Independent Aggregation in Sweden — The Implementation Gap

Sweden’s statutory framework for independent aggregation exists, and Svk has a BSP agreement template. Yet in 2026, an aggregator with 500 MW of qualified flexibility resources spends half its working time on BRP administration workarounds — and the tools that would resolve this are 2–6 years away. This synthesis maps the distance between what Swedish law says and what Swedish markets can actually do.

The obligation and the gap

The EU requirement: The Clean Energy Package (Directive 2019/944 Art. 13, 17) guarantees aggregators the right to access markets without approval from other market actors and to choose their own BRP. The Electricity Balancing Guideline (EB GL, Regulation 2017/2195) operationalises this: member states must establish national terms enabling aggregation, including cross-BRP portfolio bidding. Art. 18.1’s actual deadline is for TSOs to submit a proposal for those terms — six months from entry into force (~June 2018), not to have them approved and in force. Svk met that deadline, submitting its first proposal to Ei in 2018. There is no separate EB GL deadline for the terms’ final approval. (Source - EB GL (Regulation 2017-2195) › Article 18 — National terms obligation (the BSP implementation article))

Sweden’s response: Svenska kraftnät split the old balansansvarig role into BSP (Avtal 5937-2) and BRP (Avtal 5938) in May 2024 — nominally implementing the BSP role. Market actors describe this split as a “paper construction”: to sign a BSP agreement, an actor must already hold a BRP agreement and be BRP at the delivery points where ancillary services will be delivered. The pre-2024 cross-BRP constraint is reproduced in slightly different form. An aggregator managing resources across ten electricity suppliers must still create ten separate bid portfolios. (Source - Svk Införande BSP BRP)

The multi-year gap between Svk’s 2018 proposal and its 2024 implementation reflects Ei’s iterative national approval process (change requests in 2019 and 2020, a third consultation round still open as of November 2021), not a missed EU deadline — there was none to miss. Full functional BSP implementation (standalone BSP role) is now targeted for 2028–2029. As of mid-2026 the government had taken no official steps to accelerate it (confirmed in Betänkande 2025/26:NU25).

The policy chain — what was proposed, what was legislated, what was deferred

Step 1: Ei R2021:03 — the legislative proposal (2021)

Ei’s major report (Source - Ei R2021-03 Oberoende Aggregatorer) identified that existing Swedish law — one BRP per uttagspunkt — was incompatible with EMD Arts. 13 and 17. It proposed:

• The balanspunkt concept: a sub-meter point behind the customer’s main grid connection (uttagspunkt), at which an aggregator’s BRP takes balance responsibility for activations while the customer’s existing BRP retains responsibility at the uttagspunkt for everything else
• Two models (aligned with NordREG’s framework):
  • Model 1: multiple BRPs per metering point via balanspunkt + sub-metering
  • Model 2: single BRP with compensation mechanism (Svk designs methodology; Ei approves before use)
• New ellagen definitions, extended metering obligations, updated BRP framework, new aggregator rights chapter
• Svk mandate to design the compensation methodology

Konkurrensverket’s critique (Source - Konkurrensverket Yttrande Ei R2021-03): the proposed direct-cost cap on BRP compensation is asymmetric — compensation flows only aggregator→BRP, never in reverse, even when activations reduce BRP imbalance costs. This entry barrier grows with the aggregator’s utility. KKV’s concern directly anticipates the ellag 8 kap. 16 § requirement that BRP compensation terms “take into account the benefits the aggregation service may provide to other BRPs” — a requirement Sweden has not yet operationalised.

Step 2: Prop. 2025/26:240 — what was codified in primary law

Source - Prop. 2025-26-240 Nya lagar om elsystemet (2026) enacted a new elmarknadslag codifying the BSP/BRP framework in primary law — including a new chapter on aggregators, the FSP role (replacing the earlier LA — leverantör av aggregeringstjänster), and new definitions (flexibilitetsmätpunkt, abonnemangspunkt). The ellag 8 kap. 13–26 §§ provisions governing balance responsibility and compensation were aligned with the new framework.

