Your First Custom Scenario#

This tutorial walks you through creating and running a custom scenario.

Understanding Scenarios#

A scenario in PyPSA-GB defines:

  • The modelled year (historical or future)

  • Which FES pathway to use

  • Network model type (ETYS, Reduced, Zonal)

  • Solve settings and time period

Scenarios are defined in config/scenarios.yaml and activated in config/config.yaml.

Step 1: View Existing Scenarios#

Open config/scenarios.yaml to see example scenarios:

# Example scenario definition
HT35:
  description: "Holistic Transition 2035"
  modelled_year: 2035
  renewables_year: 2019
  demand_year: 2035
  network_model: "ETYS"
  FES_year: 2024
  FES_scenario: "Holistic Transition"
  solve_period:
    enabled: true
    start: "2035-01-01 00:00"
    end: "2035-01-07 23:00"

Step 2: Create Your Scenario#

Add a new scenario to config/scenarios.yaml:

# My custom scenario
MyScenario_2040:
  description: "Custom 2040 analysis - winter week"
  modelled_year: 2040
  renewables_year: 2018          # Weather year for renewables
  demand_year: 2040
  network_model: "Reduced"       # Use 32-bus for faster solving
  FES_year: 2024
  FES_scenario: "Electric Engagement"
  solve_period:
    enabled: true
    start: "2040-01-13 00:00"    # A winter week
    end: "2040-01-19 23:00"
  solver:
    name: "gurobi"
    method: 2                    # Barrier method

Key Configuration Options#

Parameter

Description

Options

modelled_year

Target year to model

2010-2050

renewables_year

Weather data year

2010-2024 (auto-download from Zenodo)

network_model

Network resolution

ETYS, Reduced, Zonal

FES_year

FES release year

2021, 2022, 2023, 2024, 2025

FES_scenario

FES pathway name

Varies by FES release

solve_period

Time window to solve

Any date range

Available FES Years and Scenarios#

FES releases are updated annually by NESO. You can use different FES release years to:

  • Compare scenario projections across different releases

  • Access historical FES assumptions from earlier releases

  • Model the evolution of capacity expectations over time

Note: Different newer FES releases may have different scenario names and parameters. Check available data before running.

Step 3: Activate Your Scenario#

Edit config/config.yaml to add your scenario:

scenarios:
  - MyScenario_2040
  # - HT35  # Comment out others if not needed

Step 4: Validate Configuration#

Check your scenario is valid:

python scripts/validate_scenarios.py

This checks:

  • All required fields are present

  • FES data exists for the specified year

  • Weather cutouts are available

Step 5: Run Your Scenario#

# Dry run first to see what will execute
snakemake resources/network/MyScenario_2040_solved.nc -n -p

# Full run
snakemake resources/network/MyScenario_2040_solved.nc -j 4

Step 6: Analyze Results#

Load and analyze your results:

import pypsa
import pandas as pd
import matplotlib.pyplot as plt

# Load network
n = pypsa.Network("resources/network/MyScenario_2040_solved.nc")

# Generation dispatch over time
gen = n.generators_t.p.groupby(n.generators.carrier, axis=1).sum()
gen.plot.area(figsize=(12, 6), title="Generation Dispatch")
plt.ylabel("Power (MW)")
plt.tight_layout()
plt.savefig("my_scenario_dispatch.png")

# Curtailment analysis
renewable_carriers = ['wind_onshore', 'wind_offshore', 'solar_pv']
for carrier in renewable_carriers:
    gens = n.generators[n.generators.carrier == carrier]
    if len(gens) > 0:
        available = n.generators_t.p_max_pu[gens.index].sum(axis=1) * gens.p_nom.sum()
        dispatched = n.generators_t.p[gens.index].sum(axis=1)
        curtailment = (available - dispatched) / available * 100
        print(f"{carrier}: {curtailment.mean():.1f}% average curtailment")

Scenario Variations#

Historical Scenario#

For historical validation (uses real data, not FES):

Historical_2022:
  description: "Historical year 2022 validation"
  modelled_year: 2022
  renewables_year: 2022
  demand_year: 2022
  network_model: "ETYS"
  # No FES_scenario needed for historical years

High-Resolution Solve#

For detailed analysis of specific events:

StormAnalysis_2035:
  description: "Storm period analysis"
  modelled_year: 2035
  solve_period:
    enabled: true
    start: "2035-02-01 00:00"
    end: "2035-02-03 23:00"  # 3-day detailed analysis
  timestep_minutes: 30  # Half-hourly resolution

Clustered Network#

For faster solving with reasonable accuracy:

HT35_clustered:
  description: "Holistic Transition 2035 - clustered"
  modelled_year: 2035
  network_model: "ETYS"
  clustering:
    enabled: true
    n_clusters: 100  # Reduce to 100 buses

Comparing FES Releases#

Compare how projections change across FES releases:

HT35_FES2024:
  description: "HT scenario 2035 - FES 2024"
  modelled_year: 2035
  FES_year: 2024
  FES_scenario: "Holistic Transition"
  network_model: "Reduced"

HT35_FES2023:
  description: "HT scenario 2035 - FES 2023"
  modelled_year: 2035
  FES_year: 2023                    # Different FES release
  FES_scenario: "Holistic Transition"
  network_model: "Reduced"

Run both scenarios to compare results and understand how NESO’s capacity expectations evolved between releases.

Common Issues#

“Cutout not found”#

For years 2010-2024, cutouts are automatically downloaded from Zenodo when needed. If you get this error:

  1. Check if cutout workflow was run:

    ls resources/atlite/cutouts/

  2. Generate missing cutouts (will auto-download from Zenodo for 2010-2024):

    # Configure the year first
nano config/cutouts_config.yaml
# Set: years_to_generate: [2018]

# Run cutout workflow
snakemake -s Snakefile_cutouts --cores 1

  3. For years outside 2010-2024, you’ll need CDS API credentials (see Installation).

“Infeasible optimization”#

Check that generation capacity meets demand:

n = pypsa.Network("resources/network/MyScenario_unsolved.nc")
print(f"Generation: {n.generators.p_nom.sum()/1000:.1f} GW")
print(f"Peak demand: {n.loads_t.p_set.sum(axis=1).max()/1000:.1f} GW")

Next Steps#