Flex›Source - Ei PM2026:07 Effektavgift Supervision

Source - Ei PM2026:07 Effektavgift Supervision

Ei PM2026:07 — Tillsyn av elnätsavgifter med en effektavgift (2026). Energimarknadsinspektionen supervisory review of demand-charge tariff (effektavgift) design at five DSOs serving customers with fuse ratings up to 63 A (household/SME segment). Triggered by the government assignment to repeal EIFS 2022:1 and develop a new model for effektavgifter; this tillsyn provides the empirical foundation. Raw text: raw/Tillsyn-av-elnätsavgifter-med-en-effektavgift-Ei-PM2026-07-extracted.txt.

Document metadata

  • Publisher: Ei
  • Series: PM-series (promemoria)
  • Published: 2026
  • Legal basis: Ellag + Art. 18 EMR (objektivitet, icke-diskriminering, skälighet, efficient grid use); EIFS 2022:1 explicitly not applied (being repealed)
  • Companies covered: Bergs Tingslags Elektriska (BTEA), Ellevio, Göteborg Energi Nät (GENAB), Södra Hallands Kraft (SHK), Tekniska verken Linköping Nät (TVL)
  • Customer segment: Elanläggningar with main fuse ≤63 A

Summary

Ei found that all five effektavgift models comply with current regulations. No orders to take action were issued. The review instead documents significant variation in model design — in how the billing-basis power is calculated, in time-differentiation, and in customer choice options. This variation is partly justified by differing local grid conditions, but Ei identifies a need for clearer framework rules to ensure more consistent application of the principles of objectivity, non-discrimination, and fairness across Sweden.

Notably, two of the five companies have decided to drop their effektavgifter before or shortly after the tillsyn: Ellevio from 1 June 2026 and SHK from 1 October 2026. The remaining three (BTEA, GENAB, TVL) continue operating their models.

The government has mandated Ei to develop a new effektavgift model proposal by April 2027. This tillsyn is the empirical input to that process.

Company-level model designs

BTEA (Bergs Tingslags Elektriska)

Grid profile: High share of vacation homes; strong seasonal pattern with sharp winter peaks; low utilization for most of the year.

Model design:

  • Billing basis (debiteringsgrundande effekt): the single highest hourly average (timmedeleffekt) during the month — the most aggressive approach of the five companies
  • Time-differentiation: year-round but with seasonal structure. Winter months (December–February), weekdays, daytime hours (höglastperiod): double rate (both an explicit höglastkomponent and a låglastavgift component are charged); remaining hours in winter and shoulder months: only låglastavgift; summer months (April–October) and non-höglast periods: different rate structure
  • Customer segments: divided by fuse size ≤63 A; vacation-home/residential mix handled through tariff structure

Ellevio

Grid profile: Mixed — covers Stockholm urban, suburban, and rural areas; complex customer composition.

Model design:

  • Billing basis: average of the three highest timmedeleffekter in different calendar days during the month (reduces volatility compared to single-hour methods)
  • Night-time peaks (22:00–06:00) reduced to 50% of their value in the billing calculation — a partial time-differentiation
  • Decision: dropped effektavgift from 1 June 2026, reverting to fast avgift + rörlig energiavgift for the ≤63 A segment; described as intäktsneutral at the company level

GENAB (Göteborg Energi Nät)

Grid profile: Urban network with mixed customer base — industry, commercial, residential, apartments; load influenced by both heating and activity patterns.

Model design — ordinarie (standard) model:

  • Billing basis: average of the three highest timmedeleffekter in different calendar days during the month
  • No time differentiation
  • Paused further development of the ordinarie model pending Ei’s forthcoming new rules
  • Intäktsneutral at the total level (but with some redistribution between customers)

Model design — valbar (optional) model:

  • Adds time-differentiation (höglast vs. låglastavgift component)
  • Includes a customer-chosen capacity level (effektnivå) that affects billing — gives customers stronger price signals and more control
  • Both models available; strong-signal customers can self-select into the valbar model

Previously GENAB calculated from the single highest hour; adjusted to 3-hour average to reduce month-to-month variability.

SHK (Södra Hallands Kraft)

Grid profile: Rural and coastal; dominance of villa- and fritidshuskunder; significant elvärme dependence; capacity constrained partly by limitations in overlying grid.

