A 35% Revenue Drop in Romania Tests Merchant Storage Math

Romanian battery storage revenues fell by roughly 35% in August for both two-hour and four-hour systems, which Clean Horizon attributes in part to increased competition after new capacity including the Gura Ialomiței project came online.

Romania has just supplied European storage investors with something they have mostly had to model rather than observe: a month in which grid battery revenues fell sharply, immediately after new capacity entered the market. Revenues for both two-hour and four-hour systems dropped by roughly 35% in August, according to Clean Horizon, which points to increased competition following the commissioning of fresh capacity — including the Gura Ialomiței project — as one driver.

That is a single month in a small market, and single months in merchant power are noisy. But the direction and the mechanism matter more than the magnitude, because the mechanism is the one Wood Mackenzie has separately described as "looming" across Europe: revenue cannibalization, in which each new battery narrows the price spreads that every existing battery depends on.

Why adding batteries reduces what batteries earn

A merchant storage asset makes money from difference, not from level. It charges when power is cheap, discharges when power is expensive, and captures the gap. In most European markets it layers a second income stream on top of that: payments for frequency response and other ancillary services that help keep the grid stable second by second.

Both revenue streams shrink as capacity grows, and they shrink for slightly different reasons. Ancillary markets are volumetric and small — a system operator needs a fixed quantity of fast response, and once enough batteries are bidding for it, clearing prices fall hard. That is why ancillary saturation typically arrives first and arrives abruptly. Energy arbitrage is more durable, but it too erodes: batteries charging in the cheap hours lift those prices, and batteries discharging in the expensive hours cap them. The spread compresses from both ends.

Romania's revenue fall applying to two-hour and four-hour systems alike is the detail worth sitting with. Longer-duration assets are usually pitched as the defensive play — better positioned to ride broader daily price shapes rather than depending on short, sharp scarcity events. When both durations fall by a similar order, it suggests the pressure is not confined to the fastest, shortest products.

The arithmetic that makes a 35% month uncomfortable

Storage financing tends to assume a revenue stack that declines gradually over a project's life, with equity absorbing the downside in exchange for the upside of volatile years. A step change is harder to underwrite than a trend.

The recovery math is worth spelling out, purely as illustration rather than as a forecast. A 35% decline leaves revenue at 65% of its prior level. To get back to where it started, revenue would have to rise by roughly 53.8% from the new base — not 35%. Asymmetry of that kind is why merchant storage models are usually built on multi-year averages rather than trailing months, and why a single compressed month can still move a financing conversation: it changes the assumed floor, not just the assumed mean.

Nothing in the Romanian data tells us the drop is permanent. Summer months in Southeast Europe bring their own price shapes, and August is not a typical trading month anywhere. The point is that a structural cause — new competing capacity — has been identified alongside whatever seasonality is present, and structural causes do not reverse when the weather changes.

Which European markets sit next in the queue

Romania is instructive precisely because it is early in its buildout. Markets that are small relative to the size of individual new projects reprice fastest: one commissioned site can represent a meaningful share of total flexible capacity, so the marginal impact of each connection is large. As Energy News notes, the Romanian experience reads as a live instance of the dynamic Wood Mackenzie has flagged for the continent more broadly.

The sequencing logic for the rest of Europe follows the same variables, and investors can watch them directly rather than waiting for the revenue print:

• Pipeline density versus market size. Where the connection queue is large relative to peak demand, ancillary saturation arrives sooner.
• Ancillary market depth. Small frequency-response requirements saturate quickly; markets that have already moved most revenue into wholesale arbitrage have less left to lose.
• Interconnection. Better-connected markets import price volatility from neighbours, which slows spread compression. Peripheral markets compress faster.
• Renewables mix. High solar penetration creates the midday-to-evening spread batteries live on — but also attracts the batteries that erase it.
• Contracted versus merchant share. Where capacity markets or long-term tolling agreements underpin a large share of the fleet, revenue volatility is absorbed by counterparties rather than by equity.

On those tests, the markets most exposed are the smaller, less interconnected ones adding capacity quickly — the profile Romania itself fits. Larger markets are not immune; they simply take longer to show it, and the eventual adjustment tends to be gentler because the pipeline is spread across more competing revenue layers.