What the Prop. did not specify: the detailed compensation model design. Whether Model 3 (multiple delivery points) or Model 4 (korrigerad faktura) is the primary operational path, and how the compensation methodology works in practice, was left to implementing regulations. The Prop. codified the architecture but not the plumbing.

Step 3: Svk’s compensation model design (2024)

Svk’s September 2024 government assignment reports (Source - Svk Kompensationsmodell Delrapport 1 (2024), Source - Svk Kompensationsmodell Delrapport 2 (2024)) proposed the detailed compensation architecture:

Model 3 — Flera leveranspunkter: The flexibility resource receives its own leveranspunkt (delivery point) within the customer’s internal network. A second BRP (BRP2) takes balance responsibility specifically for that resource. No compensation between FSP and original supplier is needed — no shared metering uncertainty. Can be implemented immediately once ellagen is amended, without the central information system. Transition path: available now for resources ≥1 MW (consumption) or any production/storage.

Model 4 — Korrigerad faktura (recommended): Customer keeps a single electricity contract. FSP contracts with BRP2, which takes balance responsibility for activations. The customer’s BRP1 receives compensation from the FSP, via the supplier adjusting the customer’s electricity bill for the activation volume. Svk administers and verifies volume centrally but no money flows through Svk.

• Compensation price: spotpriset (day-ahead spot price for the activation period), regardless of the market the FSP activates on. This means on the day-ahead market, FSP revenue ≈ compensation owed — near-zero margin. Model 3 avoids this.
• Reference profile: deferred to NC DR alignment
• Prerequisite: the central information system (centralt datahanteringsverktyg). The September 2026 Ei/Svk report gives 36–42 months (reuse) or 90–96 months (from scratch) from the start of implementation. Until it exists, only Model 3 is possible.

Svk recommended both models coexist: Model 3 as the immediate transition path, Model 4 as the steady state once DHV is operational.

Government has declined to implement this model (2026). Svk’s Sept 2024 report was delivered under assignment KN2023/03647. In answer to written question 2025/26:611 (late March 2026), energy minister Ebba Busch stated the government has no plans to implement the model Svk proposed in that report, or to take it further. The Riksdag reaffirmed this in Betänkande 2025/26:NU25 (15 June 2026): rejecting the C-motion’s demand for a simpler model, the utskott reasoned the demand was “tillgodosett” precisely because the government will not introduce Svk’s contested model. So the specific Model 3/4 compensation design above is, as of mid-2026, not the committed path — the design question is effectively reopened. Criticism driving this: the model is seen as complex and a market-access barrier, and C wants the flex provider liable only for imbalance in the promised flexibility (echoing KKV’s earlier asymmetry critique) — a specific “leveranspunkt/uttagspunkt vs. system as a whole, EU law requires” framing sometimes attached to this critique is not found in the Betänkande 2025/26:NU25 text itself and should be treated as unverified pending a check against the full Motion 2025/26:4138. The administrative backbone (DHV, KN2025/01781) still proceeds; the compensation methodology it will administer is now open. Ei has since delivered its national non-fossil-flexibility target proposal — Ei R2026:07 (11 Sept 2026) proposes 0 MW for both 2030 and 2035 — but the government has not yet decided on it (expected ~Jan 2027, alongside the NECP update); until that decision lands, it doesn’t yet resolve which compensation methodology the reopened Model 3/4 question converges on. (Source - Prop. 2025-26-240 Nya lagar om elsystemet (2026))

The infrastructure dependency

The central information system is the primary constraint on the entire aggregation compensation architecture.

The September 2025 government decision (Regeringsbeslut KN2025/01781) explicitly names compensation for independent aggregation as a required DHV function: “ett datahanteringsverktyg är nödvändigt för en effektiv administrering av kompensation vid oberoende aggregering.” (Source - Uppdrag Centralt Datahanteringsverktyg (2025))

Ei and Svenska kraftnät delivered their joint DHV proposal in September 2026 (Ei R2026:08); the government has not yet decided. (Source - Ei Förslag Centralt Datahanteringsverktyg (2026), Source - Ei R2026-08 Förslag Centralt Datahanteringsverktyg (2026)) The companion FIS (flexibilitetsinformationssystem) for NC DR register functions comes after DHV v1.