Model design:

  • Billing basis: average of the three highest timmedeleffekter in different calendar days during the höglastperiod
  • Time window: winter months only (November–March), weekdays, 06:00–21:00 — no effektavgift outside this window
  • Effektavgift in force from 1 October 2025 (recently introduced)
  • Intäktsneutral
  • Decision: dropping effektavgift from 1 October 2026, reverting to a model without effektkomponent for ≤63 A customers

TVL (Tekniska verken Linköping Nät)

Grid profile: Multi-voltage network; mixed customer base with large customers (relatively flat load) and 16–63 A customers (temperature-dependent, significant share of peak demand).

Model design — standardmodell:

  • Billing basis: average of the five highest timmedeleffekter in different calendar days during the month — the most hourly inputs of any of the five companies
  • Season-differentiated: higher effektavgift in winter than summer
  • Apartment and single-phase customers excluded (high aggregation, low individual peak impact)

Model design — alternativ valbar modell:

  • Effektavgift split into separate dag and natt components instead of seasonal differentiation
  • Different pricing between day and night periods

Outcomes reported:

  • TVL noted reduced peak demand and some shift of consumption to lower-load periods after introducing effektavgift — but effect described as limited, especially short-term
  • Revenue from the ≤63 A customer group was lower than expected in the first period after introduction

Ei’s analysis and conclusions

4.1 All five models comply with current regulations

Ei concludes that none of the five models violates the requirements in ellagen and the Electricity Market Regulation (Art. 18 EMR). The current framework leaves broad discretion to DSOs in how they design tariff models, and Ei found no violations of objectivity, non-discrimination, fairness, or compatibility with efficient grid use.

However, Ei notes that concepts such as kostnadsriktighet and effektivt utnyttjande allow multiple reasonable interpretations — and that different local grid conditions mean different models can each appear reasonable in their own context.

4.2 Local conditions justify variation but create inconsistency

BTEA and SHK operate rural/vacation-home networks with sharp seasonal peaks — their time-limited and high-sensitivity models reflect those realities. Ellevio, GENAB, and TVL operate more mixed networks. All five identify the 16–63 A customer group as central to peak demand.

4.3 Significant variation in model design

Key dimensions of variation:

  • Number of hours in billing calculation: 1 (BTEA) / 3 (Ellevio, GENAB, SHK) / 5 (TVL standard)
  • Time window: All-year (BTEA, GENAB, TVL, Ellevio) vs. winter-only (SHK); Ellevio partially differentiates by night
  • Customer choice: GENAB (valbar with capacity level) and TVL (alternative day/night model) offer choice; others do not
  • Retreating companies: Ellevio (1 June 2026) and SHK (1 October 2026) are dropping effektavgifter; the other three continue

4.4–4.5 Customer complexity and behavior

Effektavgifter create complexity for customers:

  • Predictability is challenging, especially with single-hour or few-hour billing calculations
  • Companies that allow customer choice (GENAB, TVL) address heterogeneity but require more customer understanding
  • Behavior change effects were noted at TVL (reduced peaks, some load shifting) but are limited and gradual
  • Initial customer reactions often centered on understanding the mechanism; later on operational use

4.6 Need for clearer framework

Ei concludes there is a significant need for clearer rules on effektavgift design. The current broad discretion leads to tariffs that are difficult for customers to compare or understand across DSOs, even where local differences do not objectively explain the variation.

The government has mandated Ei to:

  1. Repeal EIFS 2022:1 by 30 June 2026
  2. Submit a new effektavgift model proposal by 12 April 2027

Ei has begun this work and will consult DSOs, customer representatives, and other stakeholders. The aim is a framework that is transparent, proportionate, and comprehensible to customers while maintaining relevant incentives for efficient grid use. This tillsyn is described as an important input to that work.

Relevance to existing wiki topics

  • Swedish DSO Tariff Reform — Three Parallel Tracks (2025–2027) — Track 2 empirical data: five model designs, two retreating companies, Ei’s overall finding, forthcoming new model.
  • Göteborg Energi Nät — GENAB effektavgift model details (ordinarie + valbar; 3-hour average; paused development; intäktsneutral).
  • Ellevio — Effektavgift discontinued from 1 June 2026; 3-hour average with 50% night reduction was the design.
  • Ei — Adds PM2026:07 to Ei’s tariff oversight record; signals April 2027 new model proposal.
  • Demand Response — Empirical evidence that effektavgifter produce limited but detectable behavior change (TVL data); customer complexity as a barrier.
  • Effektavgift — First cross-DSO empirical comparison of effektavgift designs in the Swedish household/SME segment.