Congestion Management

The set of actions taken by system operators (TSOs and DSOs) when the physical grid cannot accommodate all requested power flows. Congestion occurs when demand for transmission or distribution capacity exceeds what the network can safely deliver. Managing it is a core driver of Flexibility needs.

Why congestion occurs

The Electric Power Transmission and Electric Power Distribution networks have finite capacity. Congestion arises when:

• Demand growth outpaces grid reinforcement (electrification of transport, heating, industry)
• Generation patterns shift — renewable generation in locations far from load centers (e.g., wind in northern Sweden, consumption in the south)
• Variable renewables create volatile, weather-dependent flow patterns
• Cross-border flows add to domestic congestion (Flow-Based Capacity Calculation determines available capacity)

In Sweden, the structural north-south imbalance across the Bidding Areas (SE1–SE4) is the most prominent congestion pattern — surplus hydro and wind generation in the north, growing consumption in the south. Svenska kraftnät‘s NordSyd initiative targets this bottleneck with SEK 225 billion in grid investment. (Source - Svk Network Development Plan 2026-2035)

Congestion at different grid levels

Transmission (TSO)

At the transmission level, congestion management has well-established mechanisms:

• Bidding zones — structural congestion is reflected in price differences between zones (e.g., SE1–SE4). Market participants see different prices, which implicitly steers production and consumption. (Regulation Art. 14)
• Redispatching — the TSO orders generators or loads to adjust output/consumption to relieve specific constraints. The Clean Energy Package requires redispatching to be market-based (Regulation Art. 13), with non-market-based redispatching only as an exception. The SO GL explicitly lists redispatching of transmission or distribution-connected system users as a remedial action category (Art. 22(1)(e)).
• Countertrading — cross-border variant of redispatching (SO GL Art. 22(1)(f))
• Curtailment — last resort reduction of generation or load
• Cross-border cost-sharing — when redispatching or countertrading is ordered for cross-border relevance, costs are shared among TSOs under regional methodologies approved by national NRAs. In CCR Hansa (the Baltic Sea region), the core principle is that the TSO in whose control area the physical congestion occurred bears the cost — creating a direct financial incentive for TSOs to manage their own domestic constraints. HVDC interconnector faults/limits are split between cable owners by Annex 1 sharing keys; Baltic Cable AB uniquely bears 100% of costs for the SE4–DE/LU border. (Source - CCR Hansa RCCS Methodology and Ei Approval (2024))

Distribution (DSO)

At the distribution level, congestion management is newer and less mature:

• Flexibility Markets — DSOs procure flexibility from local resources to resolve congestion. Art. 32 of the Electricity Market Directive requires market-based procedures.
• Villkorade Avtal — conditional connection agreements where the DSO can curtail customers during congestion events. A rules-based backstop to market-based procurement.
• Network tariff design — time-of-use or capacity-based tariffs that incentivize consumption outside peak periods (implicit Demand Response)
• Grid reinforcement — the traditional but slow solution: building more capacity

DSO-level congestion is growing rapidly as distributed energy resources (solar, EVs, heat pumps, batteries) create bidirectional flows and local capacity constraints that didn’t exist in the traditional passive distribution grid. (Source - Electric power distribution (Wikipedia))

EU regulatory framework

The Clean Energy Package establishes clear principles:

1. Market-based is the default — both redispatching (Regulation Art. 13) and DSO flexibility procurement (Directive Art. 32) must use market-based procedures
2. Non-market measures require justification — rules-based approaches are only permitted where market-based procurement is “not economically efficient” or would cause “severe market distortions”
3. DSOs as neutral facilitators — DSOs cannot own storage (Directive Art. 36) or favor their own resources; they must procure flexibility on a level playing field
4. Transparency — DSOs must publish network development plans identifying congestion areas and flexibility needs (Directive Art. 32(3))

