The Price-of-Flexibility Ladder — Why Capital Flows to FCR, Not Local Markets

Central question: why does Swedish battery capital overwhelmingly flow into FCR rather than the local flexibility markets that DSOs spent years building? Two existing wiki syntheses each answer half of this. DSO Flexibility Valuation — Methods and Swedish Evidence shows what flexibility is worth to a DSO — up to 122,000 SEK/MWh for the first sliver of a subscription cut — but that is a buyer’s ceiling, not a number any battery owner has actually been paid. The Swedish BESS Business Case — Revenue Stacking and the FCR Saturation Problem shows FCR paying 86% of Flower’s revenue, but treats that as a starting point rather than asking what it is being compared against. What’s missing is DSO cost-avoidance benchmarks, local market clearing prices, and national balancing prices on one shared unit, ranked against each other rather than read in isolation.

Units, and the trap of comparing them directly

The four families of numbers in play are not natively comparable, and conflating them is the single most common way this kind of analysis goes wrong:

1. DSO cost-avoidance benchmarks (Vattenfall’s 0.27 vs 4–8 MSEK/MW/year): what a DSO estimates it would cost to buy flexibility availability instead of building copper. This is a ceiling on willingness to pay, not a market outcome; no one has actually paid 4–8 MSEK/MW/year for Swedish flexibility.
2. DSO marginal-value figures (FlexAbility’s 10,711–122,000 SEK/MWh): the value of avoiding a specific narrow congestion event (a subscription breach, a utnyttjandegrad penalty). These are per-MWh values for a handful of critical hours, not annualized capacity prices, and are not safe to multiply by 8,760 hours; the value collapses the moment the constraining hour has passed.
3. Local market clearing prices (SWITCH’s ~2,100–16,000 SEK/MWh activation in V2025/26 (Source - SWITCH CSV Exports V2025-26 (2026)), 236,000–377,600 SEK/MW/season availability): what FSPs have actually been paid, at whatever volume actually cleared.
4. National balancing capacity prices (FCR/aFRR/mFRR, EUR/MW): quoted per hour of the trading day; annualizing them (×8,760) gives a theoretical ceiling assuming continuous acceptance, which overstates what any single provider captures across a full year but is a fair basis for comparing tiers against each other, since the same overstatement applies uniformly across FCR, aFRR, and mFRR.

The ladder below keeps categories 1 and 2 separate from 3 and 4, and anchors 3 and 4 against one real, disclosed number, Flower’s March 2026 blended revenue, so the annualized ceilings aren’t left floating as pure theory.

The ladder

(Source - SWITCH Marknadsdata (info.switchmarket.se, 2026), Source - Svk Månadsrapport Balansmarknader Februari 2026, Source - Svk Månadsrapport Balansmarknader Juni 2026, Source - Flower Website (2024-2026), Source - Vattenfall Eldistribution Nätutvecklingsplan 2025-2034)

Currency basis: EUR figures are converted at the daily EUR/SEK rate Svenska kraftnät itself publishes for ancillary services, taken as the monthly mean of that rate for the month each price refers to (10.63 for February, 10.75 for March, 10.95 for June 2026) (Source - Svk Mimer Reserve Price Exports (2026)).

A necessary caveat on the annualized national-market figures: multiplying an hourly clearing price by 8,760 assumes the same bid is accepted every hour of the year at that exact mean price — no provider actually achieves this. The annualized FCR-D/aFRR/mFRR figures are best read as rank-ordering devices, not revenue forecasts. Flower’s row is different in kind: it is not an annualized price, it is realized revenue for an actual portfolio over an actual month, disclosed under MAR reporting obligations. It sits where a real, diversified BESS operator’s revenue actually lands relative to the single-product ceilings — below FCR-N and mFRR (both of which have real access barriers keeping most capacity out, discussed below), above the FCR-D floor, and above anything a battery could earn from local flex even in the best-served Swedish market.