Timeline implication: Ei and Svk’s September 2026 report (Ei R2026:08) gives an indicative schedule in months from the start of implementation, not calendar dates: first go-live in months 36–42 if the old Elmarknadshubb work is reused, 90–96 months if the solution is built from scratch. The government has not set a start; on the wiki’s own arithmetic a 2027 start means ~2030–2031 (reuse) or ~2034–2035 (from scratch). Model 4 compensation can only operate once the DHV is live. (Source - Ei R2026-08 Förslag Centralt Datahanteringsverktyg (2026)) FIS for NC DR aggregation register: after DHV.

The Nordic Imbalance Settlement Handbook v5.2 confirms that the settlement infrastructure for cross-BRP aggregation already exists at eSett level — a dedicated Compensation mechanism, priced at the day-ahead market price (distinct from VoAA/Incentivizing Component, which is general imbalance pricing for ISPs with no dominating regulation direction, not aggregation-specific). The constraint is purely Sweden’s national-level requirement that BSPs hold BRP agreements — the one the DHV is intended to dissolve. (Source - Nordic Imbalance Settlement Handbook v5.2 (2025))

What is and is not operational today (May 2026)

Aggregators operating under current constraints (six primary aggregators or flexibility specialists registered as BSPs as of March 2026: Capalo AI, Mind Energy, Entelios, Flower, Ingrid Capacity, Vimab BESS): they operate by maintaining separate BRP agreements per delivery point cluster and submitting multiple bids. One aggregator with ~500 MW of qualified flexibility resources reports spending up to 50% of working time on BRP administration friction — a single interviewee’s account in the source, not a general figure across all six (Source - FlexAbility Delrapport 5 (2025)). CheckWatt illustrates the structural cost directly: in Sweden it must route through a BRP intermediary charging an additional 5–10% fee on top of its own take, while in Finland and Denmark it registers directly with Fingrid/Energinet, bypassing that fee — a difference CheckWatt itself reports as a plausible, unconfirmed contributor to its higher self-reported revenue multiplier in Finland (4.0×) versus Sweden (2.5×), not a figure FlexAbility’s report addresses (Source - CheckWatt Website (2025-2026)).

The estimated immediate impact of a functional free-standing BSP in 2028: +300 MW entering the market from resources that are technically prequalified but administratively blocked.

The structural barriers that persist beyond 2028

Even after the free-standing BSP arrives in 2028 and DHV resolves the compensation infrastructure gap, several structural issues remain:

The KKV cost-cap asymmetry (still unresolved): Ellag 8 kap. 16 § requires that BRP compensation terms take into account the benefits the aggregation service provides to other BRPs. No implemented compensation mechanism yet satisfies this. Until it does, BRPs that benefit from aggregation activations (by having their imbalances reduced) receive that benefit without contributing to the cost — shifting the full burden onto the FSP.

Spotpris compensation in Model 4: Pricing compensation at spotpriset makes day-ahead market FSP economics near-zero. FSPs with Model 4 resources will rationally direct those resources toward balancing markets (aFRR, mFRR) where revenues substantially exceed spot — concentrating day-ahead flexibility on Model 3 resources (larger, industrial). Household-scale flexibility accessible only via Model 4 will be structurally disadvantaged in market selection.

The 1 MW threshold for interim Model 3: Sub-1 MW consumption resources cannot self-register a separate leveranspunkt in the interim period — they must wait for DHV. The long tail of household-scale resources (batteries, EVs, heat pumps) is locked out of Model 3’s immediate path. This is the majority of the long-term flexibility resource base.

NC DR and prequalification: The Network Code on Demand Response‘s Service Providing Group qualification framework and the FIS flexibility register are the infrastructure through which aggregated household resources gain standardised access to all local DSO markets. FIS comes after DHV. Until then, aggregators must navigate each DSO’s own registration and prequalification processes — the multi-market administrative friction that the NC DR is designed to eliminate.

