Anslutningsavgift 2027 — What Changes for DSOs and Customers

From 1 January 2027, Sweden replaces DSO-by-DSO connection-fee practice with one nationwide rulebook (EIFS 2026:10). The headline principle — a “shallow” charge covering only customer-specific costs — sounds like it should make connections cheaper and more predictable across the board. In practice the effect splits sharply by customer type: small, straightforward connections get simpler and more predictable pricing; large, remote, or capacity-hungry connections can face a higher bill than under today’s looser, more negotiable practice, because the same reform that constrains “normal” charges also hands DSOs a sharper, explicit rulebook basis for charging more when a connection is genuinely expensive. A second, less visible change — a revenue-cap adjustment starting with RP5 (2028) — closes a gap where customers had effectively been paying for the same connection twice. This page works through both halves for the two audiences who actually have to act on them: DSOs redesigning their charging methodology, and customers about to connect.

What’s actually changing, in one paragraph

Today, each of Sweden’s ~168 nätkoncessionshavare designs its own connection-charge methodology, within only loose statutory guardrails — producing materially different bills for similar connections depending on which DSO’s territory you happen to be in. EIFS 2026:10 replaces this with one detailed, common rulebook: a shallow-charge default (customer pays only customer-specific costs up to the nearest feeding/receiving transformer station or network part; overlying-grid costs go to the tariff-paying collective instead), a särskilda-skäl exception requiring a deeper charge where the shallow default would leave the DSO recovering only a small share of the cost (Ei’s allmänna råd list five example situations), and mandatory standard (schablon) pricing for small connections (≤25 A) based on distance to the nearest transformer station. It implements EU Regulation (EU) 2019/943 Art. 18’s requirement for pre-established, pre-approved connection-charge methods — Sweden’s prior regime didn’t meet that bar. See Anslutningsavgift for the full rule text.

What changes for customers

If you’re connecting a small facility (≤25 A) — more predictable, not necessarily cheaper

Household and small-business connections must be priced by schablon: a fixed fee for the first 100 m (as-the-crow-flies distance to the nearest transformer station), a per-metre schablon rate from 100–600 m, and an individually-calculated rate only beyond 600 m. The practical effect is predictability, not a price cut — a DSO that today might quote a low, informally-negotiated fee for a nearby connection, or a high one for a distant one, must now apply the same published formula to everyone in comparable circumstances. Whether any individual customer ends up paying more or less than they would have under the old practice depends entirely on what that DSO was actually doing before — which is precisely the fragmented, hard-to-compare picture EIFS 2026:10 is meant to fix.

If you need a large, remote, or capacity-heavy connection — the exception is where the real money is

The shallow-charge default only holds while it doesn’t leave the DSO drastically undercompensated. Ei’s own examples of when a deeper charge is required read almost like a checklist of exactly the connections flexibility and grid-capacity policy cares most about: a remote facility needing dedicated infrastructure only it will use; a facility using shared network capacity out of proportion to others; a connection requiring overlying-grid reinforcement; connection to a capacity-constrained meshed network; or a facility that will only operate for a limited time. A large new industrial load, a data centre, an EV charging hub, or a rural connection needing real grid build-out is disproportionately likely to land in this bucket — and the DSO now has an explicit, rulebook-based basis for charging the fuller cost, where before this may have been a matter of negotiation or informal practice — though the föreskrift text itself doesn’t spell out a formal Ei prövning process for invoking the exception; Ei’s general prövning channel for connection disputes remains the backstop. The predictability reform and the “connections in constrained areas get expensive” reality are not in tension — the second is the necessary release valve for the first.

The alternative lever: paying less upfront by accepting curtailment

A customer facing a deep connection charge has, in principle, a second option Sweden has been building out in parallel: a villkorat avtal (conditional/flexible connection), trading a lower guaranteed capacity — curtailable by the DSO during congestion — for faster or cheaper access. EIFS 2026:10 doesn’t create this trade-off, but by finally codifying how expensive the “just pay for full capacity now” path can be, it makes the flexible-connection alternative easier to evaluate against a known baseline rather than an opaque, DSO-specific quote.

Microproducers: the exemption is gone

The old ellagen’s fee exemption for microproduction connections is explicitly not carried forward — Ei found it incompatible with EU law’s requirement that connection charges be objective and non-discriminatory (a blanket waiver for one customer class effectively makes everyone else subsidise it). Anyone connecting rooftop solar or similar small-scale production from 2027 pays the same anslutningsavgift regime as any other connecting customer, sized by the same rules above.

