Small DSO Capacity — The Binding Constraint on Swedish Flexibility Policy
One proposed fix doesn't require small DSOs to build capacity at all — a "digital connecting SO," a licensed third party that handles metering, baseline calculation, and dispatch signals independently of which DSO owns the physical network, letting a small municipal DSO meet its neutral-facilitator obligations without ever building internal digital infrastructure itself.
A DNDP-FNA-NC DR architecture built on thin data from the long tail (~162 of today's 168 companies) isn't worthless — the handful of large DSOs dominate by customer count and provide most of the signal — but it systematically underrepresents exactly the rural, peripheral areas where renewable connection is growing fastest and flex-market development is most absent, so policy calibrated on that data quietly optimizes for large urban DSOs' conditions, not the long tail's.
The EU’s flexibility regulatory architecture — DNDP, FNA, NC DR — is a data pipeline. DNDPs produce raw flexibility need data; the FNA aggregates it into a national picture; the NC DR uses it to calibrate market design. The pipeline is only as strong as its weakest link. Sweden has 168 elnätsföretag (Ei R2026:05, 2025). Six of them are sophisticated actors with dedicated flex teams, running live markets, and producing detailed DNDPs. The other ~162 are mostly small municipal utilities whose staff have other jobs. The binding constraint on Swedish flexibility policy is not regulatory ambition or platform technology. It is organizational capacity in the long tail.
The shape of the population
Ei’s synthesis of the first DNDP cycle categorizes the then-submitting population by size (Source - Ei PM2025-03 DNDP Sammanställning (2025); 152 of 155 first-round DNDPs reviewed). Today’s company count is 168 (Ei R2026:05); the long tail outside the six large is ~162, not the older unlabelled 149:
| Category | Customers | Number of companies | Customers served |
|---|---|---|---|
| Large | >100,000 | 6 | 3,432,500 (~60%) |
| Medium | 10,000–100,000 | 70 | 1,884,400 (~33%) |
| Small | <10,000 | 61 | 318,300 (~6%) |
| Production networks | — | 15 | ~1,500 |
The six large DSOs — Vattenfall Eldistribution, E.ON Energidistribution, Ellevio, Göteborg Energi Nät, and two others — serve around 60% of Swedish customers. The other companies serve about 40%: in Ei’s first-round table above that is 146 companies (70 medium, 61 small and 15 production networks), and on today’s count of 168 the long tail outside the six is ~162, a population Ei’s customer split does not cover. In terms of grid length and geographic extent, that ratio inverts: the long tail covers most of Sweden’s territory, including the rural and peri-urban areas where renewable generation connections are fastest-growing and grid constraints are most likely to be seasonal and localized.
Flex market activity tracks the company size distribution almost perfectly. 100% of large DSOs use flexibility services today. About 20% of medium-sized DSOs do. For small DSOs, the figure is around 10% — and these are mostly villkorade avtal or informal bilateral agreements, not structured market procurement. Only ~15% of all DSOs are considering creating or participating in a formal Flexibility Market (Source - Ei PM2025-03 DNDP Sammanställning (2025)). Sweden = the only Nordic country with permanent distribution-level flex markets (Source - Nordic Energy Research 2025-03 Current Utilisation of Flexibility in the Nordics) — but that distinction rests almost entirely on what E.ON and Göteborg Energi are doing.
The capacity deficit
Survey data makes the gap concrete. A 50-DSO survey conducted for Energiforsk 2026:1157 found:
- 28% of DSOs lack a documented forecasting methodology
- 40% lack documented power templates
- 36% do not validate their forecasts systematically against actuals
- Only 19% do continuous validation
(Source - Energiforsk 2026-1157 Nationell Metod Effekt och Kapacitetsprognoser (2026))
A parallel study drawing on UK and Portuguese DSO interviews found that approximately 1 FTE per 40,000–80,000 grid connections is required for DNDP production and maintenance — even in countries with mature DNDP practice and purpose-built tooling (Source - Energiforsk 2024-1043 DNDP Analys och Flexibilitet (2024)). For a medium-sized Swedish DSO with 50,000 connections, that means roughly one person. For a small DSO with 8,000 connections, it means a fraction of someone’s time — typically someone who also handles other grid operations. Excel dominates the workflow across all geographies and company sizes; purpose-built tooling is the exception.
