The day-ahead market clears as if each bidding zone were a "copper plate" — infinite internal transmission capacity. Reality differs: internal grid constraints exist within zones. When the market result would cause overloaded lines, the TSO must intervene via redispatch — instructing generators to increase or decrease output to relieve congestion. This is a massive and growing cost (Germany alone: >€4 billion/year in recent years) and a central challenge in the energy transition.
Since October 2021, Germany's redispatch framework expanded significantly:
| Aspect | Before (Redispatch 1.0) | Redispatch 2.0 |
|---|---|---|
| Scope | Conventional plants ≥ 10 MW only | ALL generation ≥ 100 kW including renewables + storage |
| RES participation | Separate "Einspeisemanagement" (feed-in management) regime | Unified framework — RES treated same as conventional |
| Compensation (conventional) | Cost-based (variable costs + opportunity cost) | Cost-based; detailed cost verification |
| Compensation (RES) | 95% of lost market revenue under EEG | At market value of curtailed energy (reference: DA price or PPA price) |
| DSO involvement | Limited | DSOs identify local congestion; coordinate with TSO |
| Data exchange | Bilateral | Standardized via CONNECT+ platform; real-time availability schedules required |
Upward (increase): TSO pays generator's marginal cost for additional output. Southern gas plant starts: fuel + carbon + start-up cost.
Downward (decrease): TSO compensates Northern wind for curtailed output at market value (DA price they would have earned).
Net cost: Southern gas MC (high) − Northern wind curtailment compensation ≈ spread between constrained-on and constrained-off cost. Socialized via § 13 EnWG through grid tariffs to all consumers.
Congestion management design is the most contested question in European power market architecture:
| Zonal (current EU) | Nodal (US ISOs: PJM, CAISO, ERCOT) | Smaller Zones | |
|---|---|---|---|
| Model | Copper plate within zone; congestion managed ex-post via redispatch | Every node gets its own LMP; congestion priced ex-ante | Split large zones at structural bottlenecks (e.g., DE-North / DE-South) |
| Price signal | Uniform within zone — no locational signal | Full locational signal — every node different | Coarse locational signal at zone boundaries |
| Investment signal | Build anywhere in zone — no price penalty for congested location | Build where prices are high → naturally relieves congestion | Some locational differentiation; better than single zone |
| Redispatch cost | High — socialized (no incentive to locate optimally) | Zero — congestion incorporated in market clearing | Reduced — less internal congestion within smaller zones |
| Liquidity | High (large zone = large market) | Fragmented at each node (but financial hubs aggregate) | Moderate — depends on zone size |
| Political feasibility | Status quo | Massive overhaul; strong resistance from incumbents | Contested but technically feasible (cf. SE split in 2011) |
In US ISOs, the market clearing itself handles congestion:
Energy: System marginal cost (same everywhere if uncongested).
Congestion: Shadow price of binding transmission constraints — positive at import-constrained nodes (load pockets), negative at export-constrained nodes (generation pockets).
Losses: Marginal loss factor at the node (typically small, 1-5%).
Result: No redispatch needed — the dispatch IS optimal given constraints. Generators in congested-export areas (like North Germany would be) face lower prices, creating incentive to relocate or for grid to be built. The congestion rent (price difference × flow) funds transmission rights (FTRs/CRRs).
Between the extremes of "ignore congestion" (current zonal) and "full nodal" (radical reform), several intermediate approaches are emerging:
| Approach | Mechanism | Examples |
|---|---|---|
| Local flex markets | DSO/TSO procures flexibility from local assets via competitive platform for congestion relief | GOPACS (NL), Enera (DE pilot), Piclo Flex (UK), NODES |
| Dynamic grid tariffs | Time-and-location varying grid fees incentivize consumption/generation to match grid constraints | Emerging in NL, DE (§14a EnWG for flexible loads) |
| Redispatch market-based | Instead of cost-based instruction, TSO runs a local auction for upward/downward adjustment | Proposed in some reform discussions; risks strategic behaviour |
| Non-firm connection agreements | New generators accept curtailment risk in exchange for faster/cheaper grid connection | UK "connect and manage"; Dutch congestion areas |
Dutch TSO (TenneT) and DSOs operate a platform where flexibility providers can offer capacity for congestion management. Market-based: providers bid price/volume; TSO/DSO selects lowest-cost solution. Integrates with ETPA (Energy Trading Platform Amsterdam) for matching. Represents the most advanced operational local flex market in Europe.
| For… | Congestion/Redispatch Impact |
|---|---|
| Generator in export-constrained area | Risk of curtailment (downward redispatch). Under Redispatch 2.0: compensated at market value, but operational disruption + uncertainty. In a nodal world: permanently lower prices → lower asset value. Mitigation: diversify location; hybrid storage to time-shift output. |
| Generator in import-constrained area | Opportunity: frequently constrained-on (upward redispatch). Additional revenue from redispatch activation on top of market schedule. In a nodal world: permanently higher prices → higher asset value. Risk: regulatory change (zone split → lose premium). |
| Flexible asset (battery/DR) | Can provide BOTH upward and downward redispatch. Local flex markets create new revenue stream. Co-optimization: reserve + arbitrage + redispatch. Behind-the-meter battery can defer grid reinforcement → network value beyond wholesale. |
| RES developer (siting decisions) | Under zonal: no locational price signal → build where wind/solar resource is best (often = congested). Under nodal or split zones: lower capture price in congested areas → shifts optimal siting toward load centers. Grid connection queue and curtailment risk already de-facto creates locational signal. |
| Portfolio optimizer | Must model curtailment probability per asset location. Expected revenue = market price × (1 − P(curtailment)) + redispatch compensation × P(curtailment). Curtailment correlated with high-wind events (when market prices are low) → impact is modest in €/MWh but significant in volume. |
| Year | Redispatch Volume (GWh) | Redispatch Cost (€ billion) | Key Driver |
|---|---|---|---|
| 2017 | ~20,000 | ~1.4 | Growing wind in North; grid expansion delayed |
| 2019 | ~25,000 | ~1.2 | Lower gas prices reduced unit cost |
| 2021 | ~16,000 | ~2.3 | Higher commodity prices inflated compensation |
| 2022 | ~12,000 | ~4.2 | Energy crisis; extreme commodity prices |
| 2023 | ~15,000 | ~3.1 | Stabilizing; new interconnectors coming online |
For the zonal vs. nodal debate in European context:
Neuhoff, K. et al. (2013). "Renewable electric energy integration: Quantifying the value of design of markets for international transmission capacity." Energy Economics, 40, 760–772.
For the German redispatch framework:
BNetzA (Bundesnetzagentur), Monitoring Report (annual) — contains detailed redispatch statistics, costs, and volumes. Available at: bundesnetzagentur.de
For local flexibility markets:
Schittekatte, T. & Meeus, L. (2020). "Flexibility markets: Q&A with project pioneers." Utilities Policy, 63, 101017.
Why does the current zonal market design lead to redispatch costs rather than preventing congestion through pricing?
In a nodal pricing system (LMP), the congestion component at a node is:
Why are German redispatch costs growing despite relatively stable congestion volumes?
What distinguishes redispatch from balancing activation?
Rechtliches: Impressum