DLA) Scholarships from the School of Engineering. Funding to Home / UKRI level (£19,237 for 2024/25) is preferred and enquiries from exceptional overseas candidates welcomed (which need to have an excellent MSc/MEng thesis in a related topic, and/or a track record of scientific publications and/or experience working in the offshore renewable energy industry).

This project is also eligible for the Osborne Reynolds top-up Scholarship which provides an additional £1,500 per year top-up to other funding sources for outstanding candidates. Successful applicants will be automatically considered for this top-up.

Tidal-stream turbines operate in harsh environmental conditions, with ocean waves likely leading to extreme conditions that can trigger large structural loads or modulate the turbine’s wake dynamics. This project will adopt and extend an in-house high-fidelity numerical simulation tool (DOFAS – Digital Offshore FArms Simulator) to represent realistic ocean conditions for bottom-fixed and floating tidal turbines. This numerical model is essential to accurately quantifying the wave-current-turbine interaction, as they allow to control the variation of the wave’s characteristics, turbine operating point and environmental turbulent flow conditions. Understanding of all the former conditions is critical to inform industry about the performance of tidal turbines and to develop a machine learning model trained on flow and turbine loading data from a set of flow conditions that would allow to extrapolate to the any wave climate (wave height and period) to could be found at any tidal site.

High-fidelity simulation data will be used to build the machine-learning model to be then embedded into a GPU-accelerated blade-element momentum solver that will be made open-source to enable direct impact with world-leading industry, with whom the supervisory team has established collaborations for more than a decade. This multi-fidelity model integration is ground-breaking.

The project has two main objectives:

O1: Generation of a comprehensive dataset from large-eddy simulations using actuator line method which also needvalidation for laboratory- and full-scale tidal-stream turbines.

O2: Inform a machine-learning using both LES and blade-element momentum code to enable turbine loading calculations at unprecedented speed which will be expanded for a wider range of wave conditions.

Candidates must have a 1st or high 2i in a degree, ideally at Masters level, in an Engineering subject, Physics, Mathematics, or Atmospheric Sciences. Knowledge in fluid mechanics, numerical methods and computational modelling is necessary. The student is expected to have prior experience writing codes on Fortran or C/C++, and experience on Linux systems. No previous experience with machine learning is required, although it would be advantageous. The ideal candidate is expected to have a strong interest in renewable energy, be enthusiastic about physics-based computational modelling, be able to have a proactive attitude towards problem solving independently, and ability to work in multidisciplinary teams.

This project has to start between 1^st June 2025 and 30^th September 2025.

To apply, please contact: Dr Pablo Ouro (pablo.ouro@manchester.ac.uk) and Prof Tim Stallard (tim.stallard@manchester.ac.uk). Please include details of your current level of study, academic background and any relevant experience and include a paragraph about your motivation to study this PhD project.

Advert information