Data Sources#

This page documents all external data sources used in PyPSA-GB.

NESO Future Energy Scenarios (FES)#

The primary source for future capacity projections.

Overview#

Attribute

Details

Publisher

NESO (formerly National Grid ESO)

Frequency

Annual (typically July)

Coverage

GB electricity system 2025-2050

Granularity

Technology × GSP × Year

Access

Open API + Excel files

Available Releases

FES 2021, 2022, 2023, 2024, 2025

FES Releases and Scenarios#

NESO publishes updated FES reports annually. You can model any FES release to compare how projections evolve.

Data Fields#

FES provides capacity by:

  • Technology: Wind, Solar, Nuclear, CCGT, etc.

  • Technology Detail: Offshore Wind, Onshore Wind, etc.

  • GSP: Grid Supply Point (connection location)

  • Year: 2025-2050

Location#

data/FES/
├── FES_api_urls.yaml          # API endpoints
├── FES_2024_*.xlsx            # Raw Excel downloads
└── gsp_boundaries/            # GSP geographic data

Usage in Model#

# Processed FES data - example for 2024 release
resources/FES/FES_2024_data.csv

When configuring a scenario in config/scenarios.yaml, specify both the FES release year and the pathway name:

HT35:
  modelled_year: 2035
  FES_year: 2024              # Which FES release to use
  FES_scenario: "Holistic Transition"  # Which pathway within that release

For future years (modelled_year > 2024), PyPSA-GB uses the specified FES release to provide:

  • Thermal generation capacity

  • Renewable capacity by technology

  • Storage capacity (GSP-level)

  • Demand projections: ED1 consumer demand totals and Dem_BB003 GSP shares

Demand accounting: future demand does not use raw Dem_BB003 as the national total. Dem_BB003 is GSP-facing demand and is used as the spatial allocation across GSPs. The national annual target is read from the FES workbook ED1 sheet as total consumer electricity demand, then allocated to GSPs using the Dem_BB003 shares.

Note: Different FES releases may have different scenario names and capacity values for the same modelled year. Use this to explore sensitivity to FES updates.

DUKES (Digest of UK Energy Statistics)#

Historical power station data from UK Government.

Overview#

Attribute

Details

Publisher

DESNZ (UK Government)

Frequency

Annual (March/July)

Coverage

All major UK power stations

Granularity

Individual power stations

Access

Open data (Excel)

Key Tables#

Table

Content

DUKES 5.11

Power stations (capacity, fuel, location)

DUKES 5.10

Plant efficiencies

DUKES 5.6

Fuel input statistics

Data Fields#

  • Station name

  • Installed capacity (MW)

  • Fuel type

  • Company/owner

  • Commissioning year

  • Grid reference (for mapping)

Location#

data/generators/
├── DUKES_5.11_2025.xlsx       # Latest DUKES
├── DUKES_5.10_*.xlsx          # Efficiency data
└── fuel_prices/               # Fuel cost data

Usage in Model#

For historical years (modelled_year ≤ 2024), DUKES provides:

  • Thermal generator capacities

  • Fuel types and efficiencies

  • Approximate locations

REPD (Renewable Energy Planning Database)#

Site-level data for all renewable projects in the UK.

Overview#

Attribute

Details

Publisher

DESNZ (UK Government)

Frequency

Quarterly

Coverage

All renewable projects (planned + operational)

Granularity

Individual sites

Access

Open data (CSV)

Data Fields#

  • Site name and reference

  • Technology type

  • Installed capacity (MW)

  • Development status (Operational, Under Construction, etc.)

  • Location (postcode, coordinates)

  • Planning authority

  • Connection voltage

Location#

data/renewables/
├── repd-q2-jul-2025.csv       # Latest REPD
└── repd_technology_mapping.yaml

Usage in Model#

  • Historical: Direct use of operational sites

  • Future: Distribution patterns for FES capacity

Technology Mapping#

REPD Technology

PyPSA Carrier

Wind Onshore

wind_onshore

Wind Offshore

wind_offshore

Solar Photovoltaics

solar_pv

Battery

battery

Hydro

large_hydro / small_hydro

ETYS (Electricity Ten Year Statement)#

Transmission network data from NESO.