The Network Code on Demand Response (NC DR) adds an operational layer: detailed rules for local services markets, TSO-DSO coordination on shared congestion, observability areas, and congestion forecasting across multiple time horizons. (Source - NC DR Proposal (ENTSO-E and EU DSO Entity, 2024), Source - ACER Recommendation 01-2025 on NC DR)

The 70% minimum capacity rule (Regulation 2019/943, Art. 16(8)) — which requires at least 70% of each CNEC’s physical capacity limit to be available for cross-zonal trade — has historically been interpreted as a floor: TSOs were expected to use costly remedial actions (redispatch, countertrading) to provide more than 70% where economically efficient. The EU General Court ruling in BNetzA and Germany vs ACER (October 2025) changed this: TSOs providing 70% satisfy all regulatory obligations and have no legal duty to go further. The Nordic TSOs incorporated this reinterpretation into the Nordic CCM Third Amendment (submitted May 2026): costly remedial actions are now limited to meeting the 70% minimum, managing temporary outages, and enabling TATL in N-1 scenarios — virtual capacity above the physical limit is abandoned. This means the 70% rule now functions as a ceiling of TSO cross-zonal capacity obligations in the Nordic CCR, constraining available capacity in congested periods. (Source - Nordic CCM Third Amendment Package (2026), Flow-Based Capacity Calculation › The 70% rule: floor or ceiling?)

The European Grids Package (COM/2025/1005, 2025) adds two further EU-level instruments relevant to congestion management:

• Cross-border cost allocation reform in COM(2025) 1006 — enables use of congestion income (flaskhalsinkomster) to finance cross-border infrastructure; Sweden objects to this restriction on national flexibility in using these revenues (see Svenska kraftnät › HVDC interconnectors — three projects paused (May 2026))
• 2-way CfD guidance (C/2025/8479) — non-binding guidance for member states on designing contracts for difference between generation operators and a public counterpart; draws on Art. 19d Reg 2019/943 (as amended by the Electricity Market Design Reform 2024) which from 2027 requires public interventions for new generation to use 2-way CfDs. Relevant to Sweden’s debate on capacity mechanisms for dispatchable fossil-free production (see Elmarknadsutredningen and Svenska kraftnät › Strategy 2030)

Congestion management vs grid expansion

A key strategic question: when should congestion be managed through flexibility, and when should the grid simply be expanded?

• Grid expansion is permanent, expensive, and slow (5–15 years for major transmission projects)
• Flexibility is faster to deploy but has operating costs and requires digital infrastructure
• In practice, both are needed: flexibility bridges the gap while grid expansion proceeds, and in some cases permanently defers investment where congestion is infrequent

Svenska kraftnät‘s experience illustrates this: the connection queue exceeds 175 GW against ~25 GW peak load. Even with the massive NordSyd investment, flexibility is essential to manage interim congestion and may permanently reduce the need for some reinforcement. (Source - Svk Network Development Plan 2026-2035)

Phantom congestion: luftbokning and the faktisk belastning doctrine

A structural source of apparent congestion in Swedish distribution grids is luftbokning (air booking): DSOs planning grid capacity from customers’ contracted power (abonnerad effekt) rather than their actual physical load (faktisk belastning). A customer who contracts 100 kW but consistently draws only 60 kW locks up 40 kW as a phantom reservation. At aggregate level, a distribution grid may appear congested on paper while being physically underloaded — causing new connections to be refused or delayed when the physical capacity is already available.

Ei‘s government assignment report (R2024:14) reaffirms the faktisk belastning doctrine: DSOs must calculate available capacity using actual physical load with diversity effects, not the sum of contracted capacities. Using AMI data to measure actual load — and applying statistical diversity factors to reflect that customers never simultaneously reach their contracted maximums — resolves phantom congestion without requiring grid reinforcement.

Conclusion: Ei R2024:14 found no new regulation is needed for capacity redistribution. Existing tools (faktisk belastning methodology, forthcoming Article 6a FCA framework, TOTEX reform) are sufficient once correctly applied. The primary action is supervisory: Ei planned a tillsyn review to check which DSOs are still using contracted power rather than actual load in their capacity calculations.