What changes for developers and for capital

The practical response to cannibalization has never been to stop building. It is to change what gets built and how it is paid for. Three shifts follow logically from a revenue print like Romania's.

First, duration. If short-duration economics are the first to compress, the incentive tilts toward longer systems that can trade broader daily shapes — although Romania's data, showing four-hour revenues falling alongside two-hour ones, is a caution against treating duration as a complete hedge.

Second, contracting. Tolling agreements, in which an offtaker pays a fixed fee for the right to dispatch the battery, transfer merchant risk to a party better able to hedge it. Expect more of them, on tighter terms, wherever merchant revenue has visibly wobbled.

Third, the cost of capital. Lenders size debt against the revenue they believe is contractually or statistically secure. A market that has just demonstrated a 35% monthly swing invites lower leverage, higher coverage ratios, or both — which raises the levelised cost of storage even if battery hardware keeps getting cheaper.

Wider markets were untroubled by any of this. The S&P 500, tracked by SPY, closed at $771.35, up 0.54% on the day, as of 20:00 GMT on 25 September 2026, with the Nasdaq 100 proxy QQQ at $744.50 and the Dow 30 proxy DIA at $517.49. Storage revenue compression in a single Southeast European market is not an equity-index event. It is a project-finance event, and those tend to show up in the terms of next year's deals rather than in today's close.

The number to watch next

The useful follow-up is not whether Romanian revenues rebound in the autumn — they may, on seasonality alone. It is whether they rebound to the pre-August level or settle visibly below it once further capacity connects. A partial recovery that plateaus under the old baseline would confirm the structural reading. A full recovery would suggest August was mostly weather and maintenance. Developers in every European market with a crowded connection queue have reason to watch which one it turns out to be.

Key facts

• Revenue decline: Romanian battery storage revenues fell roughly 35% in August
• Durations affected: Both two-hour and four-hour systems
• Cited cause: Increased competition after new capacity, including the Gura Ialomiței project, was commissioned (Clean Horizon)
• Benchmark close: S&P 500 proxy SPY at $771.35, +0.54%, as of 20:00 GMT, 25 Sep 2026

Frequently asked questions

What happened to Romanian battery storage revenues?

Revenues for grid-scale battery storage in Romania fell by roughly 35% in August, and the decline applied to both two-hour and four-hour systems. Clean Horizon attributes part of the fall to increased competition following the commissioning of new capacity in the market, including the Gura Ialomiței project. The figure covers a single month.

What is revenue cannibalization in battery storage?

Cannibalization is the process by which each newly connected battery reduces the earnings available to existing batteries. Storage makes money from price spreads and from ancillary grid services. More batteries lift cheap-hour prices, cap expensive-hour prices and drive down clearing prices for a fixed volume of ancillary services, compressing both revenue streams at once.

Why does it matter that four-hour systems fell too?

Longer-duration batteries are usually pitched as more defensive, since they trade broader daily price shapes instead of depending on short scarcity events. When four-hour revenues fall alongside two-hour revenues by a similar order, it suggests the pressure is not limited to the fastest, shortest products and that duration alone is not a complete hedge.

Has anyone else warned about this across Europe?

Yes. Wood Mackenzie has separately flagged revenue cannibalization as a looming issue for European battery storage markets. The Romanian August decline is notable because it is an observed instance of the dynamic rather than a projection, arriving in a market that is relatively early in its storage buildout.

Which European markets are most exposed next?

The markets most exposed are generally those that are small relative to the size of individual new projects, weakly interconnected with neighbours, and adding capacity quickly into shallow ancillary service markets. Larger, better-interconnected markets with more contracted capacity tend to compress more slowly because merchant risk is spread across more revenue layers.

How do developers respond to compressed storage revenues?

Typical responses include building longer-duration systems, signing tolling agreements in which an offtaker pays a fixed fee for dispatch rights, and accepting more conservative debt sizing. The last of these raises the levelised cost of storage even when hardware costs fall, because lenders size debt against revenue they consider secure.

Sources

• Romania’s Battery Revenue Crash Shows What’s Coming for the Rest of Europe — Energy News (H2)