A second, equally necessary caveat on the Södra Skåne local-flex figure: the 377,600 SEK/MW headline seasonal rate is not what an FSP nets. SWITCH’s own portal text states its Tillgänglighetspriser (availability prices) are quoted “exklusive påverkansfaktor” — excluding the impact factor — which the page reads as meaning the posted rate is scaled down at settlement (the portal quote concerns the hourly availability price; extending it to the seasonal rate is this wiki’s inference). No source states a specific numeric impact factor for Södra Skåne; this page uses a data-derived 0.56–0.625 range, computed from five independent ratios already published in the SWITCH portal’s own data (allocated/requested volumes, and seasonal-compensation and activation-price premiums across two seasons — full derivation in Source - SWITCH Marknadsdata (info.switchmarket.se, 2026) › Säsongstillgänglighet (scheduled seasonal availability) — Vinter 2025/2026). Applied to the headline rate, real FSP remuneration is ≈211,000–236,000 SEK/MW per contracted season, not the headline figure. The contract also runs a shorter window than the platform’s own “season” label suggests: SWITCH’s wrapper season spans Nov–Mar (5 months, covering the hourly Tillgänglighetsordrar product too), but E.ON’s LFM-p/seasonal-availability product family is designed around a 3-month contract window (Dec–Feb, core winter) for Säsongstillgänglighet specifically. Annualizing the corrected per-season range the same way this page annualizes the national-market rates (rate ÷ months in the real contract window, ×12) gives ≈846,000–944,000 SEK/MW/year. The remaining imprecision — a range rather than a point estimate — is structural, not a research gap waiting to close: DSOs don’t publish per-FSP impact factors, because doing so would reveal grid-topology and node-relation detail they treat as sensitive against foreign-adversary mapping.

Reading the gap: 69–76% at best, not 15–30×

The number DSOs cite — Vattenfall’s own 15–30× ratio between grid capex and flexibility availability cost — describes the gap between rows 1 and 8 of the table above: what grid investment costs the DSO versus what the DSO believes it would have to pay for guaranteed availability. That comparison has done real regulatory work (it is Vattenfall’s stated justification for treating flexibility markets as not viable, per Vattenfall vs E.ON — DSO Approaches to Flexibility › Same regulation, different conclusions), but it answers a different question than the one this page asks. It says nothing about what a battery owner actually receives, because no Swedish DSO has ever paid 4–8 MSEK/MW/year for flexibility — that figure is Vattenfall’s internal estimate of a hypothetical contract, not an observed price.

The number that actually explains capital allocation is the gap between rows 2 and 5: Södra Skåne’s best-case local-flex rate (~846,000–944,000 SEK/MW/year, accounting for its impact factor and true contract window) against Flower’s real, disclosed blended revenue (~1,235,000 SEK/MW/year). The single most active Swedish local market pays at most about 69–76% of what national balancing markets deliver in practice — a real gap, but a much smaller one than the raw 15–30× DSO figure suggests. Every other local market pays 236,000 SEK/MW per season at real, undiscounted value (they’re all radial grids with impact factor 1, unlike meshed Södra Skåne) — about 62% of Södra Skåne’s headline rate but equal to the top of its impact-factor-adjusted range, so ~944,000 SEK/MW/year on the same annualization — or nothing: 7 of 12 SWITCH markets awarded zero seasonal capacity in V2025/26, and the hourly availability-order side is thinner still — 5 of 11 markets awarded zero Tillgänglighetsordrar in the same season (Source - SWITCH Marknadsdata (info.switchmarket.se, 2026) › The thin market problem in practice).

A battery owner deciding where to point a fixed MW of capacity is not choosing between a 15–30× gap and nothing; that gap belongs to the DSO’s own investment calculus. The owner is choosing between a market that pays about 1.2 MSEK/MW/year with near-certain acceptance (FCR, given 4,510 MW of prequalified FCR-D up supply against a 547 MW need, bid-and-hold is close to a guaranteed clear) and a market that pays at most ~69–76% of that in the one location where it clears at all, with most markets (7 of 12) clearing no seasonal capacity at all — a smaller price gap than the headline local-flex rate implies, but the access gap (below) does most of the remaining work.