Sweden is not an outlier. An EU-wide industry outlook states that in 2025 most member states had not fully transposed independent aggregation into national law; in its business-as-usual scenario, lagging states keep a “soft veto” through complex notification procedures or missing compensation rules (smartEn and LCP Delta 2026).

Why the gap persisted

The Swedish statutory basis (ellag 8 kap. 13–26 §§) exists. The NordREG model was published in 2020. Svk submitted its compensation design in September 2024 — but this is a delay measured against the NordREG model’s publication and against how long Ei’s own approval process took, not against a missed EU deadline (Svk’s actual EB GL obligation, submitting a national-terms proposal, was met on time in 2018 — see above).

Several factors compound the delay:

1. Infrastructure dependency: Model 4 — the model for household-scale resources with single meters — requires a central information system that Sweden decided in 2015 (Elmarknadshubb), paused in 2020, and re-decided in September 2025. A 10-year gap in infrastructure governance translates directly into a 10-year gap in operational aggregation capability.

2. BSP as “paper construction”: Svk’s May 2024 BSP implementation satisfied the formal regulatory requirement without delivering operational substance. Ei issued an injunction; Svk responded with a revised implementation plan targeting 2028. The regulatory oversight cycle added years.

3. Compensation design complexity: The spotpris reference price, the reference profile methodology, the KKV competition concern, and the NC DR alignment requirement created legitimate design dependencies — each requiring resolution before the next step. Svk’s 2024 reports are the first authoritative design, six years after the NordREG model they build on.

4. Small market, large structural change: Sweden’s 168 DSOs, ~10–15 historically dominant BRPs, and thin aggregation market create limited commercial pressure to accelerate. The aggregators most harmed by the delay are the ones building the market — a constituency too small to drive legislative urgency.

Practical implications for aggregation market participants

For aggregators already in the market: Plan around the 2028 free-standing BSP as the most near-term structural change. Model 3 (multiple delivery points) is available now for resources ≥1 MW — assess which portfolio segments qualify. For sub-1 MW household portfolios, DHV (~2030–2031 at the earliest, later without reuse) is the operative horizon. CheckWatt’s model — optimise across FCR-D, mFRR, and local DSO markets with a BRP intermediary for TSO markets — remains the practical template until 2028.

For new market entrants: The EB GL Art. 18 requirement for national terms enabling aggregation (already satisfied on paper since 2018, per Svk’s Ei-approved terms) is an enforcement lever that ACER/Ei can apply against the gap between formal and operational compliance. The existing regulatory obligation is not new; what changes in 2026–2028 is the combination of: free-standing BSP approaching, DHV architecture being designed, NC DR entering force, and TOTEX creating DSO demand for third-party flexibility. The window for building scale ahead of infrastructure completion is now.

For DSOs building flexibility markets: The FIS infrastructure for NC DR register functions is 5+ years away. Local flexibility market design should accommodate direct bilateral aggregator registration in the interim — the standardised SPG qualification the NC DR promises cannot be assumed operational before 2029 at earliest.

For investors: The +300 MW BSP unlock in 2028 is the near-term catalyst. The Model 4 / DHV unlock of sub-1 MW household resources is the longer-term story — and the larger market, since the majority of Sweden’s distributed flexibility resource base is in this segment.

Relationship to other wiki pages

• Aggregation — full technical and structural detail on Swedish aggregation barriers, actor landscape, BSP architecture, consumer acceptance
• Balancing Markets — BSP/BRP reform legislative background; EB GL timeline; Nordic comparison
• Svenska kraftnät — BSP implementation authority; compensation model mandate; DHV assignment
• Elmarknadshubb — DHV/FIS infrastructure; 2015 history and 2025 restart
• Network Code on Demand Response — SPG qualification and FIS register; the NC DR’s aggregation architecture
• Distribution System Operator — DSO neutrality obligations; metering infrastructure for balanspunkt
• CheckWatt — primary case study for aggregation under current constraints
• Source - Ei R2021-03 Oberoende Aggregatorer — the legislative origin; balanspunkt; NordREG two-model framework
• Source - Svk Kompensationsmodell Delrapport 1 (2024) — Model 3/4 compensation design specification