Green hydrogen can help store surplus renewable electricity and decarbonise sectors that are difficult to electrify directly. However, hydrogen production from wind power is highly variable and depends on uncertain factors such as wind conditions, turbine performance, wake losses, electrolyser efficiency and operational constraints.
This project will build on an existing model of the proposed Firlough Wind Farm and Hydrogen Plant in Ireland. The existing model uses 34 years of half-hourly wind data to simulate wind-farm power and hydrogen production. It provides a useful baseline, but most model parameters are fixed and its reported uncertainty primarily represents differences between windy and calm years. It therefore does not fully quantify uncertainty in the wind resource, turbine and wake models, electrolyser performance, system losses or operating conditions.
The aim of this project is to develop a more comprehensive, uncertainty-aware model and investigate how uncertainty affects predicted hydrogen production and system-design decisions.
The student will first reproduce and examine the existing modelling results. They will then identify the most important uncertain inputs, which may include wind-resource bias, turbine power and thrust curves, wake-model parameters, turbine availability, electrical losses, electrolyser efficiency, auxiliary consumption and minimum operating load. Suitable probability distributions or structured scenarios will be assigned using published data.
Monte Carlo or other efficient sampling methods will be used to propagate these uncertainties through the wind-to-hydrogen model. Global sensitivity analysis, such as Morris screening or Sobol indices, will then identify which assumptions have the greatest influence on annual hydrogen production and low-production years.
Depending on progress and the student’s interests, the project may also examine the optimal size and operation of the electrolyser plant. Possible extensions include hydrogen storage, trailer-collection constraints, module start-up and shutdown, degradation, and the allocation of wind electricity between hydrogen production and grid export. The analysis could investigate whether the currently proposed 80 MW electrolyser—or the approximately 70 MW capacity suggested by the existing physical model—remains preferable when uncertainty and operational constraints are considered.
Expected outputs include:
- A reproducible probabilistic wind-to-hydrogen model;
- Uncertainty ranges for annual hydrogen production;
- A ranking of the assumptions and parameters that most strongly influence the results;
- An assessment of the robustness of alternative electrolyser capacities;
- Recommendations on what additional data would most improve future predictions.
The project is intended for MSc in Statistics and Sustainability students. Experience with Python is required. Prior knowledge of wind energy or hydrogen systems is not required, as the relevant engineering background will be developed during the project.