Source - Flexsäsongen i Götene 2024-2025 Sweco

Metadata

Summary

The fullest operational report on any Swedish local flexibility market outside the large DSO programmes. Sweco describes the 2024–2025 season of Kinnekulle Energi‘s local flex market in Götene — the first commercially active season after a 2023–2024 proof-of-concept pilot. The report covers FSP recruitment, product design, the complete operational process, season results with exact figures, a quantitative analysis of the grid’s flex potential, and a scenario analysis showing how much additional load the market could accommodate.

The main conclusion is that the market worked as a concept but is currently constrained by the total MW that can be activated simultaneously (5.28 MW) — not by endurance or energy. The biggest strategic decision is the removal of MaxUsage from the 2025–2026 season, replaced by a pure LFM-h availability model, due to both baseline validation difficulties and the forthcoming time-differentiated power tariffs (EIFS 2022:1, mandatory from January 2027).

Context

Götene is an industrial municipality in Västra Götaland (SE3), ~13,000 inhabitants. Major industries: Dafgårds (food), Semper (food), Arla, Nolato (manufacturing), Paroc. The local DSO — Götene Elförening ek. för., trading as Kinnekulle Energi — connects to Ellevio’s regional network. Growing industrial electricity demand means capacity shortage is forecast even after Ellevio granted a subscription increase at the start of the season. The flex market is the chosen strategy to defer or avoid capital investment in grid reinforcement.

The pilot season (2023–2024) tested an energy activation product (LFM-e) at D-2 lead time. The 2024–2025 season switched to a LFM-h availability product to allow activation closer to the actual need hour (H-1), improving forecast accuracy and reducing over-procurement.

Four FSPs recruited

Available capacity: up to 5.28 MW per hour, of which up to 5 MW has effectively unlimited endurance (Dafgårds elpanna via continuous HVO supply).

Recruitment process: digital letter to all customers with >63A fuses or power subscription, direct dialogue with 18 high-voltage customers, aggregator outreach (one of several aggregators contacted had sufficient aggregated capacity in the area), website interest form integrated with CRM via API.

Products and market design

LFM-h (tillgänglighetsprodukt): DSO calls for available capacity D-7 to D-2 (latest 14:00). Activation occurs from D-1 07:00 to D-1 15:00 (or up to H-1). The two-stage process — avrop first, activation later if needed — reduces unnecessary activation while guaranteeing a reserve. Only operational hours 07:00–15:00 per contract.

MaxUsage: FSP holds power consumption below a defined contractual maximum during high-load hours (weekdays 06–16). GKT’s contract: reference 450 kW, maximum 350 kW. Validated by counting measurement values per hour that fall under the limit. 95% delivery rate over the season.

Pricing: Bilateral contracts negotiated before the season. Basis: alternative cost of energy from HVO combustion plus a fair markup per MW-hour. Activation prices ~2,300–2,500 SEK/MW/h; availability payments ~200–300 SEK/MW/h; MaxUsage ~300–350 SEK/MW/h.

Operational process

• Forecast tool: Aiolos software by VITEC, with integrated mathematical models; prognostic meetings twice weekly (Monday and Thursday 09:00); automated email alert when high-load predicted; improved from one summaserie to per-station data toward end of season
• Avrop/activation: fully manual via email and/or phone within agreed format and time windows
• Validation: manual in Microsoft Excel at month-end; data collected from FSPs in .csv format + grid operator database
• Payments: monthly
• Transparency: market data (avrop, activations, validated MaxUsage, grid load, abonnemangsgräns) published monthly on website

Season results — exact figures

LFM-h (tillgänglighetsprodukt):

• 20 avrop occasions; 151 hours of first-prognosis availability called; 15 of those hours proceeded to second-prognosis activation
• Activation rate: 15.6% of avropad capacity was actually activated

Weekly breakdown (Table B1): LFM-h avrop occurred in weeks W50, W3, W5, W7, W8, W11, W14. W7 was the single largest week (41.08 MWh avropad, 4.84 MWh activated). MaxUsage ran every week at ~4–5 MWh/week (except shortened weeks W53 and W14).

