Germany’s grid-scale battery storage capacity stood at over 1.5 GW across 207 commissioned projects as of June 2025 and was projected to reach 2.5 GW by year’s end. Grid connection applications submitted to German operators represent a further pipeline exceeding 200 GW of proposed BESS capacity. Against that backdrop, the aFRR capacity market, which has become Germany’s largest balancing market by TSO expenditure, is absorbing a structural shift in its supplier base while simultaneously being reshaped by the solar generation patterns it was not originally designed around. Data covering January 2025 through May 2026 makes the combined effect legible in price terms for the first time with sufficient granularity.
Germany’s aFRR capacity market operates on a pay-as-bid basis, structured into daily blocks of four hours that produce six separate auctions per day in each direction. Upward and downward capacities are procured separately, reflecting their different physical and economic profiles. The market runs under the ALPACA framework, which enables capacity sharing with Austria and the Czech Republic, though the price heatmaps examined here pertain specifically to German auction results. Each auction produces both an average price, reflecting what capacity providers are actually paid across selected bids, and a marginal price, reflecting the highest bid that cleared.
The Solar Imprint on Downward Prices
The directional asymmetry in German aFRR capacity pricing is now pronounced enough to constitute a structural feature rather than a seasonal pattern. Upward capacity prices track demand: they are highest during morning and evening peak load hours and lowest at night, a profile consistent with the opportunity costs faced by flexible thermal assets and pumped hydro that form the supply stack. Downward capacity pricing follows a different logic entirely. Prices in the downward direction are materially higher during daylight hours with high solar output, reflecting the value placed on the ability to reduce net generation into the grid when solar is pushing the system toward oversupply.
This solar imprint on downward balancing prices is a direct consequence of Germany’s generation mix. In 2025, Germany recorded almost 575 hours of negative day-ahead electricity prices, a metric that captures the frequency with which solar generation exceeded dispatchable demand. Each of those hours creates a condition in which downward balancing capacity, the ability to reduce injection on command, carries enhanced value to the system operator. The price signal in the downward aFRR market is therefore functioning as intended: it is pricing the scarcity of flexible capacity to absorb renewable surplus. A secondary pattern is visible at night in winter, where slightly elevated prices in both directions reflect the higher wind generation during those periods and the corresponding increase in variability that the reserve stack must accommodate.
Batteries, Price Compression, and the 2026 Divergence
The 2026 price data, though not yet sufficient to establish statistical significance over a full year, shows average aFRR capacity prices running modestly below their 2025 equivalents across several time blocks. The most plausible candidate explanation is the accelerating penetration of large-scale battery storage in the aFRR segment. Batteries are structurally well suited to aFRR provision: they respond within the required timeframe, can cycle multiple times daily across the six four-hour auction blocks, and carry no fuel cost that would set a price floor in the way that a gas peaker’s opportunity cost does.
As battery capacity competing for aFRR contracts grows, the marginal cost of providing aFRR in normal conditions falls. The ancillary services market, combining FCR and aFRR, currently stands at roughly 4 GW of prequalified capacity, and forward projections indicate saturation within two to three years at current installation rates. One forecast places ancillary services at 57% of German BESS revenues during the 2025 summer months but projects a decline to roughly 5% of revenues by 2030 as wholesale arbitrage becomes dominant. A 2-hour BESS system that generates approximately €240,000 per MW per year in the near term is projected to earn around €115,000 per MW per year by 2030 as market compression takes hold.
The implications for aFRR pricing are consequential. If batteries erode capacity prices in normal market conditions, the baseline cost of procuring aFRR falls, which is efficient from a system cost perspective. But the distribution of outcomes is becoming more skewed: typical days show average and marginal prices converging, while extreme days show increasingly large divergences between the two.
Extreme Price Events and the Pay-as-Bid Structure
The German pay-as-bid mechanism creates a dynamic that becomes particularly significant during peak price events. On a normal day, the merit order for aFRR capacity is relatively flat: average and marginal prices sit close together, and the difference between pay-as-bid and a hypothetical pay-as-cleared outcome is modest. On an extreme day, the merit order becomes sharply convex, with prices rising steeply at the top end of selected bids. On September 16, 2025, a single aFRR upward capacity bid cleared at €3,784 per MW per hour, an outlier but not without precedent in European balancing markets: a comparable event in the Netherlands produced an aFRR upward bid clearing at €1,789 per MW per hour for a four-hour block.
Under pay-as-bid, the providers whose bids are selected receive exactly what they bid rather than the marginal clearing price. On extreme days, this means that bidders who correctly anticipated the price spike and submitted high bids capture a large premium, while those whose bids were conservative relative to the eventual cleared price leave substantial revenue on the table. Accurately forecasting which four-hour blocks will produce extreme price events is, by the nature of those events, extremely difficult. They tend to arise from coincident scarcity conditions, unusual demand patterns, or unexpected generation shortfalls, none of which are predictable with the precision that strategic bidding would require.
The contrast with the pay-as-cleared mechanism used in PICASSO for aFRR energy exchange is instructive. Under pay-as-cleared, all selected providers receive the marginal clearing price, eliminating the information rent that accrues to accurate bidders and redistributing it across the supply stack. Germany’s choice to retain pay-as-bid in the capacity layer while participating in pay-as-cleared energy exchange through PICASSO creates a split incentive structure: capacity procurement rewards bidding accuracy and strategic information asymmetries, while energy activation is settled on a uniform marginal basis. Whether this design remains optimal as the supplier base shifts toward batteries, which have lower marginal costs and potentially different strategic bidding behaviour than the thermal assets that shaped the original auction design, is a question the data is beginning to pose, even if the market rules have not yet engaged with it.
What the Pattern Suggests About Market Evolution
The structural trends in German aFRR pricing point toward a market in transition between two equilibria. In the first, which characterised the 2023 to 2025 period, conventional flexible assets dominate supply, prices are anchored by their opportunity costs, and solar affects the downward direction but does not yet restructure the market as a whole. In the second, which the 2026 data hints at early stages of, large-scale batteries are numerous enough to compress normal-condition prices, solar’s role in the downward price signal is fully embedded in bidding strategies, and extreme price events become the primary source of above-average returns for providers with the forecasting capability to anticipate them.
Germany’s 15-minute market time unit reform, effective from September 2025, adds granularity to intraday trading that will affect how BESS operators stack revenues across aFRR capacity, aFRR energy, and wholesale arbitrage. Finer time resolution in the intraday market increases the value of assets that can switch between market participation modes quickly, which advantages batteries over slower thermal assets. The combined effect of MTU reform, ALPACA capacity sharing, continued BESS installation, and the solar-driven asymmetry in downward pricing is a market that will look materially different by 2028 from what it looks like today, with the price heatmaps accumulating month by month as the most direct empirical record of the transition in progress.
