Data for 'Uncertainty-Aware Capacity Calculation and Congestion Management'

This dataset accompanies the thesis "Uncertainty-Aware Capacity Calculation and Congestion Management".

It includes all input data, forecast scenarios, and simulation results for two case studies modeled using the open-source framework `munacco`. All result files are provided in .pkl (Python pickle) format for efficient loading in Python environments.

Dataset Structure

thesis-data/
├── stylised-4node/
│   ├── base_model/                  # Input data (CSV) for the stylised 4-zone test system
│   ├── 01_analyzer_data_*.pkl      # Results from Uncertainty Impact on capacity calculation and validation (Section 4.1.1) 
│   ├── 02_meshpoints_*.pkl         # Domain meshpoints for plotting
│   ├── 03_multiple_scenarios.pkl   # Results from multiple scenario experiment (Section 4.1.3)
│   └── 04_results_sensitivity_*.pkl# Risk sensitivity experiments (Section 4.1.4)
│
└── pypsa-eur-50node/
    ├── cwe_network_data/           
    │   ├── config/                 # Model configuration files (.yaml)
    │   ├── input_data/            # PyPSA-Eur derived data for the CWE region
    │   │   ├── RES availability (NetCDF)
    │   │   ├── Load profiles
    │   │   ├── Bus and line mappings
    │   │   ├── Power plant data
    │   │   └── Regional shapes (GeoJSON)
    │   ├── networks/              # PyPSA networks (.nc) used as the base for simulations
    │   └── solved_network/        # Pre-solved network instance
    │
    └── result_data/ # Results for 50 node case stud (Section 4.3.3)
        ├── all_data_det_*.pkl     # All scenario-level KPIs for deterministic validation
        ├── all_data_robust_*.pkl  # All scenario-level KPIs for robust (chance-constrained) validation
        ├── snapshot_overview_det_*.pkl    # Aggregated KPIs per snapshot (deterministic)
        └── snapshot_overview_robust_*.pkl # Aggregated KPIs per snapshot (robust)

Notes

•  All simulations were performed using a modified version of `PyPSA-Eur` (for the 50-node case) and CSV-based models (for the 4-node case).
•  Forecast scenarios were generated with stochastic RES models, including configurable forecast error levels and timing.
•  The data enables full reproduction of the model experiments discussed in the thesis and supports further analysis of uncertainty-aware zonal capacity calculation.
•  Some experiments use random draws without fixed seeds. Exact results may differ slightly.

For the simulation code, see the accompanying repository:  
https://github.com/mvoitl/munacco