Konzept
Modellierung Strommärkte
Our power modeling approach…
… combines heuristic optimization and player behavior into sophisticated market models describing realistic short and long-term market dynamics
Key Drivers
- Fuel / CO2 prices
- Existing power plants
- Cost & efficiency of new power plants
- New technologies (e.g. electric mobility)
- Interconnectors
- Load growth, small CHP, renewables
- Future of nuclear
European long-term Power Market model (EltPM)
- Mittel- / langfristige Preisprognosen
> Strom: Europa
> Gas: Global
- Spread Vorhersagen
> commodity spreads
> saisonale spreads
> regionale spreads
- Anlagenspezifische Auswertungen
- Identifikation von Haupttreibern und Interdependenzen
- Übersetzung von “Key Beliefs” in quantifizierte Szenarios
Output
- Mittel- / langfristige Preisprognosen
> Strom: Europa
> Gas: Global
- Spread Vorhersagen
> commodity spreads
> saisonale spreads
> regionale spreads
- Anlagenspezifische Auswertungen
- Identifikation von Haupttreibern und Interdependenzen
- Übersetzung von “Key Beliefs” in quantifizierte Szenarios
A closer look at our power modeling approach…
Basic Model Design
- Geography: EU – plus CH, NO, Balkans, Turkey but (currently) excl. IRE, CY, MT
- Time horizon: 2015 – 2035(+) – 8760 hrs/yr
- Price Rationale: marginal cost of last power plant online + risk premium (RP) ONLY in tight situations. market design sets RP range
- Market Clearing: heuristic (modeling single trades until transmission used)
- New Build: decision based on economic evaluation
- Competition: hurdle rates, min. risk premium timing and volume of new build
Input Data
- Deterministic(scenarios-based): list of power plants (age, type, efficiency) ,fuel prices, interconnector, capacities, …
- Stochastic: plant outages & overhaul probability
- Cases: historic weather years (load shape, wind & PV feed-in shapes)
- Data: All data (in-/output) stored in fast ORACLE database
- Result GUI: Easy/fast graphical representation of results within predefined tools
- Performance: 25 years (long-term run) on single multicore computer approx. 6 hrs
Model IT
- Text
> Strom: Europa
> Gas: Global
- Lorem
> text
> text
> text
- Ipsum
- text
- text
Unique features of our power modeling approach…
… modelling capabilities:
Coverage
- Geography: All regions covered fully synchronous in each model run (no “single region” approach with fixed import/export flows or “simplified neighboring regions)
- Time horizon: hist all single years 2015 – 2035(+) – 8760 hrs/yr chronologically
Data Quality
- Permanent update of mothballing, retirement and new grid connection for power plant
Outstanding Risk Assessment
- 31 consistent weather yrs. (hourly wind-, PV-, demand-profile) for all EU countries
- … can be combined with monte-carlo-based plant outages
- Contribution margins based on hourly data for all 2000 power plants and for all “cases” (=combination of weather year and heuristic plant outages)
- User-defined scenarios for political framework (e.g. carbon, nuclear, sector coupling, …)
