Code and Data associated to the manuscript "Ensuring Reliability for Socio-Ecological Transformation of the Chemical Industry supported by Chance Constrained Life Cycle Optimization"

This is a reproducible case study demonstrating the uncertainty-analysis capabilities of PULPO (v1.5.1) on a European ammonia-production system. It accompanies the methodological paper introducing chance-constrained (CC) life-cycle optimization and shows how epistemic and aleatory uncertainties can be propagated through a real-world prospective LCA. The repository contains two self-contained Jupyter notebooks: 01_uncertainty_case.ipynb – full CC pipeline: LCI import (ecoinvent 3.10 cutoff + foreground in data/ammonia.xlsx), uncertain IPCC 2013 GWP100 characterisation, uncertainty-strategy assignment, chance-constrained Pareto sweep over risk level λ, and Sobol global sensitivity analysis across multiple λ settings. 02_uncertainty_vs_MC.ipynb – comparison of the analytical CC formulation against Monte Carlo sampling, analytical impact PDFs, and technology-choice analysis. Python 3.11, uv, Brightway2, Pyomo, Gurobi 13.0.0, SALib, JupyterLab. A valid ecoinvent 3.10 licence is required to reproduce the results. Reproducibility: dependencies are fully pinned in pyproject.toml and uv.lock; run uv sync followed by the notebooks in order. Intermediate results are cached under data/results/.