Capacity Expansion Model

How we build future supply stacks for European markets

Future build of the generation stack across Europe is determined by Modo Energy’s market knowledge in the short term, and a techno-economic capacity expansion model in the longer term.

The short term means to circa 2030 - though this can vary for different technologies and regions, depending on the available information. Generally the first three years are fixed, based on Modo Energy’s research, which covers project pipelines, policy, press releases, market conversations, and detailed analysis.

The mid- to longer-term forecast is determined by our in-house capacity expansion model, reflecting the project economics of various power generators in our future modelled world for all 15 regions of our European forecast.

Evolution of power generation across Europe

The interactive map below shows how capacity additions grow over time across the continent. Regional differences in resource quality, demand growth patterns, interconnection constraints, and policy frameworks lead to different generation types dominating in different countries.

Use the selectors to explore capacity buildout by technology and year.

In every year we model, investment decisions for new power generation combine with an hourly dispatch to reflect new and existing assets’ operation. If a generator’s existence will reduce the overall cost of running the system, it will be built. Build decisions are made given the same information as the full market dispatch (generator technicals, fuel & carbon, renewable profiles, outages, transmission limits are common to both). Thus investment decisions are made for the same world as the market dispatch model actually sees, rather than a simplified proxy.

More information on our underlying capacity expansion model is here.

At a glance

   
Geography 18 countries: Great Britain, Germany, France, Spain, Portugal, Italy (as seven bidding zones), Belgium, Netherlands, Austria, Switzerland, Poland, Norway, Sweden, Denmark, Ireland, the Czech Republic, Slovenia and Montenegro
Horizon Annual decisions to 2060
Temporal resolution Hourly dispatch on 100 representative days per model year (50 two-day chunks)
Windowing Rolling one‑year optimisation (co-optimised investment + dispatch → state carry‑over)
Reliability Country‑level energy unserved limits and planning reserve margins with de-ratings for each technology
Outputs Annual capacity additions/retirements by zone & technology, country reliability metrics, and dispatch time series for sampled days

Building a future generation stack

Technologies governed by economics are included in the model:

  • Gas OCGT
  • Gas CCGT
  • CCS Gas CCGT (retrofits)
  • Solar PV
  • Onshore wind
  • Offshore wind
  • Hydrogen peakers (H₂)
  • Battery energy storage (BESS): 2h, 4h, 6h, 8h

Note: Storage is modelled with power/energy limits, round‑trip efficiency, and state‑of‑charge tracking. Retrofit options (e.g., CCGT → CCS) are constrained by age, size, and retrofit-ability rules where applicable.

Where the expansion model does not include a technology as a build option (e.g. nuclear, large hydro), we look to official TSO/DSO datasets and national development plans to guide sense-checks on installed capacity levels. A full list of these inputs is shown at the bottom of this page.

Modelling representative time periods

The capacity expansion model runs at hourly resolution on 100 representative days per year (50 two-day chunks). Generator revenues are calculated and profits (or costs) determined. Days are selected to capture various conditions across each year (peak demand, low‑renewables, high‑renewables, shoulder periods, etc.). This means investment choices are robust to the operating realities the system faces, while maintaining a reasonable run-time.

Each year’s build and retire decisions shape the starting point for the next year. This keeps track of things like storage capacity and retrofit commitments as they evolve.

Ensuring a reliable power system

We use several reliability constraints to ensure sufficient generation is built to meet demand.

Energy unserved limits

This refers to the maximum amount of electricity demand that can go unmet before it becomes unacceptable to the system operator or regulator. In practice, no grid runs with zero risk of shortages - because building to meet every possible peak demand would be prohibitively expensive. Instead, markets or regulators define an acceptable “energy unserved” threshold. Energy unserved limits are applied to every country in the model.

Planning reserve margin

This is a reliability standard that defines how much “extra” generation capacity the system must have above expected peak demand. For example, if peak demand is forecast at 100 GW, and the reserve margin is set at 15%, the system needs 115 GW of available capacity (de-rated capacity).

