# README.txt

## Title of the dataset
Forecasting Price Electrolysers — An Empirically Grounded Approach to Forecast the Cost of Electrolysers

## General introduction
This code supports the MSc thesis project "Forecasting Price Electrolysers — An Empirically Grounded Approach to Forecast the Cost of Electrolysers" at Delft University of Technology (TU Delft). The project investigates probabilistic deployment and cost scenarios of Alkaline Electrolysis Cells (AEC) and Proton Exchange Membrane (PEM) electrolyzers using stochastic logistic (S-curve) models and Wright’s Law formulations.

The repository is intended to enable reproducibility of thesis figures and tables and to support further research in green hydrogen cost modeling.

## Important note on data dependency
This code and dataset are part of an integrated set of files. The datasets described in the complementary README file are required to run this code successfully. Both sets of code and data must be used together, as they depend on each other.

## File and folder descriptions

### Code files
- Deployment AEC.py: Probabilistic logistic forecast for AEC capacity using Monte Carlo simulation.
- Deployment PEM.py: Probabilistic logistic forecast for PEM capacity.
- CAPEX AEC.py: Wright’s Law-based stochastic cost forecast for AEC using the deployment forecast as input.
- CAPEX PEM.py: Wright’s Law-based stochastic cost forecast for PEM.
- Historical CAPEX data for AEC and PEM (2003-2023).py: Visualization of historical CAPEX data points and median trends for AEC and PEM.

### Data files
- DEPLOYMENT FORECAST AEC.xlsx: Percentile-based deployment forecasts of AEC capacity (GW), including historical data and simulated scenarios.
- DEPLOYMENT FORECAST PEM.xlsx: Same as above, for PEM.
- CAPEX DATA.xlsb: Historical CAPEX data for AEC and PEM electrolyzers (2003–2023), sourced from Glenk & Reichelstein and IEA.

## File structure relationships
The deployment forecast Excel files are generated by Deployment AEC.py and Deployment PEM.py scripts. These forecasts are then used by CAPEX AEC.py and CAPEX PEM.py to generate cost projections.

## Methodological information

### Data generation and modeling
- Deployment projections use a logistic S-curve model with stochastic parameters: initial capacity, growth rate, and anticipated demand pull.
- Cost projections use Wright’s Law extended with stochastic shock propagation, linking cumulative deployment to projected cost reductions.
- Historical CAPEX data compiled from Glenk & Reichelstein (2003–2018) and IEA (2019–2023).

### Software and dependencies
- Python 3.10+
- Packages: numpy, pandas, matplotlib, scipy, scikit-learn, tqdm, openpyxl, pyxlsb

Example installation:
pip install numpy pandas matplotlib scipy scikit-learn tqdm openpyxl pyxlsb

## Data-specific information

### Column definitions (Excel outputs)
- Year: Calendar year (e.g., 2025–2050).
- Deployment capacity: Electrolyzer capacity in GW.
- CAPEX: Capital expenditure in EUR/kW (inflation-adjusted).

### Units
- Capacity: GW
- Cost: EUR/kW

### Missing values
No explicit missing values; rows without data points are omitted.

## Sharing and access information

### Licenses
- Code: MIT License — you are free to use, copy, modify, and distribute the code with attribution.
- Data: Creative Commons Attribution 4.0 International (CC BY 4.0) — you may share and adapt with attribution.

## Citation
If you use this dataset or code, please cite:

Bram van Eijden, "Forecasting Price Electrolysers — An Empirically Grounded Approach to Forecast the Cost of Electrolysers," MSc Thesis, Delft University of Technology, Faculty of Technology and Policy Management, 2025.

## Contact
Author: Bram van Eijden
Faculty: Faculty of Technology, Policy and Management, Delft University of Technology

## Supervisors
- Dr. S.J. Pfenninger-Lee (Chair)
- Dr. F. Lombardi (First supervisor)
- Dr. S. Azimi Rashti (Second supervisor)
- Ir. I. Ruiz Manuel (Daily supervisor)

---

This README file was created following the 4TU.ResearchData guidelines (2020).