Capacity Adequacy and Flexibility as the Missing Reserve

Sweden’s resource adequacy is deteriorating — ERAA 2025 puts Loss-of-Load Expectation at 6.5–10.3 hours/year through 2035, six-to-ten times the 1 h/year norm — yet Sweden has deliberately chosen not to build a capacity market, relying on an energy-only market plus a thin strategic reserve. The reserve’s first procurement failed outright (winter 2025/26): every bid exceeded the CONE price cap, which was set to household demand response — the cheapest theoretical flexibility — while only expensive production-side resources actually bid. That failure is the whole story in miniature: demand flexibility is now the named adequacy mitigation in Svk’s reports, but the flexibility the system needs (cheap, abundant, demand-side) is not the flexibility that shows up. This page connects the adequacy gap, Sweden’s policy choice, the strategic-reserve saga, and why flexibility is the structurally right answer that is not yet bankable for adequacy — drawing on Balancing Markets, Demand Response, Svenska kraftnät, Bidding Areas, and The Swedish BESS Business Case — Revenue Stacking and the FCR Saturation Problem.

The adequacy gap is real and growing

ERAA 2025 (the EU-wide European Resource Adequacy Assessment) shows Sweden well above the adequacy norm for the whole horizon (Source - Svk Kraftbalansen Vår 2026):

SE3 and SE4 carry the risk. SE1/SE2 have surplus hydro but inadequate transmission to move it south — the same north-south constraint that drives the bidding-zone debate. The structural drivers are familiar: nuclear closures (Ringhals 1–2, Oskarshamn 1–2), variable renewables displacing dispatchable capacity, rapid electrification demand, and the binding north-south transmission limit.

This is the “missing money” problem in concrete form (Balancing Markets › The missing money problem): zero-marginal-cost wind and solar clear first and compress wholesale prices during high-VRE hours, so dispatchable plant cannot recover capital at average prices, capacity exits, and adequacy risk rises.

Sweden’s choice: energy-only market plus a thin strategic reserve

Sweden has rejected a market-wide capacity mechanism. The Elmarknadsutredningen (SOU 2025:47) examined the question and recommended a strategic reserve instead — a small, ring-fenced volume held outside the market and activated only in scarcity, rather than capacity payments across the whole fleet.

This aligns with the EU framing: capacity mechanisms are a residual instrument, permissible only when proportionate to an ERAA-demonstrated adequacy problem and when market reforms cannot resolve it (Source - Capacity Mechanisms (EC DG ENER)). The 2024 EMD Reform softened the stance — mechanisms are no longer “last resort,” approvable for up to 10 years (Art. 21(8), Regulation 2024/1747) — but the “market reform first” principle holds. Sweden’s bet is that energy-only price signals plus flexibility plus interconnection deliver adequacy, with the strategic reserve as a narrow backstop.

The strategic-reserve saga — failure, then a quiet success

The mechanism’s first two seasons are a natural experiment in whether that bet holds.

The setup. The old effektreserv (standby, mostly fossil, ~20 years in operation) was terminated March 2025. Its replacement, the strategisk reserv under Lag 2025:50, received EU state-aid approval in July 2025 (Source - Svk Kraftbalansen Höst 2025).

First procurement (autumn 2025, for winter 2025/26): FAILED. Every submitted bid exceeded the CONE (Cost of New Entry) price cap, so no capacity was contracted. Sweden entered winter 2025/26 with no contracted strategic reserve. Svk judged the ~1,000 MW normalvinter surplus meant no immediate adequacy threat — but the mechanism designed to backstop adequacy had simply not functioned.

Second procurement: succeeded. 350 MW was contracted for 15 January – 15 March 2026 from Mälarenergi Kraftvärme and Sydkraft Termisk Kraft. It was held at 100% availability and never activated — the market balanced without it (Source - Svk Kraftbalansen Vår 2026).

Uppdrag 3.1 — quantifying the volume gap and bridge measures

Svk’s June 2026 government assignment report Planering för ett leveranssäkert elsystem gives the most detailed quantitative picture of the gap between what ERAA requires and what can realistically be procured (Source - Svk Uppdrag 3.1 Leveranssäkert Elsystem (2026)):

Strategic reserve: need vs. procurable

The implication: on the ERAA calculation the 1h norm is not met with a ~350 MW reserve. Svk’s 3h norm and higher CONE are its own sensitivity assumptions (Ei’s review, planned autumn 2026, has proposed nothing yet), but under them procurable volume would exceed need — making the norm review a highly consequential near-term decision. Svk also judges the short-term real position better than ERAA shows, since ~800 MW of plants outside ERAA are available and överbelastningshantering (~1,300 MW) is assessed sufficient to hold risk under 1h.