  • De-rating factors (or capacity credits) are applied to the nameplate capacities of different generators to reflect that not all capacity is firm or available in every hour. The de-rating factors vary by technology
  • The planning reserve margins are framed to be conceptually consistent with capacity market mechanisms used in GB, Poland, Ireland, and France
  • Margins are enforced annually across all countries in scope
  • Great Britain: 10%. Other countries in the model use 15%. The GB margin matches the de-rated margin in NESO’s Winter Outlook 2025/26: 6.1 GW of spare capacity after adjusting each plant for expected availability, or 10% of average cold-spell peak demand.

De-rating factor sources

De-rating factors come from published capacity market rules where a country has them, and from national adequacy studies where it does not.

Country Source Type
GB NESO capacity market de-rating factors, 2029/30 Capacity market
Belgium Elia CRM calibration report, 2030/31 Capacity market
France RTE capacity mechanism parameters, 2026/27 Capacity market
Ireland SEM-O T-4 capacity auction pack, 2030/31 Capacity market
Italy Terna capacity market factors, 2027 delivery Capacity market
Poland Capacity market regulation, 2029 auction Capacity market
Norway NVE power-system report, 2024 Adequacy study
Portugal ERSE reliability-standard study Adequacy study
Spain CNMC capacity-mechanism assessment Adequacy study
Sweden Svenska kraftnÀt winter power-balance report, 2026 Adequacy study
Netherlands Uses Belgium’s factors Country proxy
All other countries Equal-country average of the capacity-market factors above (Belgium, France, GB, Italy and Poland). Italy counts once Cross-country proxy

Where a source has no value for a model technology, we use the closest published class. If no class fits, we use the cross-country average. Interconnectors inside Italy and between Spain and Portugal get zero capacity credit.

Great Britain long-term build follows system need

GB capacity with a Contract for Difference (CfD) stays on its published trajectory, including delivery delays. Beyond that contracted capacity, the capacity expansion model decides how much to build.

Subsidy support for new solar, onshore wind, offshore wind and hydrogen peakers is assumed to continue while the system needs the capacity. The model builds only where that capacity is needed.

Build limits keep each year’s wind and solar decision inside a range. The build can be anywhere from 20% to 100% of that year’s addition in the Modo Energy GB buildout tracker. This applies from 2031 for solar, 2032 for onshore wind and 2034 for offshore wind.

Technology Cost Projections

Our capacity expansion model uses capital and operational cost projections based on industry-standard sources including NREL’s Annual Technology Baseline, IRENA’s renewable cost data (for solar), and Modo Energy’s BESS CapEx Survey.

Example Capacity Buildout

The model produces annual capacity expansion pathways for each technology across all modeled countries.

Data sources

GB‑specific

  • FES 2023 (Future Energy Scenarios) - Scenario pathways for GB capacity to 2050 (National Grid ESO)
  • Summer Outlook 2025 - Near‑term operational margins and available capacity
  • SOR25 (System Operability Report 2025) - System needs (inertia, voltage, frequency response) under the projected 2025 mix

EU‑wide

  • ENTSO‑E TYNDP scenarios - Pan‑EU generation, demand, and grid development outlooks to 2050
  • ERAA 2024 - Probabilistic adequacy framing and cross‑border resource sufficiency

Country plans

  • Germany – NEP 2037/2045 - Grid development and capacity scenarios
  • France – PPE3 (Programmation pluriannuelle de l’énergie) - Nuclear life extensions and the EPR2 programme, plus solar and offshore wind targets
  • Luxembourg – Creos scenarios - National capacity and interconnection outlooks
  • Spain – ESIOS 2025 update - Updated capacity/demand projections
  • Norway – NVE publications - Hydropower expansion and market‑integration assumptions

Medium‑term adequacy

  • ENTSO‑E Mid‑term Adequacy (now within ERAA) - Ten‑year adequacy framing for stress testing

Financial data sources

Category Key input sources
CapEx IRENA 2024 (solar), NREL Annual Technology Baseline (ATB) (other technologies)
OpEx IRENA 2024 (solar), NREL Annual Technology Baseline (ATB) (other technologies)
BESS CapEx Modo Energy CapEx Survey, informing Modo Energy central view
Asset Lifetimes 25 years, 20 years for BESS
WACC 5.00% real (~7.10% nominal)
De-rating factors Country capacity-mechanism and adequacy studies (Elia, Terna, PSE, SEM-O, RTE, SvK, CNMC, ERSE)
✏ Suggest an edit on GitHub