See also Distribution System Operator › Capacity calculation doctrine: faktisk belastning.

Swedish redispatching framework

Ei‘s analysis (Source - Ei Villkorade avtal (2023)) establishes the regulatory hierarchy for distribution-level congestion management in Sweden:

1. Kostnadsreflektiva tariffer (cost-reflective tariffs) — first line of defense for predictable congestion; EIFS 2022:1 had required capacity tariff reform by January 2027, but is being repealed by June 2026; permissible under Art. 18 EU reg; new mandatory model due April 2027 (Source - Ei Effektavgifter webb (2026)). Ei’s ställningstagande Ei2025:06 (Source - Ei Ställningstagande Tariffer Ei2025-06) clarifies the aggregate-load principle: the time-differentiation of the effektavgift must reflect the sammanlagda belastningen (aggregate load) on the DSO’s entire grid — not only the individual customer’s load curve. Setting high-price periods solely from each customer’s own peak hours, without checking whether those hours coincide with high total grid load, is explicitly non-compliant. A four-component structure applies: energiavgift (losses, per kWh), effektavgift (forward-looking capacity signal, time-differentiated), kundspecifik avgift (metering/admin), and fast avgift (residual fixed recovery).
2. Market-based redispatching (marknadsbaserad omdirigering) — competitive procurement on Flexibility Markets; required as default by Art. 13 of the Regulation
3. Non-market-based redispatching — including Villkorade Avtal; only permitted when Art. 13(3) exceptions are met (no market available, resources exhausted, too few providers, or congestion too predictable for market solutions)

DSOs must report their redispatching activities annually to Ei, including volumes, prices, and procurement method — creating a transparency mechanism for tracking the transition from non-market to market-based approaches.

Multi-level congestion: which DSO bears the obligation?

Ei‘s ställningstagande Ei2025:01 (Source - Ei Ställningstagande Ei2025-01 Villkorade avtal (2025)) clarifies a common ambiguity in multi-level grid congestion:

The DSO in whose own network the physical constraint exists bears the obligation to resolve it. This is true at all grid levels — from local DSO to regional DSO to TSO.

Practical consequence: when a local DSO’s connection queue is blocked because the overlying regional grid lacks capacity, it is the regional DSO (or TSO if the constraint is at transmission level) that must act — not the local DSO. The local DSO cannot sign Villkorade Avtal to work around a constraint it doesn’t control and isn’t responsible for resolving.

What the local DSO can do:

• Request raised subscription/connection capacity from the overlying operator
• If unreasonably delayed: report to Ei for review (4 kap. 13 § ellagen)
• Procure flexible resources in its own network
• Design tariffs that relieve the connection point to the overlying grid

What the overlying DSO can do: contract flexibility directly with resources in the underlying network, with the underlying DSO administering procurement and activation — but the financial responsibility stays with the overlying operator (since flexibility service costs are generally not customer-specific and spread across the overlying DSO’s customer collective).

A key practical driver of Swedish DSO congestion management has been the subscription to the overlying grid (abonnemang mot överliggande nät): regional DSOs have fixed power subscriptions from Svenska kraftnät, and exceeding them during peaks creates the primary motivation for local flexibility procurement. Svk’s temporary subscription increases sometimes act as a price cap on flexibility, suppressing willingness to pay on local markets. (Source - Ei Flexibility in Distribution Grids (2023))

The subscription mechanism as market driver

The CoordiNet demonstration provides concrete evidence of how the subscription mechanism shapes market outcomes. (Source - CoordiNet D4.7.2 Swedish Demonstration (2022))

The subscription level (abonnemang mot överliggande nät) is the annually contracted maximum power a regional DSO may draw from the TSO grid without prior notice. It differs from physical capacity — the physical components allow higher flows, but the subscription defines the operational limit. The TSO also grants temporary subscriptions (up to 7 days ahead, retractable at any time) for short-term needs.