Production resources outside ERAA2025 (just under 800 MW counted by Svk)

ERAA2025 does not include Karlshamnsverket block 3 (330 MW, återuppbyggnadstjänst contract 1 April 2026–1 April 2027) or Öresundsverket (448 MW, kept for ö-drift for five years from 2025), together just under 800 MW. Mothandelsreserv contracts (Ryaverket 300 MW, of which ~200 MW is already in ERAA, and Tekniska verken 45 MW) expire 1 January 2027 and are outside Svk’s sum, though 100+ MW may remain available.

Bridge measure: överbelastningshantering (1,350 MW)

Uppdrag 3.1 puts contracted volume at ca 1,300 MW to end-2029 (1,350 MW including just over 50 MW of earlier störningsreserv agreements), activatable year-round without advance warning as a last measure before MFK; it is not in the ERAA data and its purpose is sudden disturbances and production losses, not a substitute for a strategisk reserv. Svk judges it sufficient to keep expected LOLE under 1h in the short term even without a strategic reserve. Per Ei’s flaskhalsinkomster decision it is 1,309 MW across 20 resources at 10 plants, funded by 520 mnkr in 2027 (Source - Ei Beslut Flaskhalsinkomster 2027 (2026)).

Long-term volume trajectory

2033 looks the harder year in ERAA2025 (partly because more nuclear is assumed in 2035); reaching 1h in 2033 would require a much longer reserve endurance, which could make procurement harder.

Kapacitetsåtgärder — the TSO-DSO bridge concept

Svk and Vattenfall Eldistribution jointly developed a concept to procure temporary flexibility (production, storage, demand reduction) to bridge connection gaps before grid reinforcement is complete. A pilot in Västra Götalandsregionen was explored in 2025 but concluded without procurement — industrial plans shifted and Svk’s transmission expansion covered the need to 2035. The methods and tools developed remain available for future applications.

Stöd till icke-fossil flexibilitet

Regulation (EU) 2024/1747 requires member states to set national non-fossil flexibility targets (via the FNA process) and permits financial support mechanisms — auction-based procurement — for non-fossil flexibility that also contributes to adequacy. Svk coordinates the Swedish FNA, with results to government/EC/ACER in 2026. This is the EU-framework route to subsidising storage, demand response, and dispatchable renewables in ways that address both adequacy and the flexibility market simultaneously. (Flexibility Need Assessment)

Fixed-price contract concern

Svk identifies fixed-price electricity contracts (long-term contracts at predetermined prices, as used in neighboring countries) as an adequacy concern: when buyers lock in prices regardless of real-time scarcity, they remove demand elasticity from the system at precisely the scarcity hours. Demand response from neighboring countries is a component of the ERAA model — its erosion tightens Sweden’s effective adequacy envelope.

The CONE paradox — what the failed procurement actually revealed

The reason the first procurement failed is the most analytically important fact on this page. The Swedish CONE was calculated using household demand response as the reference technology — the cheapest theoretical new-entry resource. But the actual bidders were production-side units (generators, diesel/gas) with substantially higher true entry costs. All production bids exceeded a cap benchmarked to cheap flexibility, so nothing cleared (Source - Svk Kraftbalansen Höst 2025).

Read carefully, the failure is a measurement: the price cap is set to the flexibility the system wishes it had; the bids reveal the capacity the market actually offers. The gap between them — household DR’s notional cost vs. real production-side entry cost — is the precise size of Sweden’s flexibility shortfall expressed in price terms. The instrument meant to procure adequacy instead quantified the absence of cheap demand-side flexibility. Until demand response is abundant and cheap enough to bid below CONE, a reserve benchmarked to it will keep mispricing the resources that do exist. The revised CONE methodology is an open policy question and a tracked data gap.

Winter 2025/26 reality check — passed on margins, not on flexibility

The winter that followed was a genuine stress test, and how Sweden passed it matters (Source - Svk Kraftbalansen Vår 2026):

• Coldest in decades in Norrland; peak load 24,800 MWh/h on 12 January 2026 (up from 22,470 the prior year).
• Nuclear at 99% utilization at peak; wind at only 18% (vs 63% the year before) — low wind during the cold snap was the defining stress.
• Prices rose sharply: SE4 winter average 90 EUR/MWh, day-ahead peak 493 EUR/MWh (SE4, 19 Feb), and a balancing energy price of 7,363 EUR/MWh (7 Jan).
• Svk price-adjusted 11 quarter-hours of mFRR balancing energy under its temporary manual-correction analyses.

