Merit-Order Price Suppression from Weather-Dependent Generation

Weather-dependent generation (wind, solar) suppresses wholesale electricity prices by entering the merit order at near-zero marginal cost — the standard economic story. What’s less obvious, and what a 2026 Energiforsk/Profu empirical study of Sweden’s SE4 bidding area demonstrates concretely, is that the size of the effect scales nonlinearly with the volume added, and the mechanism driving that nonlinearity is specific to bidding areas that are only partially interconnected with larger neighbouring markets — exactly the situation SE4 (and by extension much of the wiki’s Sweden-focused congestion and export material) describes. (Source - Energiforsk Elprispåverkan Vindkraft Skåne SE4 (2026))

The mechanism: price-coupling regimes

A bidding area like SE4 does not have one merit-order curve; it effectively switches between three price-coupling regimes hour by hour, depending on whether its interconnectors to neighbouring areas are saturated:

1. Coupled to a larger market (for SE4: the continent — DK2, DE-LU, PL, LT aggregated). Price is set by that much larger market and barely moves in response to local supply changes — SE4 is a price-taker.
2. Isolated. A binding export constraint has decoupled the area from the larger market; price is instead set by the area’s own supply-demand balance including full export, and is highly sensitive to local supply.
3. Coupled to a neighbouring same-size area (for SE4: SE3). When the local price falls toward the neighbour’s level, the import constraint from that neighbour stops binding and the area re-couples to it — price sensitivity drops again, but not as far as regime 1.

Which regime holds in a given hour is set by whether the gap between allocated transfer capacity and actual flow on the relevant border falls below a threshold. This is why the same amount of added local generation has a different price effect depending on how full the export corridor already is: a first tranche of added wind mostly gets exported at the continental price with little local effect, but as more is added, hours where the export corridor is already full become more frequent, decoupling the area more often and exposing its own, more volatile balance.

Empirically, for SE4: with a 650 MW addition, the area decoupled from the continent (regime 2) in 9–11% of hours; doubling the addition to 1,300 MW pushed that to 22–25% of hours — the added generation doesn’t just lower the price within a fixed regime, it changes which regime applies, and it’s this regime-shifting that produces the nonlinear scaling (a doubled build more than doubling the price effect, not merely doubling it). (Source - Energiforsk Elprispåverkan Vindkraft Skåne SE4 (2026))

Nonlinear scaling — the empirical result

For SE4 (2024–2025 counterfactual), a Kriegers Flak-sized offshore wind addition (650 MW, ~2 TWh/yr) cut the load-weighted price by ~4%; doubling the addition (1,300 MW) cut it by ~9% — not double the effect, but more than double. The same volume of solar has a much smaller effect on the annual average (solar production is concentrated in a minority of hours, most of them already low-price), but a much larger effect within the hours it actually produces.

This has a direct implication for how to read a “we added X MW of renewables, prices should fall by Y%” claim: Y is not a fixed rate per MW. It depends on how close the area already is to saturating its export capacity, which is itself a function of how much weather-dependent capacity is already installed.

Capture rate and the cannibalization effect

The mechanism that suppresses the area’s average price also erodes the added generator’s own revenue, and it does so unevenly across technologies:

• Capture price (intjäningsförmåga): the generator’s own production-weighted average price — what it actually earns per MWh, as opposed to the arithmetic system average
• Capture rate (värdefaktor): capture price as a percentage of the arithmetic average price. A capture rate under 100% means the technology earns less than the average MWh sold in that market — a structural feature of weather-dependent generation, not a market failure, since it tends to produce more exactly when its own output (and everyone else’s like it) is pushing the price down

This is the cannibalization effect: the more of a weather-dependent technology is built, the more its own output overlaps with the hours it depresses, and the lower its own capture rate falls. In the SE4 study, wind’s capture rate declined from ~80% (650 MW) to ~74% (1,300 MW) in 2024; solar’s was already lower at 1 GW (~60%) and fell further at 2 GW (~54%). Wind degrades gradually because its production is spread across more hours with a flatter profile (more full-load hours); solar degrades faster because its output concentrates in a narrow midday/summer window that it saturates quickly.