This creates two distinct market equilibria:

1. Low-price, high-volume equilibrium (Uppland model): When the TSO frequently denies subscription raises, the DSO must use the market regularly. Flexibility priced below the temporary subscription fee (~240–280 SEK/MWh) is attractive → DSO buys routinely → stable market with meaningful volumes (9,965 MWh over three winters, avg 248 SEK/MWh in Uppland).

2. High-price, low-volume equilibrium (Skåne model): When the TSO routinely grants temporary subscriptions, the DSO only buys flexibility when subscription is denied (penalty ~2,800 SEK/MWh for exceedance without permission) → small volumes, high prices (206 MWh over three winters, avg 2,285 SEK/MWh in Skåne). The market tests processes but doesn’t regularly manage real congestion.

This suggests that the TSO’s subscription policy (how readily it grants temporary subscription increases) is a more important determinant of DSO flexibility market liquidity than market design choices.

GOPACS (Netherlands) — common market model benchmark

The DSO Entity and EC LFM study both cite GOPACS (Grid Operator Platform Alleviating Congestion on the Subnet) as the most advanced TSO-DSO coordination model for congestion management in Europe. Technical specifications (from DSO Entity report, 2026):

GOPACS scores highest among all 37 EU LFM initiatives in the EC LFM study (VITO, 2025) on value stacking and market efficiency metrics. The 150% overbooking ensures that the market has sufficient volume even when some bids are not delivered, while the shared order book means resources are always allocated to the SO with the greatest need.

SE4 vs SE3 congestion price spread

Sweco’s 2025 market survey provides the clearest evidence of the SE4/SE3 structural congestion difference (Source - Sweco Kartläggning av lokala flexibilitetsmarknader (Ei, 2025)):

SE4 prices are 2–60× higher than SE3, reflecting the structural deficit in the southern grid: high demand, constrained transmission links (historically), and limited local generation. The October 2024 Söderåsen-Barsebäck 400 kV upgrade (+600 MW, equivalent to Malmö’s entire cold-day demand) is expected to reduce this premium over time.

Grid expansion as congestion resolution — the market closure dynamic

A key insight from Sweco’s mapping: local flexibility markets are often triggered by temporary grid bottlenecks. When the underlying grid constraint is resolved through reinforcement, the market rationale disappears and the market closes. This is not a market failure — it is the intended dynamic where flexibility bridges the gap until investment catches up.

Södra Skåne example: E.ON’s Skåne markets addressed congestion on the Söderåsen-Barsebäck-Sege transmission corridor. October 2024: the 400 kV Söderåsen-Barsebäck link came online, the third sequential upgrade on the west coast. Combined with the earlier Hurva-Sege (2021) and Barsebäck-Sege (2023) upgrades, +600 MW of capacity was added. This made the flex market “much less necessary or even unnecessary” for the primary bottleneck — though markets remain running for other sub-area constraints. (Source - Sweco Kartläggning av lokala flexibilitetsmarknader (Ei, 2025))

The “chicken race” between grid levels

A structural barrier identified through interviews: when a capacity constraint spans multiple grid levels (e.g., regional DSO constraint affects local DSO connection queue), each level waits for the other to procure flex first — to avoid bearing the cost. Neither level has a legal obligation to initiate. This creates inaction and delays market development. Sweco recommends Ei develop a methodology to attribute responsibility across grid levels (Rec 3 and Rec 4). (Source - Sweco Kartläggning av lokala flexibilitetsmarknader (Ei, 2025))

Quantified economic value of DSO flexibility procurement

FlexAbility (2025) provides three case studies from Ellevio’s network that quantify the economic value of different flexibility procurement use cases for DSOs. These are among the first publicly available concrete cost comparisons in Sweden. (Source - FlexAbility Delrapport 3 (2025))

Utnyttjandegrad — regulatory revenue incentive

EIFS 2023:6 introduced utnyttjandegrad (utilization rate) as a metric in the Swedish revenue cap regulation. It measures how flat a DSO’s demand profile is at its grid connection point:

utnyttjandegrad = daily average demand / average of the four highest daily peaks

A low score triggers a revenue penalty. Station A case: normvärde 44.02%, actual 2023 value 38.88% → penalty −924,000 SEK/year. A 20 MWh flexibility intervention targeted at peak hours has an average value of 10,711 SEK/MWh (range 0–40,000 SEK/MWh depending on timing — only hours that affect the four highest peaks have any value). This makes utnyttjandegrad one of the highest-value but most timing-sensitive DSO flexibility use cases.