Sweden met demand through high nuclear output and Nordic interconnection (net export to Finland rose from 2.1 to 4.9 TWh, helped by the strengthened SE1 link from the Aurora line expansion) — not through demand flexibility or the strategic reserve (unactivated). The system held, but on the strength of baseload and imports, exactly the resources the adequacy outlook says are getting scarcer.

8 June 2026 — a reserve-exhaustion event in miniature

A second, sharper data point arrived outside winter. On 8 June 2026 two independent faults hit simultaneously — an SE2↔SE3 line de-energised when people climbed a transmission tower, and a −500 MW loss on the Swe-Pol Link (SE4↔Poland) — dropping frequency to 49.80 Hz. The response is the tell: Svk activated mFRR until every bid in SE3 and SE4 was exhausted, then had to start four gas turbines and lean on imports (200 MW each from Lithuania and Norway). mFRR activation peaked at 701 MW (SE3) + 255 MW (SE4) in the 15:30–15:45 quarter. (Source - Svk Driftstörningar 8 Juni 2026)

This is the missing-reserve thesis observed live, and on an ordinary June afternoon rather than a winter peak: when a real two-contingency event hit, the bid stack ran dry and the fallback was fossil gas turbines plus Baltic imports — not demand-side flexibility, which was absent from the response entirely. It also extends the winter signal of balancing-market fragility (the 11 price-adjusted quarter-hours above): here too the activations drove high mFRR and imbalance prices in SE3/SE4 that Svk will likely adjust down, on the stated principle that major grid incidents should not economically harm BRPs. That ex-post correction is the settlement-layer backstop (eSett, BSP and BRP Roles) — but it is a financial patch after the fact, not added physical reserve depth. The structural reading: SE3/SE4 reserve depth is thin enough that a non-extreme double fault can exhaust it, and the only fast firm capacity that answered was fossil and foreign.

Flexibility becomes the named mitigation

Against this backdrop, Svk’s spring 2026 report again names demand-side flexibility as an increasingly needed resource — consistent with earlier assessments, not a new framing — stressing that electricity need and the need for demand-side flexibility are both rising, and warning flexible resources (production and consumption) may not be built out in pace with need (Source - Svk Kraftbalansen Vår 2026).

The technical potential is large (Source - FlexAbility Delrapport 1 (2025), Demand Response):

On paper, this dwarfs the adequacy gap. The problem is not potential but deliverability for adequacy specifically.

Why flexibility is the right answer but not yet bankable for adequacy

Adequacy needs capacity that is firm at the scarcity hour — cold, still, dark winter evenings. Most flexibility revenue today comes from the wrong places for that:

• Batteries chase FCR, not scarcity. As the BESS business case shows, batteries earn ~86% from FCR capacity at 0.8 cycles/day — they are optimized for frequency containment, not for discharging through a multi-hour winter peak. Most are 1-hour FCR batteries, not the 2–4h+ duration adequacy needs.
• Local flex markets are thin and congestion-driven, not adequacy-driven (Why Swedish Local Flex Markets Are Thin — Structural Causes).
• Demand response at the scarcity hour is the least developed segment — the electric-boiler potential (975 MW) is outright policy-blocked by energy tax, and household DR (the CONE reference) is precisely what failed to materialize in the reserve auction.

The bankable bridge is emerging in two places, both narrow:

1. Svk capacity-service contracts (Hisingen/Stenungsund, ~2027) — multi-year bilateral payments that give BESS firm, financeable revenue for being available in scarcity, the first structured adequacy-adjacent revenue outside FCR and spot (Energy Storage › Svk capacity services for Hisingen/Stenungsund).
2. The strategic reserve itself — once the CONE methodology is fixed so demand-side and production-side resources can compete on comparable terms.

In other words: flexibility is the structurally correct adequacy answer, the technical potential is more than sufficient, but the contracting and revenue architecture that would make flexibility firm at the scarcity hour barely exists. That gap — not a shortage of MW on paper — is Sweden’s real adequacy problem.