Why the higher rungs (FCR-N, mFRR, aFRR) don’t already dominate FCR

If FCR-N, aFRR, and mFRR up all annualize higher than FCR-D, why does Flower’s blended revenue sit at 1.23 MSEK/MW/year rather than closer to those ceilings? Access, not price: FCR-N’s entire requirement is only 224 MW, too small to absorb a large portfolio; mFRR and aFRR both carry a 1 MW minimum bid against FCR’s 0.1 MW, and aFRR is additionally gated to just 6 prequalified BSPs by its 5-minute activation requirement (detail in Swedish Balancing Market Prices and Volumes › Market supply side — BSP counts and concentration and The Swedish BESS Business Case — Revenue Stacking and the FCR Saturation Problem › The pivot markets — ranked by the volume arithmetic). FCR is not the top of the ladder; it is the rung every small battery can actually stand on. Local flex sits below even that accessible floor, for most markets and most hours.

Falsifiability — what would overturn this

This synthesis rests on two public, checkable datasets, and is deliberately built so the comparison can be re-run:

1. Svk’s public balancing data — the BSP register (which BSPs are authorised for which products) and the prequalified volumes by technology in the Balancing Market Outlook data update — lets the supply overhang be re-checked. The “near-certain acceptance” premise behind FCR’s accessibility advantage rests on an ~8× supply overhang (4,510 MW prequalified against ~547 MW need); if a future update shows that ratio falling below roughly 2×, bid-and-hold can no longer be assumed close to guaranteed, and the case for FCR as capital’s default destination should be re-examined.
2. Flower’s own monthly MAR-mandated trading disclosures are the anchor for the “actual, not theoretical” row in the ladder. If a later month shows the energy-market share of Flower’s revenue rising materially above 14% — the leading indicator The Swedish BESS Business Case — Revenue Stacking and the FCR Saturation Problem › Data gaps already flags as unmonitored — that would be evidence the pivot toward mFRR/arbitrage predicted elsewhere in the wiki is underway, and the ladder’s ranking of national-market rungs should be re-annualized against the new mix rather than the March 2026 snapshot used here.

A third, harder check would falsify the local-flex comparison specifically: if a Swedish local market other than Södra Skåne began clearing seasonal availability (impact-factor-adjusted) above ~1.23 MSEK/MW/year — matching or exceeding Flower’s realized blended rate — the “local flex is structurally below the accessible national floor” claim would need revision for that market. Nothing in the V2024/25 or V2025/26 SWITCH data shows this; Södra Skåne’s own rate would need to rise by roughly 31–46% (depending on which end of its own impact-factor range applies) to close the gap, a real possibility worth re-checking each season — and Södra Skåne is already the platform’s highest-clearing market by a wide margin, so a rise elsewhere is the more plausible route to this check tripping.

What is deliberately not listed as a gap

The Södra Skåne impact-factor range (0.56–0.625) and the 236,000 SEK/MW “typical market” rate for every other market are not listed as gaps. Södra Skåne is a meshed grid, where an FSP’s contracted MW only delivers a fraction of that MW as relief at the constrained point; every other SWITCH market is a radial (non-meshed) local grid, where the platform sets the impact factor to exactly 1 (no discount) (Source - SWITCH Marknadsdata (info.switchmarket.se, 2026) › Säsongstillgänglighet (scheduled seasonal availability) — Vinter 2025/2026). The per-FSP impact factor within Södra Skåne’s range is permanently non-public by design: disclosing it would reveal grid-topology and node-relation detail DSOs treat as sensitive against foreign-adversary mapping, so this is a structural precision limit, not a pending data point.

Related pages

DSO Flexibility Valuation — Methods and Swedish Evidence · The Swedish BESS Business Case — Revenue Stacking and the FCR Saturation Problem · Swedish Balancing Market Prices and Volumes · Vattenfall vs E.ON — DSO Approaches to Flexibility · Why Swedish Local Flex Markets Are Thin — Structural Causes · SWITCH · Flower · Flexibility Market