Abonnemangsoptimering — subscription reduction

Swedish regional DSOs pay Svk 195,000 SEK/MW/year for their power subscription. The marginal value of flexibility used to cut peak demand and reduce the subscription level is extremely steep at small reductions and collapses rapidly:

At a flexibility procurement cost of 10,000 SEK/MWh, optimal subscription reductions were: Station A 55→51 MW (−4 MW, saving ~712,000 SEK/year), Station B 59→57 MW (−2 MW, saving ~482,000 SEK/year). The report cautions that local savings may not represent system-level benefits — the question of whether abonnemangsoptimering reduces costs for the system as a whole or merely redistributes them is an open research question.

Alternativkostnad — grid investment deferral (bus depot case)

Station C: two 20 MVA transformers, N-1 capacity 28 MVA. A bus depot adding 8.7 MW planned for 2028. Traditional solution: upgrade to 2×25 MVA ≈ 22 MSEK.

Hourly demand forecasting showed that N-1 is exceeded only 4 hours per year after the bus depot connects — totaling 2.6 MWh. Flexibility cost at 10,000 SEK/MWh: 26,000 SEK/year. Present value over 45 years at WACC 4.53%: ~1.2 MSEK vs 22 MSEK for grid investment. Despite this overwhelming economic case (17:1 ratio), the CAPEX bias in current regulation still favors the investment (see CAPEX bias).

Generalized transformer upgrade cost

Based on Ei’s normprislista (standard price list) across 29 transformer configurations:

Rule of thumb: 1 MW of grid upgrade capacity ≈ 100–200 hours of flexibility at 5,000–10,000 SEK/MWh. This gives DSOs a practical first-order test for whether flexibility is likely to be cost-effective before committing to detailed analysis.

Nodal pricing as a theoretical benchmark — SOU 2025:47 analysis

The Elmarknadsutredningen conducted a detailed analysis of nodal pricing as an alternative to the current zone-based price system. The conclusion is that nodal pricing should not be implemented in Sweden but is useful as a theoretical reference standard.

The theoretical case: Nodal pricing (locational marginal pricing, used in parts of the US) assigns each grid node its own electricity price based on the balance between supply and demand plus congestion at that node. This provides the most precise possible signal for where to locate flexible resources, batteries, and new generation.

Why the committee rejected it:

• Liquidity: many small markets, each with few participants → low competition, higher prices, potential market power abuse
• Financial hedging difficulty: no liquid forward market at node level → increases investment risk for generators and industrial customers
• Investment uncertainty: congestion is dynamic; a node-specific price can change as grid topology evolves, making long-term investment calculations unstable
• Complexity: requires advanced real-time calculation systems; difficult for market participants to navigate

Recommended use: The Flow-Based Capacity Calculation method active since October 2024 in the Nordics can partially serve as a nodal pricing benchmark — it incorporates grid topology data down to 70 kV and allows modeling of the value of reducing specific network constraints. This makes it useful for evaluating whether current price signals in a given area are appropriately reflecting congestion costs.

Redispatching volumes in Sweden — 2022–2024 trend

Ei publishes an annual Art. 13.4 EMR redispatching report. The most recent is Source - Ei R2025-13 Omdirigering i Sverige 2024 (October 2025), covering 2024 data. Historical data comes from Source - SOU 2025-47 Elmarknadsutredningen (2025).

