MSc Dissertation · 2026
Educational PEM Fuel Cell Performance
Performance Analysis of an Educational Proton Exchange Membrane Fuel Cell Stack. My dissertation examined the operation of a reversible educational PEM fuel-cell system through repeated bench tests, alongside a separate analysis of automotive-scale simulation outputs.
Project at a Glance
- Approach
- Ten educational bench runs, with a separate automotive-scale simulation analysis.
- Key Finding
- Run-to-run variation makes controlled measurements and repeatability the next research priority.
- Tools
- Electrical measurements · Excel · MATLAB / Simscape
- Evidence
- Experimental setup, original dissertation extract and research presentation.
Research Question
How does the electrical output and operating duration of a small educational PEM system vary during discharge and across repeated runs? The work also examined how an automotive-scale model represents electrical and thermal behaviour under a varying load.
My Contribution
I carried out practical tests, recorded hydrogen filling and discharge duration across ten runs, measured voltage and current, calculated electrical power and organised the results in Excel. I also analysed MATLAB/Simscape outputs, reviewed relevant literature and discussed the limits of comparing the experiment with the simulation.
Experimental Method
The educational setup used a Horizon FCJJ-11 reversible PEM teaching kit. The study examined electrolysis and subsequent fuel-cell operation, recording filling time, operating duration and electrical measurements during discharge. Power was calculated from voltage and current. Repeated runs provided a descriptive comparison of operating behaviour.
Findings
The ten bench runs showed variation in operating duration and electrical output during discharge. Hydrogen supply, sealing and measurement consistency are important factors to control in follow-up tests. Separately, the automotive-scale simulation showed voltage changes with current demand and a warm-up followed by a more stable stack temperature. The two studies do not establish quantitative agreement between the bench system and the model.
Simulation Scope
The automotive-scale Simscape analysis includes electrochemical, gas-flow, electrical-load and thermal subsystems. Figure 4.8 shows a varying load over approximately 2,500 seconds. My contribution focused on interpreting the electrical and thermal outputs using the physical modelling framework.
Duration Data
Table 4.2 labels run time in minutes and records entries such as 3.40 and 2.30. The original timing log is needed to confirm decimal minutes versus minutes-and-seconds notation. The table below preserves those entries as text; no converted durations or headline average are claimed.
Limitations
The educational bench system was not a direct validation platform for the automotive-scale model. The ten runs are a small descriptive dataset, and the published figures do not establish measurement uncertainty or isolate the causes of variation. Reported durations should not be used to infer automotive-stack efficiency or durability.
Calculation Review
A review of the original tables found that the percentages labelled as experimental efficiency do not establish hydrogen-to-electricity efficiency. The power-percentage pairs imply a reference near 1.24–1.25 W, whose physical meaning is undocumented. Several reported powers also differ from the tabulated voltage-current products. These quantities require reconciliation with the original records; no corrected experimental efficiency is claimed. Efficiency requires electrical energy output divided by hydrogen energy input over the same interval, with the measurement boundary and heating-value convention specified.
Research Materials
Both downloads begin with a dated technical review notice. The original academic pages are retained as a historical record; the notice supersedes their experimental efficiency interpretation and direct simulation-validation claims. Use the current portfolio summary for the supported scope of the work.
Publication Status
I am preparing a research manuscript based on this dissertation for journal submission. It is currently a manuscript in preparation, rather than a published or accepted paper.
Directions I Would Like to Develop
A future study could use controlled load conditions, synchronised measurements and an explicit uncertainty analysis. I would like to develop a model matched to the tested stack and compare its predictions with repeatable experiments before extending the analysis to parameter sensitivity and thermal management.
Recorded Run Times
| Run | Recorded entry (labelled min) |
|---|---|
| 1 | 3.40 |
| 2 | 3.28 |
| 3 | 2.35 |
| 4 | 2.30 |
| 5 | 3.20 |
| 6 | 3.32 |
| 7 | 4.25 |
| 8 | 3.28 |
| 9 | 3.26 |
| 10 | 4.15 |
Dissertation and Presentation
Technical review · 8 October 2026. Each PDF begins with a review notice. Experimental efficiency percentages and direct validation claims in the original academic pages are superseded by that notice. Original timings and several power calculations still need verification.
Source: MSc dissertation, Performance Analysis of an Educational Proton Exchange Membrane Fuel Cell Stack, dated 27 April 2026; original research presentation and progress report supplied 7 October 2026. Academic research, not a published journal article.