Energiforsk Elprispåverkan Vindkraft Skåne SE4 (2026)

Mikael Odenberger & Viktor Walter (Profu), “Elprispåverkan av ökat utbud av vindkraft i Skåne/SE4” — Energiforsk rapport 2026:1213, september 2026. A short-term, empirical counterfactual market analysis commissioned by Region Skåne (Regional Utveckling), asking: how would SE4 electricity prices in 2024 and 2025 have looked if more weather-dependent generation (offshore/onshore wind, solar) had already been built?

Method

For every hour of 2024 and 2025, the authors recalculate the SE4 price as if a hypothetical new wind or solar plant had delivered power that hour, using the plant’s real production profile under actual weather conditions (offshore wind profile from observed DK2 production, scaled; onshore wind and solar from observed SE4 production, scaled). The added low-cost supply shifts the merit-order/supply curve; the price adjustment follows the curve for that hour’s price-coupling regime (see below). Price is never allowed to go negative — supply that would push it below zero is counted as spill instead, a simplification since SE4 did see real negative-price hours in 2024–2025 (attributed to support-scheme or contract structures that keep some generators producing even at negative prices).

Price-coupling regimes — the mechanism that determines how sensitive SE4’s price is to local supply changes in a given hour:

1. Coupled to the continent (DK2, DE-LU, PL, LT treated in aggregate) — price is set by a much larger market and barely responds to local supply changes
2. Isolated — a binding export constraint has decoupled SE4 from the continent; price is set by SE4’s own balance including full export, and is highly sensitive to local supply
3. Coupled to SE3 — when SE4’s price falls toward SE3’s level, the SE3→SE4 import constraint (the binding direction) stops binding and SE4 re-couples to SE3, again lowering sensitivity

Whether a border is “coupled” or “binding” is determined hour by hour from the gap between allocated transfer capacity and actual flow (a rolling 30-day 97.5th-percentile cap, since explicit NTC data was only available for the DE-LU–DK2 link; other borders are approximated from scheduled day-ahead exchange). A border counts as binding when the margin to that cap is under 100 MW.

Scenarios and headline results

Six build-out scenarios, each expressed as added capacity (MW) and added annual energy (TWh/yr):

Reference load-weighted price: 54.5 €/MWh (2024), 65.2 €/MWh (2025).

The scaling is nonlinear: doubling the offshore wind build more than doubles the price cut (from ~4% to ~9%), because a larger share of the added output pushes SE4 out of continental price-coupling more often (9–11% of hours at 650 MW, 22–25% at 1,300 MW) and pushes local prices further toward zero in favourable-weather hours, increasing spill (0.36 TWh spilled at 650 MW → 1.16 TWh at 1,300 MW in 2024) even as it lowers profitability for the added capacity itself.

Concentration in high-price hours: during winter weekday evenings (17:00–21:00, Dec–Feb), where SE4 averaged 88–99 €/MWh, the Kriegers Flak-sized park cut prices by 2.8–6.1 €/MWh (more than the annual average), while during summer midday hours (May–Aug, 10:00–15:00, ~18–19 €/MWh baseline) the same park moved price only ~0.2 €/MWh. Solar shows the mirror pattern: negligible effect on winter evenings, but a 2 GW build cuts summer midday prices by ~2.2–4.6 €/MWh — over a quarter of the price level in that window.

Capture rate and cannibalization

Värdefaktor (capture rate) — the added generator’s own production-weighted average price (intjäningsförmåga / capture price) as a share of the arithmetic average price:

Wind’s capture rate declines with volume (80%→74% at 2024 prices) but stays well above solar’s (60%→54%), because wind’s flatter, higher-full-load-hour profile overlaps less with its own price-depressing effect than solar’s concentrated midday output does. This is the cannibalization effect: the more of a weather-dependent technology is built, the harder its own output depresses the price in the hours it actually produces, directly eroding its own revenue. A smaller offshore park can even show a higher capture rate than the existing SE4 wind fleet, because the DK2-derived offshore profile has more full-load hours and correlates less with SE4’s existing low-price hours than the local onshore fleet does.