2024 breakdown — who redispatched and why:

• Svenska kraftnät: 59,334 MWh — primarily värmekraft (81%) and hydro (16%); avoid overload in own transmission network
• Ellevio: 220,150 MWh — hydro (91%), wind (7%), demand response (2%); caused by ongoing works at the regional grid level requiring sustained redispatching. Ellevio alone accounts for 99% of all DSO redispatching and 78% of the national total. This is a structural event linked to a specific investment programme, not an ongoing market mechanism.
• Umeå Energi Elnät: 988 MWh — other resources; avoid overload
• Göteborg Energi Nät (GENAB): 403 MWh — demand response; explicitly testing Effekthandel Väst local flex market and managing capacity peaks pending subscription increases from overlying network reinforcements
• Mölndal Energi Nät: 168 MWh — demand response; avoid overload in own and overlying network
• E.ON Energidistribution: 193 MWh — demand response + other + heat + storage; avoid overload and overlying network constraints
• Götene elförening: 4 MWh — demand response; avoid exceeding subscription limit + pilot testing

Market-based vs non-market: 4 of 6 DSOs described their redispatching as market-based. 2 DSOs used Villkorade Avtal — non-market-based redispatching — the report does not name which two.

Demand response growth in redispatching:

A 48× year-on-year increase in 2024, though still trivial (1.8% of total redispatching) compared to hydro and wind. The growth is consistent and traceable to expanding flex market activity and villkorade avtal portfolios becoming operational.

The low uptake paradox: Despite widespread reported capacity constraints across Sweden’s ~170 DSOs, only 6 reported using redispatching in 2024. Ei explicitly flags this gap — volumes are “fortfarande låg” (still low) and “de företag som omdirigerar är få” (few companies redispatch) despite stated capacity shortages. Most DSOs appear to be managing capacity through villkorade avtal that were not activated, connection refusals, or queue delays — none of which appear in the Art. 13 report. Ei calls on all grid companies to investigate whether they could use redispatching as a grid efficiency tool.

“Increase-decrease game” risk: When redispatching is used repeatedly and predictably at the same location, market participants can game it. An actor sitting in a congested area can bid high for upward redispatch and low for downward redispatch — earning systematic profit from the congestion. This is the “increase-decrease game” documented in international literature (Holmberg, 2024). Norway encountered this problem in the late 1990s and introduced cost-based compensation rights in 2002. Germany uses non-market-based redispatch systematically for this reason. Art. 13.3 EMR explicitly permits non-market-based redispatch with cost-based compensation in situations of insufficient competition or highly predictable congestion.

Kapacitetszoner — upstream congestion prevention at TSO level

Svenska kraftnät‘s April 2026 government assignment report introduces a prospective upstream tool for preventing transmission congestion: kapacitetszoner (capacity zones) within the proposed Anvisningssystem. Rather than managing congestion after connections are made, kapacitetszoner shape which customer categories connect where, steering new large-scale connections toward areas where their load profile reduces rather than exacerbates transmission constraints. (Source - Svk Anslutningsprocessen Rapport (2026))

This is a locational congestion prevention mechanism rather than an operational congestion management tool. Key features:

• Reserved capacity zones for specific customer categories at defined transmission connection points
• A kapacitetskarta (capacity map) showing available capacity at current, 5-year, and 10-year horizons
• Zone entry conditions can require flexibility provision, voltage regulation capability, or production profile matching — directly embedding congestion management capability as a connection condition
• Complements Villkorade Avtal and Flexibility Markets, which manage congestion operationally after connection

Data gaps

• Congestion costs in Sweden: redispatching prices paid; countertrading volumes and costs at Svk level
• Which 2 DSOs used villkorade avtal (non-market redispatching) in 2024 — unnamed in Ei R2025:13
• Comparison of congestion management approaches across Nordic DSOs
• Svk subscription policy: criteria for denying vs granting temporary subscriptions
• Post-2024 congestion profile in Södra Skåne following the 400 kV upgrade
• Whether abonnemangsoptimering creates net system benefits or merely redistributes costs from DSOs to Svk