And the supply side runs on a delay it cannot make up. Even where new firm capacity is economic, it has to clear permitting and grid connection first. Energiforsk’s NEPP lead-time study finds Sweden already carries an investeringsskuld — a shortfall of initiated projects relative to the 2045 scenario pathways — with permitting and grid connection as the binding bottlenecks, and even large lead-time improvements only partly closing the gap (Source - Energiforsk 2026-1185 Ledtider för Energiomställningen (2026)). This sharpens the adequacy picture: the bankability gap (no revenue architecture for scarcity-firm flexibility) compounds with an execution gap (capacity decided too late to arrive in time). Flexibility’s appeal as the adequacy bridge is partly because it has shorter lead times than the generation and grid it stands in for — see Long-Term Market Analysis › Planeringsscenario — Svk’s expanded planning mandate, Grid Capacity Utilization.

The NRAA pivot (end 2026)

Sweden’s first NRAA (National Resource Adequacy Assessment) is due end 2026 under the Art. 19e–19f EMD Reform framework, with ERAA 2025 as its baseline (Source - Svk Kraftbalansen Vår 2026, Flexibility Need Assessment). It is the document that will formally determine whether Sweden’s adequacy gap justifies a larger mechanism — and whether demand flexibility is credited quantitatively or only qualitatively. Closely linked: the Art. 19f non-fossil flexibility targets due ~2027 (six months after the first FNA), which would for the first time set national flexibility volume objectives (Electricity Market Design Reform 2024, the regulatory timeline).

What to watch

• Ei reliability norm review (autumn 2026) — whether the 1h LOLE norm is relaxed to ~3h; a change here is the single decision most likely to make the strategisk reserv viable without structural reform. At 3h, ~350 MW procurement could close the gap; at 1h, 750–1,100 MW is needed with only ~350 MW procurable.
• Revised CONE methodology — the single most important fix for the strategisk reserv mechanism; whether the reference is re-set so demand-side and production resources compete fairly.
• NRAA end-2026 — whether it quantifies a demand-flexibility contribution or leaves it qualitative; whether it triggers a move beyond the strategic reserve.
• Svk capacity-contract expansion beyond Hisingen/Stenungsund — the main route to firm, financeable flexibility for adequacy.
• Electric-boiler energy-tax reform — unlocks 975 MW of dispatchable demand reduction currently policy-blocked.
• mFRR manual-adjustment frequency and bid-stack exhaustion — a running indicator of how often the balancing market hits its scarcity limits (11 price-adjusted quarter-hours winter 2025/26; SE3/SE4 bids fully exhausted on 8 June 2026).
• Strategic-reserve volume trajectory — whether 350 MW grows as LOLE worsens toward 2033; this is the key annual procurement question.
• Non-fossil flexibility support mechanism — whether Sweden implements financial support (auction-based) under Regulation 2024/1747 to accelerate demand-side and storage resources that improve adequacy.

Data gaps

• Whether the strategic-reserve volume scales with the worsening LOLE trajectory (350 MW today vs 10.3 h/year LOLE in 2033)
• Realized contribution of demand response at the actual scarcity hour (cold, low-wind winter peaks) — potential is large but scarcity-hour firmness is unproven
• Targeted energy tax reform for district-heating electric boilers — general tax reduced Jan 2026 (−7.9 öre/kWh); specific elpanna/fjärrvärme targeted reform (~1.2 GW potential, ~500 MSEK/yr cost) remains unapproved as of August 2026; Energiföretagen’s specific ask (found via follow-up search) is a temporary cut from 43.9 to 0.6 öre/kWh for 2026–2031, which the organisation separately frames as unlocking up to ~1.6 GW of flexibility in its broader (not government-committed) proposal — still no government response found to either framing

Related pages

• Balancing Markets — strategic reserve, missing-money problem, ERAA LOLE, reserve volumes and prices
• Demand Response — demand-flexibility potentials and the scarcity-hour deliverability question
• Svenska kraftnät — Kraftbalansen reports; strategisk reserv operation; elberedskap
• Bidding Areas — SE3/SE4 as the adequacy-risk zones; north-south constraint
• The Swedish BESS Business Case — Revenue Stacking and the FCR Saturation Problem — why batteries optimize for FCR, not scarcity; capacity contracts as the adequacy bridge
• Why Swedish Local Flex Markets Are Thin — Structural Causes — why local flex is congestion- not adequacy-driven
• Flexibility Need Assessment — NRAA baseline; Art. 19f flexibility targets
• Electricity Market Design Reform 2024 — capacity-mechanism rules; Art. 19e–19f adequacy framework
• Energy Storage — battery duration and the Svk capacity-service contracts