This is also why a new offshore wind park can show a higher capture rate than an area’s existing onshore wind fleet: if the new park’s production profile (e.g. taken from a nearby offshore reference area) has more full-load hours and correlates less with the existing fleet’s own low-price hours, it captures more value even while adding to total wind capacity — profile shape matters as much as technology type.

Why this matters for investment: a technology’s capture price must exceed its production cost for the addition to be economic. Published Swedish cost estimates for new wind (~27–32 €/MWh onshore, ~36–45 €/MWh offshore, per Green Power Sweden) sit close to or below the SE4 study’s estimated capture prices for smaller additions, but a real August 2026 Danish offshore CfD auction cleared at ~67–73 €/MWh — well above both the cost estimate and the capture prices modelled here, for projects several years out. The gap illustrates that capture price and headline production cost don’t fully capture what it actually takes to secure investment; permitting, grid-connection cost, and risk allocation matter independently. Vattenfall’s September 2024 pause of the (separate) Swedish Kriegers Flak project despite holding all main permits, citing grid-connection cost as 25–30% of total project cost, is the wiki’s concrete anchor for this point. (Power Purchase Agreement › Relationship to two-way CfDs)

New load headroom

A region banking on wind/solar buildout to lower prices faces a corollary: the price cut isn’t permanent if new demand arrives to absorb it. The SE4 study quantifies this directly — a given volume of added wind creates room for roughly the same energy volume of new flat (baseload-shaped) demand before the load-weighted price rebounds to its pre-buildout level. A Kriegers Flak-sized 2.0–2.3 TWh/yr wind addition creates room for ~2.2–2.3 TWh/yr of new flat load; the doubled build, ~5.1–5.2 TWh/yr. Solar creates much less headroom per TWh added, because a flat new load can’t concentrate its consumption into solar’s narrow production window the way it can spread evenly across wind’s flatter profile.

This reframes “more renewables → lower prices” as “more renewables → more room for new load at the old price” — directly relevant to a region like Skåne trying to use wind/solar buildout to attract new industrial or data-centre connections without raising prices for existing customers. It is a market-level headroom only: neither the added generation nor the added load is checked here against actual local grid capacity to connect at a specific point, which is a separate, unaddressed constraint. (Grid Capacity Utilization, Distribution Network Development Plan)

Scope and limits

This is a short-term, “all else equal” counterfactual, not a forecast: it recalculates historical prices assuming more capacity had already existed, holding water values, fuel prices, and bidding behaviour fixed. It is explicitly not valid for arbitrarily large additions — the further the added volume from a “moderate” share of the area’s total energy and power volumes, the less reliable the linear supply-curve assumption becomes. Negative prices are excluded by construction (treated as zero-price spill), which understates real-world outcomes in an area that already sees negative-price hours. And it says nothing about whether the physical grid — as opposed to the market — can actually carry the added generation or load to and from a specific connection point.

Relationship to other wiki topics

• Bidding Areas — the price-coupling regimes are a previously undocumented mechanism for why SE4’s (and by extension other partially-interconnected areas’) price sensitivity to local supply varies hour to hour rather than following one fixed merit-order curve
• Skånes Effektkommission — same SE4 supply-adequacy problem and same commissioner (Region Skåne) as this body’s existing work in the wiki
• Power Purchase Agreement — the Danish CfD clearing prices and Vattenfall’s Kriegers Flak pause are concrete data points on the gap between capture price and what actually secures investment
• Electricity Export Restrictions / Congestion Management — export-capacity saturation is the physical event underlying regime-switching
• Nord Pool — the underlying day-ahead market and merit-order mechanism this page’s price-suppression effect operates within
• Flexibility Need Assessment / Distribution Network Development Plan — the new-load-headroom result is a market-level upper bound; whether it can actually be used depends on the same grid-capacity questions these frameworks quantify

Data gaps

• Whether a longer-horizon (~5–10 year) market-model analysis of the same SE4 buildout scenarios — recommended by the study’s own authors as a natural next step — has since been commissioned or published
• Whether this price-coupling-regime framework has been applied empirically to any other Swedish bidding area (SE1–SE3) or generalizes as stated