A qualitative study of eight Swedish DSOs confirmed the picture from a different angle: many lack data-driven forecasting models entirely, and when asked what they want, the consistent answer is holistic end-to-end solutions — not methodology guidance, but someone else to do it (Source - Energiforsk 2026-1168 AI-modeller Prognostisering Efterfrågan El (2026)). That framing reveals something important about organizational reality: the constraint is not access to methods but capacity to apply them.
The reasons DSOs give for not using flexibility alternatives to grid investment are equally revealing. PM2025:03 records: no need assessed, flex can’t replace certain investments, lack of analysis capacity, insufficient competence (Source - Ei PM2025-03 DNDP Sammanställning (2025)). The last two are organizational constraints, not market or regulatory ones.
The compliance stack
Three mandates have landed on all 168 DSOs in rapid succession:
1. DNDP (EIFS 2024:1) — Submit by 31 December 2024; mandatory biennial from now. Requires capacity forecasting, flex need quantification in MW, investment planning, and six-week public consultation. Done: 152 of 155 submitted by January 2025. But Ei explicitly acknowledges that the submissions are not comparable: different reference years, different coincidence factor assumptions, some reporting accumulated totals and others annual increments, no standardized data format. The mandate was met; the data quality is variable.
2. FNA (ACER Decision 05-2025) — DSO submissions were due April 7, 2026; Svk’s national report was published June 29, 2026 and approved by Ei in July 2026. Requires structured flexibility need quantification to FNAM standards, barrier analysis across six categories, and Tabell 15 data in a defined format. This is a higher analytical bar than DNDP: the FNAM methodology is specific, the barrier categories require qualitative assessment, and the data must be aggregable across DSOs by Svk and Ei.
3. NC DR — T&C process to be designed within 12 months of entry into force (expected 2026–2027). Every DSO with a local services market obligation under the NC DR must participate in the national Terms & Conditions process. Ei has explicitly flagged Sweden’s “large number of small and medium-sized DSOs” as the primary design challenge for this process (Source - Ei NC DR Förberedelser (2025)).
The three mandates compound sequentially. DNDP data feeds FNA; FNA shapes NC DR T&C design. A DSO that cannot produce reliable DNDP flex need quantification contributes noise to the FNA aggregate. A DSO with no experience of flex need analysis has little constructive input to NC DR T&C discussions about local services market design.
Why this blocks flex development
The capacity problem interacts with the revenue cap problem to produce a particularly stubborn equilibrium. Intäktsreglering is ranked the single largest structural barrier to flex market development by far — 38 points out of ~250 barriers catalogued, first place by a large margin (Source - FlexAbility Delrapport 5 (2025)). The TOTEX reform in RP5 (2028–2031) is designed to fix the CAPEX bias by making DSOs indifferent between capital and operational expenditure solutions (lösningsneutralitet). But RP5 doesn’t begin until 2028, and even with better incentives, a DSO with no flex methodology, no vendor relationships, and no staff capacity is poorly placed to change its procurement behavior.
The geographic concentration of existing flex activity means that the case for flexibility as a system solution is built almost entirely on evidence from large DSOs with atypical internal capacity. When E.ON runs 12 SWITCH markets and Göteborg Energi runs Effekthandel Väst, it looks as if Swedish DSOs have embraced flex markets. But this reflects the behavior of 6 companies out of 168. The remaining ~162 represent a large portion of the geographic grid, including the rural and semi-urban areas where EV charging growth is fastest, renewable generation connections are most active, and local congestion is most likely to be addressed — for now — by waiting for someone to build more cable.
Possible structural responses
Derogations and exemptions: The NC DR draft allows derogations from market-based flex requirements where it is not cost-effective. Small DSOs with low load density and no viable FSP base can plausibly argue for derogation. This is the path of least resistance for the smallest networks — but it institutionalizes a two-tier system in which the geographic and organizational periphery of the Swedish grid remains outside the flex architecture indefinitely.
