This GitHub repository provides the 150 MV feeders with all underlying LV networks used in Seddik Yassine Abdelouadoud's study. These networks were selected from among French distribution networks as being representative of it. The details of the procedure used to select them are available in a report accessible via the previous link (French only).
In this repository, an updated version of the files representing these networks is available for direct use with Roseau Load Flow.
In order to start using these networks, roseau-load-flow>=0.12.0 must be installed. It can be achieved via uv:
# Install a virtual environment (optional)
uv python pin 3.13
uv venv
# Use the `pyproject.toml` and `uv.lock` files provided in this repository
uv syncor via pip:
# In your virtual environment
pip install roseau-load-flow>=0.12.0For more information, see the "Installation" page of the Roseau Load Flow documentation available here.
The networks can be used directly in the following way:
import roseau.load_flow as rlf
# Read a network from this repository
en = rlf.ElectricalNetwork.from_json("./data/networks/11_MVFeeder0725.json")
# <ElectricalNetwork: 336 buses, 324 lines, 10 transformers, 1 switch, 626 loads, 1 source, 1 ground, 1 potential ref, 11 ground connections>
# Solve a load flow
# The environment variable "ROSEAU_LOAD_FLOW_LICENSE_KEY" must be defined in your environment
en.solve_load_flow()
# (3, 2.1200179389779805e-09)
# Access the results
# By instance, voltages as data frame
en.res_buses_voltages
# voltage ... nominal_voltage
# bus_id phase ...
# MVVoltage_source an 11777.945491+ 0.000000j ... NaN
# bn -5888.972746-10200.000000j ... NaN
# cn -5888.972746+10200.000000j ... NaN
# 11_BUZEN ab 17666.918237+10200.000000j ... 20000.0
# bc 0.000000-20400.000000j ... 20000.0
# ... ... ... ...
# 11_LVBus0536527 bn -4.022833- 230.169477j ... 400.0
# cn -196.023130+ 119.090298j ... 400.0
# 11_LVBus0536528 an 202.935062+ 115.256587j ... 400.0
# bn -4.011705- 230.092439j ... 400.0
# cn -195.982822+ 119.122054j ... 400.0
# [1008 rows x 6 columns]For more information about the license key for Roseau Load Flow, please visit this page.
The name of the networks' files can be read as follows: for instance, with 11_MVFeeder0725.json:
11is the code of the French region the network is located (see below);MVFeeder0725is the (internal) name that was given to this MV feeder.
The regional codes are as follows. Only French metropolitan regions were considered in this study.
| Code | Name |
|---|---|
| 11 | Île-de-France |
| 24 | Centre-Val de Loire |
| 27 | Bourgogne-Franche-Comté |
| 28 | Normandie |
| 32 | Hauts-de-France |
| 44 | Grand Est |
| 52 | Pays de la Loire |
| 53 | Bretagne |
| 75 | Nouvelle-Aquitaine |
| 76 | Occitanie |
| 84 | Auvergne-Rhône-Alpes |
| 93 | Provence-Alpes-Côte d'Azur |
See this Wikipedia page for more information.
This study also provides a size for each cluster, indicating its representativeness within France. These sizes are stored in the file ./data/Cluster_Size.csv.
For example, the representativeness size of the network 11_MVFeeder0725 is 326.
| network_id | cluster_size |
|---|---|
| 11_MVFeeder0725 | 326 |
| 11_MVFeeder1363 | 258 |
| 11_MVFeeder1365 | 408 |
| 11_MVFeeder2143 | 232 |
| 11_MVFeeder2298 | 151 |
| ... | ... |
The total cluster size is 24135, so the network 11_MVFeeder0725 represents 1.3507% of the French distribution network.
The scripts/ folder contains two scripts:
convert_networks.py: Downloads network files in their original format and converts them to the latest Roseau Load Flow JSON format.plot_networks.py: Plots the networks using the folium library and generates regional maps showing all networks within each region.
To run these scripts:
uv run scripts/convert_networks.py
uv run scripts/plot_networks.pyThe following section is a translation of the dataset description made available by Mr. Abdelouadoud on the data.gouv.fr website.
This dataset provides a selection of representative medium-voltage (MV) feeders from French distribution networks. Designed to facilitate planning and operational studies of electrical networks, it is a key resource for researchers, engineers, and decision-makers.
- MV and LV Feeder Traces: Line geometries, including location and typology information (overhead or underground).
- Technical Characteristics: Conductor section and type per line, rated power of MV/LV transformers.
- Consumption: Examples of consumption drawn per node and per phase.
This dataset was developed from open data from reliable sources, such as network traces published by the ORE agency, local energy consumption data, and building geometries provided by the Institut Géographique National (IGN i.e. National Geographic Institute) (BDTOPO). The feeders were selected using advanced clustering algorithms, ensuring a balanced distribution of representative cases across the French territory.
- Network Planning: Assessment of hosting capacities and anticipation of local constraints.
- Energy Studies: Simulation of scenarios for the deployment of renewable energies and electrification.
- Academic Research: Development and testing of new electrical network modeling methods.
The dataset is made available under the Open License version 2.0, promoting its use, reuse, and sharing. You can download the files and access the full documentation via this page.
For a comprehensive report describing the methodology: Methodological Report
Contribute to the future of electrical networks by exploring and utilizing this unique dataset in France!