Cost context: Green Power Sweden cites production costs of ~27–32 €/MWh onshore and ~36–45 €/MWh offshore. Against that, this study’s onshore capture prices clear the cost range; the single Kriegers Flak-sized park sits at the top of the offshore cost range and the doubled build at the bottom. A real-world data point cuts the other way: Danish offshore wind cleared an August 2026 CfD auction (Nordsøen Midt, Hesselø — Vattenfall) at 504 and 542 DKK/MWh (≈67.4 and 72.5 €/MWh), well above both the industry cost estimate and this study’s capture prices — for projects entering service 2032, not directly comparable to today, but indicating the price needed to actually secure investment exceeds simple capture-price/LCOE comparisons. Vattenfall itself paused the (unrelated) Swedish Kriegers Flak project in September 2024 despite holding all main permits, citing grid-connection cost — stated at 25–30% of total project cost — as decisive.

New load headroom

A second counterfactual: how much new flat (baseload-shaped) demand could be added before the load-weighted price returns to its unmodified reference level? Roughly one-to-one in energy terms with the added wind production — a Kriegers Flak-sized park (2.0–2.3 TWh/yr) creates room for ~250–260 MW / 2.2–2.3 TWh/yr of new flat load; the doubled build, ~580–590 MW / 5.1–5.2 TWh/yr. Solar creates much less headroom per TWh added (only ~50–195 MW-equivalent for 1–2 GW), because its concentrated hours don’t overlap well with a flat load shape. The added supply is absorbed mostly by export (SE4 exports to the continent in >90% of hours in the dataset, with ~1,100 MW average headroom; a new 650 MW offshore wind park fits within that headroom ~90% of coupled hours) and secondarily by reduced import need from SE3 (a reduction in need, not in available import capacity — the interconnector itself is unchanged).

Explicit scope limit: this is a market/price-level analysis only. It does not check whether the local grid has the physical capacity to connect either the new generation or the new load at a given point — that is a separate condition the authors flag as outside scope.

Method limitations (as stated by the authors)

• Short-term, “all else equal” counterfactual — assumes water values, fuel prices and bidding behaviour are unaffected by the modelled change; not a forecast, and explicitly not valid for arbitrarily large additions (reliability degrades as the added volume grows relative to SE4’s total energy/power volumes)
• Negative prices excluded by construction (treated as spill at zero price), while SE4 in reality saw a material number of negative-price hours in 2024–2025
• Coupling-regime classification is approximated (rolling-percentile proxy for NTC on most borders; the 2025 shift to flow-based capacity allocation makes “capacity” itself harder to define; aggregating the continental side can mask transit constraints further out in the network)
• Repowered-onshore scenario applies today’s onshore production profile to the added capacity; real repowered turbines (taller, technology-improved) likely have a flatter profile than modelled

Relevance to existing wiki topics

• Merit-Order Price Suppression from Weather-Dependent Generation (new concept page) — primary target; this source supplies the empirical Swedish anchor for the nonlinear scaling, capture-rate/cannibalization, coupling-regime and load-headroom mechanisms
• Bidding Areas — the three price-coupling regimes are a concrete, previously undocumented mechanism for how SE4’s price sensitivity to local supply/demand actually varies hour to hour
• Skånes Effektkommission — same commissioner (Region Skåne) and same SE4 supply-adequacy problem framing as this body’s other work already in the wiki
• Power Purchase Agreement — the Danish Nordsøen Midt/Hesselø CfD clearing prices and the Vattenfall Kriegers Flak pause are concrete data points for the existing two-way CfD and PPA-economics discussion
• Electricity Export Restrictions / Congestion Management — export-capacity utilization and the coupling/decoupling dynamic bear on both pages’ treatment of SE4’s price formation

Data gaps

• Whether a follow-up longer-horizon analysis (the authors’ own recommended next step, modelling ~5–10 years out with a full market model) has been commissioned or published
• Grid (not market) capacity to actually connect either the added generation or the new load at specific SE4 locations — explicitly out of scope here, and not yet covered by any wiki page combining this study’s price case with a network-capacity check