MaxUsage (GKT cold storage):

Each validated hour = 100 kWh (450 kW reference − 350 kW max = 100 kW reduction).

Grid situation — actual need

The 2024–2025 winter was mild. No hard breach of the abonnemangsgräns occurred. Week 50 (early December) was the tightest. The DSO deliberately continued procuring flex throughout the season to build market liquidity, noting this is the same approach taken by Effekthandel Väst and E.ON markets where congestion is anticipated but not yet acute. At year-end, the subscription limit was raised (from ~55 MW to ~60 MW) when a new large industrial connection was formalized, though the new customer had not yet started drawing power.

Data analysis — flexibilitetspotential (Chapter 4)

Full-year 2024 data was used as a reference (a cold year — 2024 was a leap year). Summaserie from six measurement time series across four grid stations (four load, two production).

Key finding: The abonnemangsgräns (~54 MW avg over 2024) was exceeded 17 times in 2024, representing:

• Total excess energy: 31.5 MWh
• Maximum exceedance in any single hour: 5.4 MW

The 2024–25 market (5.28 MW available) would have handled all but the single worst hour of 2024 (5.4 MW > 5.28 MW available).

Binding constraint: Available MW per hour — not endurance, not total energy. The elpanna’s HVO endurance and the total energy traded (125.4 MWh) both far exceed what was actually needed in 2024 (31.5 MWh).

Scenario analysis — effect of lowering the abonnemangsgräns to different percentile thresholds:

Even the most modest scenario (top 1% flexed) requires 8.9 MW — well above current 5.28 MW. Recruiting more FSPs is the binding growth constraint.

The inverse reading: given a known flex capacity, the DSO can quantify how much additional load can be admitted into the network despite the subscription ceiling.

Key decisions and conclusions

1. LFM-h is the right product — real-time activation (H-1) maximises forecast accuracy; plan to develop toward LFM-p (multi-season) contractual form
2. MaxUsage removed from 2025–26 onwards — two reasons: (a) baseline/reference value problem when energy efficiency improvements occur between seasons (confounds the reference); (b) time-differentiated power tariffs mandatory from January 2027 (EIFS 2022:1) will implicitly achieve the same load shifting, making MaxUsage redundant
3. Multi-year contracts (5–10 years) recommended to secure FSP participation and enable FSP investment, and to prioritize the local flex market over competing revenue streams (Dafgårds already participates in Svk FCR — longer contracts would give local market preference)
4. Villkorade avtal as backstop — the voluntary flex market cannot guarantee capacity reserve from year to year; conditional connection agreements should be established as insurance
5. Digitalization next — after two manual seasons, Kinnekulle Energi is ready to specify and procure an IT system; deliberately waited to first understand requirements before committing to procurement

Relevance to wiki topics

• Kinnekulle Energi — primary operational data for the entity page
• Flexibility Market — resolves data gap on Götene Flex; only Swedish market operating without SWITCH or NODES; first confirmed multi-market stacking (local + Svk FCR)
• Villkorade Avtal — explicit DSO recommendation for villkorade avtal as backstop; aligns with E.ON’s sequential model
• Baseline Methods — MaxUsage baseline problem: reference value erodes when energy efficiency improvements occur between seasons; a real-world documented failure mode
• Flexibility Market › LFM-h / LFM-p / LFM-e — confirms LFM-h as the correct product for real-time need activation at small DSOs
• Aggregation — Checkwatt named as the aggregator; battery contracted with 12-hour availability blocks and guaranteed 1-hour activation
• The Flexibility Provider Base — Structure, Barriers, and the Aggregator Constraint — Dafgårds as anchor FSP combining local flex + Svk FCR; confirms multi-market stack is operationally achievable
• Distribution Network Development Plan — scenario analysis demonstrating how LFM capacity can be used to quantify permissible new connections at a subscription ceiling