What does not change

This is a rulebook for new connections, reconnections, and increases to contracted power (effekthöjningar) — it says nothing about, and does not retroactively touch, the network tariffs (nättariffer) that existing connected customers already pay for ongoing grid use. If you’re not connecting, reconnecting, or asking for more effekt, EIFS 2026:10 doesn’t change your bill directly.

What changes for DSOs

A compliance rebuild, not a rate change

Every nätkoncessionshavare must redesign its connection-charge methodology around the shallow/deep-charge structure and either adopt or develop compliant schablonvärden for ≤25 A connections by 1 January 2027. The föreskrift’s allmänna råd say DSOs may use the schablonvärden Ei publishes on its website (today, the PM2013:03 values for 16–25 A connections, see Source - Ei Anslutningsavgift 2026 Schablonvärden (2025)), which smaller DSOs without in-house tariff-modelling capacity may lean on — but the sources don’t say whether Ei will issue new or adjusted values for the 2027 regime. DSOs that want to use the deeper-charge exception now carry a documentation and justification burden: they need to be able to show why the shallow default would leave them under-recovering — the föreskrift text doesn’t itself impose a formal justification or Ei-prövning procedure for the exception, but a disputing customer can still bring it to Ei via the general connection-dispute prövning channel, so this remains an untested source of case law and dispute risk (Ei itself acknowledges “viss osäkerhet” in the initial period).

The revenue mechanics — and the RP5 connection most DSOs will feel more than the fee rule itself

This is where the two 2026 reforms this wiki has tracked separately turn out to be one story. Ei’s own RP5 consultation documents a structural flaw in current revenue-cap practice: assets whose cost was already covered by an anslutningsavgift are nonetheless included in the DSO’s capital base (kapitalbas) — meaning the DSO earns a regulated return on capital it never actually put up, and the connecting customer pays for the same asset twice: once directly via the connection fee, once indirectly via everyone’s network tariffs. Starting with RP5 (2028–2031), Ei intends to deduct connection-fee-funded assets from the revenue frame (same discount rate and depreciation period as the capital-cost calculation, applied to connection-fee income from RP5 onward) — closing that gap. See RP5 Revenue Cap Methodology.

Read together: EIFS 2026:10 (2027) fixes how much a DSO can charge for a connection up front; the RP5 deduction (2028) fixes what happens to that money afterward. For a DSO, the net effect over 2027–2031 is a genuine shift in revenue composition — less latitude to recover connection costs informally or to keep double-dipping via the capital base, and a harder requirement to get the upfront anslutningsavgift right, because it increasingly has to actually cover what it’s supposed to cover.

Exposure is uneven across the DSO population

A DSO serving a dense urban/suburban area, where most connections are close to existing infrastructure, will mostly be applying the ≤25 A schablon and rarely reaching for the särskilda-skäl exception. A DSO with a large rural or geographically dispersed service area — exactly the profile of many of Sweden’s smaller municipal and cooperative DSOs — will use the deeper-charge exception far more often, and carries proportionally more of the new documentation/prövning burden per connection. This is a distinct, narrower version of the general small-DSO capacity strain already tracked in Small DSO Capacity — The Binding Constraint on Swedish Flexibility Policy.

On the horizon, not yet decided: Dir. 2026:83

None of this is affected in the near term by kommittédirektiv 2026:83 (decided July 2026) — it’s an investigation mandate reporting 24 November 2027, after EIFS 2026:10 is already in force and after RP5’s first revenue-cap decisions are already in motion. But it signals where policy pressure is heading next: possible caps on how fast tariffs can rise (Finland’s 8%/year model is the cited reference), a rebalancing of appeal rights toward the customer perspective, tighter limits on överrullning (deferred billing of shortfalls), and — separately — a possible restructuring of the grid into the EU’s two-level model and consolidation of small DSOs. See the full breakdown. None of it is decided; treat it as “watch this space,” not as something to plan around yet.

Timeline

Data gaps

• Whether any DSO-level data exists yet estimating the aggregate revenue-cap effect of the RP5 connection-fee deduction (i.e., how large the “double payment” gap actually was in practice)