Shared services and common infrastructure: The most promising response. Energiforsk 2026:1157 Phase 2 will develop a shared forecasting data platform based on national data sources and a standardized methodology — modeled on the Netherlands capacity map (Netbeheer Nederland/TenneT, jointly maintained, monthly updates) and Norway’s wattapp.no (built in eight weeks, shared data structure). If Ei secures bemyndigande for structured DNDP data reporting, a common digital pipeline becomes feasible: DSOs contribute structured data to a central platform rather than producing standalone Excel-based plans. Energiföretagen Sverige already coordinates DSOs through AG Helhet Flex and AG Kapacitetsprognoser — the governance infrastructure for shared services exists at least in outline.
A smaller-scale version of this pattern is already running: Arvika Flex, an Energiforsk-financed Glava Energy Center pilot, develops a reusable flexibility-potential-mapping methodology using Arvika Elnät (via grid manager Teknik i Väst) as the case study, with Ellevio and Ei as a reference group and Flexia Consulting providing calculation support. The finding echoes the diagnosis above directly: local businesses in the studied grid area have real, unexploited flexibility potential, but “mapping resources takes time” — exactly the staff-capacity constraint this page describes, being addressed here via an externally-run methodology rather than in-house DSO tooling. (Source - Glava Energy Center Arvika Flex Pilot (2026))
Digital connecting SO: The EU DSO Entity’s Expert Group on Distributed Flexibility describes DSOs evolving toward a role as Digital Connecting System Operators — a licensed entity providing the digital interface between flexible customers and the grid, handling metering, baseline calculation, and dispatch signals independently of which DSO owns the physical network. This is a general framing in the report’s Recommendation R8 and conclusions, not itself a numbered critical action (Source - DSO Entity Distributed Flexibility Practices (2026)). Applied to Sweden’s small-DSO problem, it would allow small DSOs to fulfill their neutral-facilitator obligations without building internal digital capacity. It is a medium-term structural idea rather than a near-term solution, and it raises new questions about accountability and data access.
Consolidation: The Swedish DSO sector is slowly consolidating, driven by regulatory complexity and capital requirements. The cumulative compliance burden of DNDP + FNA + NC DR may accelerate this among the smallest operators. But consolidation is constrained by municipal ownership structures and political attachment to local grid control; it is slow and does not solve the near-term compliance timeline.
This is no longer just organic drift — it’s now an explicit policy question. Kommittédirektiv 2026:83 (July 2026) directly tasks a government inquiry with analysing how consolidating and reducing the number of small DSOs could be enabled and carried out cost-effectively, citing ACER’s finding that >35% of Swedish (and Finnish) distribution customers are served by a DSO with fewer than 100,000 customers, against an EU-wide average of ~8% — see Swedish DSO Landscape › Is 168 DSOs too many? A government inquiry now asks. If the inquiry’s November 2027 report leads to legislative or regulatory pressure toward consolidation (rather than leaving it to market/compliance forces alone), the “slow and does not solve the near-term timeline” assessment above may need revisiting — a directed consolidation push operates on a different timeline than organic drift.
The structural implication
Ei is aware of the problem. The bemyndigande proposal, the DNDP map tool, the early study on NC DR process design for a multi-DSO environment, the Energiforsk methodology program — these are all responses to the same diagnosis. But they are infrastructure-building measures that take years to deliver. The FNA 2026 was the first real stress test: Svk published Sweden’s national flexibility need report on June 29, 2026 (Ei approved it in July), aggregated from contributions by DSOs (up to 168 companies) of wildly varying analytical capacity.
Whether that report will be reliable enough to serve as the foundation for NC DR T&C design — and whether the NC DR T&C process can be designed so that small DSOs can participate meaningfully rather than being represented by large DSOs through Energiföretagen — are open questions. The answers will be visible in the FNA 2026 output and in how Ei structures the NC DR T&C process in 2026–2027.
The deeper question is what the regulatory pipeline produces if the weakest links don’t strengthen. A DNDP-FNA-NC DR architecture built on thin or inconsistent data from the long tail (~162 of today’s 168 companies) is not worthless — the large DSOs that dominate by customer count provide the bulk of the signal. But it systematically underrepresents the geographic periphery of the grid, the areas where rural electrification and renewable connection are most active, and the areas where flex market development is most absent. Policy designed on that data will tend to optimize for the conditions of large urban and peri-urban DSOs, not the long tail.