Overview#

Attribute

Details

Publisher

NESO

Frequency

Annual (November)

Coverage

GB transmission network

Granularity

Individual circuits and transformers

Access

Open data (Excel appendices)

Key Appendices#

Appendix

Content

B-1

Substation data (buses)

B-2-1a/b/c/d

Base circuit data (SHE-T / SPT / NGET / OFTO)

B-2-2a/b/c/d

Circuit upgrades (SHE-T / SPT / NGET / OFTO)

B-3-1a/b/c/d

Base transformer data (SHE-T / SPT / NGET / OFTO)

B-3-2a/b/c/d

Transformer upgrades (SHE-T / SPT / NGET / OFTO)

B-5-1

HVDC data

Data Fields#

Substations (Buses):

  • Name and identifier

  • Voltage level (400kV, 275kV, 132kV)

  • Coordinates (OSGB36)

Circuits (Lines):

  • From/To buses

  • Resistance, Reactance, Susceptance

  • Thermal rating (MVA)

  • Length (km)

Location#

The ETYS publication year is selected via etys.year in config/defaults.yaml (supports 2022, 2023, 2024). Files are mapped by scripts/network_build/etys_file_registry.py.

data/network/
├── ETYS/
│   ├── ETYS 2024 Appendix-B V1.xlsx    # ETYS 2024 (default)
│   ├── ETYS Appendix B 2023.xlsx        # ETYS 2023
│   ├── ETYS Appendix B 2022.xlsx        # ETYS 2022
│   ├── GB_network.xlsx                  # Offshore WF edges, BMU mappings
│   └── substation_coordinates.csv       # Bus coordinate overrides
├── reduced_network/
│   ├── buses.csv
│   └── lines.csv
└── zonal/
    ├── buses.csv
    └── links.csv

ERA5 Weather Data#

Reanalysis weather data for renewable generation profiles.

Overview#

Attribute

Details

Publisher

ECMWF / Copernicus

Frequency

Hourly

Coverage

Global (subset for GB)

Granularity

~30km grid

Access

Zenodo (2010-2024) or CDS API

Zenodo Record

10.5281/zenodo.18325225

Variables Used#

  • 100m wind speed (for wind power)

  • 10m wind speed

  • Surface solar radiation

  • Temperature (for PV efficiency)

  • Runoff (for hydro)

Location#

resources/atlite/cutouts/
├── uk-2019.nc                 # Weather cutout for 2019
├── uk-2020.nc
├── uk-2021.nc
└── ...

Acquisition Strategy#

PyPSA-GB uses a tiered acquisition strategy for weather cutouts:

  1. Data directory - Check data/atlite/cutouts/ for cached files

  2. Zenodo - Download pre-built cutouts (~5-10 minutes per year)

  3. ERA5 API - Generate from scratch via atlite (~2-4 hours per year)

Quick Start (Years 2010-2024)#

For years 2010-2024, cutouts are automatically downloaded from Zenodo:

# Configure desired years in config/cutouts_config.yaml
snakemake -s Snakefile_cutouts --cores 1

No CDS API credentials required for these years!

Custom Years (Outside 2010-2024)#

For other years, you’ll need CDS API credentials:

# 1. Register at: https://cds.climate.copernicus.eu/user/register
# 2. Set up ~/.cdsapirc with your API key
# 3. Generate cutout
snakemake -s Snakefile_cutouts --cores 1

Pre-built Cutouts (Zenodo)#

Pre-built cutouts for years 2010-2024 are available on Zenodo:

  • Repository: PyPSA-GB Atlite Cutouts

  • License: CC-BY-4.0

  • File size: ~765 MB per year

  • Download time: 5-10 minutes per year

  • MD5 verification: Automatic

These cutouts are automatically used by the workflow and require no manual intervention. | Frequency | Half-hourly | | Coverage | GB total demand 2009-2024 | | Granularity | National total | | Access | Open data (CC-BY-4.0) |

Data Fields#

  • Timestamp

  • Demand (MW)

  • Settlement period

Location#

data/demand/
├── ESPENI/
│   ├── demand_2019.csv
│   ├── demand_2020.csv
│   └── ...
└── profiles/

Usage#

  • Historical years: Direct use

  • Future years: Profile shape only; annual totals come from FES ED1 consumer demand and are spatially allocated using Dem_BB003 GSP shares

TEC Register#

Transmission Entry Capacity from NESO.

Overview#

Attribute

Details

Publisher

NESO

Frequency

Monthly

Coverage

All transmission-connected generators

Access

Open data (CSV)

Data Fields#

  • Generator name (BMU ID)

  • TEC capacity (MW)

  • Connection point

  • Fuel type

  • Effective dates

Location#

data/generators/
├── tec-register-july-2025.csv
└── tec_fuel_mapping.yaml

Usage#

Used to cross-reference DUKES data and validate capacities.

Data Quality Notes#

Market And Validation Data#

Market scenarios can use additional historical operational data:

Source

Used For

Notes

ELEXON BMRS BOD

Historical BM bid and offer prices

Used when market.balancing.bid_offer_source: "elexon" or when auto resolves to ELEXON for historical scenarios.

ELEXON BOAV/BOALF

Historical BM validation

Used by validate_bm_results to compare model redispatch with accepted bid/offer actions.

ELEXON MID

Wholesale price validation

Used by wholesale notebooks and validation plots for historical scenarios.

NESO thermal constraint data

Constraint-cost validation

Used by validate_neso_constraints to compare model constraint costs and boundary flows.

ESPENI generation by fuel

Physical dispatch validation

Used as an independent comparison for historical balancing dispatch.

These files are cached under data/market/, data/validation/, or resources/market/ depending on whether they are persistent raw inputs, validation caches, or per-scenario processed outputs. See Market Dispatch for the workflow entry points.

Known Issues#

Source

Issue

Mitigation

DUKES

Missing coordinates for some stations

Manual geocoding

REPD

Duplicate entries

Deduplication script

ETYS

Some circuits under construction

Filter by status

FES

GSP names change between years

Mapping tables

Validation#

The model validates data during build:

python scripts/validate_scenarios.py

This checks:

  • All generators have valid coordinates

  • Capacity totals match expected values

  • No orphan buses in network

Comprehensive Data Sources Table#

Type

Data

Data Processing

Source and License

Network

Reduced network model

Matpower file converted to buses and lines

Bell and Tleis, Bukhsh et al. – GPL-3.0

Zonal model

Excel data converted to buses and links

National Grid’s ETYS – NG ESO Open Data Licence v1.0

Electrical Demand

ESPENI

Excel file converted to loads

Wilson et al. – CC-BY-NC-4.0

Marginal Costs

Fuel costs from FES

Excel data converted to marginal price with addition of EU-ETS and CPS

National Grid’s FES – NG ESO Open Data Licence v1.0

EU-ETS

Excel data converted to marginal price in addition to fuel costs and CPS

Ember Climate – CC-BY-4.0

Carbon Price Support (CPS)

Excel data converted to marginal price in addition to fuel costs and EU-ETS

UK Gov BEIS – Open Government Licence v3.0

Thermal Power Plants & Hydropower

Historical data (location, fuel, type, capacity)

Conversion from Excel data to generators

DUKES dataset – Open Government Licence v3.0

Coordinates

Coordinate data converted to generator attributes

OpenStreetMap (ODbL), Global Energy Monitor (CC-BY-4.0), Google Maps

Technical characteristics

Data from papers converted to generator attributes

Schröder et al., Angerer et al.

Hydropower power output timeseries

Excel data converted to generator power timeseries

Elexon – BSC Open Data Licence

Future installed capacities and locations

Excel data converted to generator attributes

National Grid’s FES – NG ESO Open Data Licence v1.0

Renewable Power

Renewable power timeseries

ERA5 weather data converted to generator power timeseries using Atlite

ERA5 (CC-BY-4.0), Atlite (MIT)

Historical location, type, capacity

Excel data converted to generator attributes

REPD – Open Government Licence v3.0

Historical annual generation

Report data used to scale REPD for <150kW installations

UK Gov BEIS – Open Government Licence v3.0

Future installed capacities and locations

Excel data converted to generator attributes

National Grid’s FES – NG ESO Open Data Licence v1.0

Offshore wind (near-term)

Report data for near-term spatial distribution

REPD, Scottish Sectoral Marine Plan – OGL v3.0

Tidal lagoon and stream

Generator power timeseries using Thetis coastal ocean model

Thetis – MIT

Wave power generation

ERA5 wave climate converted using power matrix

ERA5 (CC-BY-4.0), Power Matrix

Storage

Historical pumped hydro

Excel data converted to storage unit attributes

DUKES dataset – Open Government Licence v3.0

Future storage capacities and locations

Excel data converted to storage unit attributes

National Grid’s FES – NG ESO Open Data Licence v1.0

Efficiency and losses

Report data converted to storage attributes

Moseley and Garche

Emissions

Direct emissions factors

Data from various sources converted to carbon factors

Staffell, Schlömer et al.