Helios CapitalA modern large-scale battery storage system earns its money not in one but in three different electricity markets — and this is not a technical detail but structurally relevant to the risk profile of a direct investment. Day-ahead arbitrage, intraday response and balancing power have different drivers, different volatilities and different entry barriers. A storage system that is active in all three markets simultaneously smooths its revenue curve across hours, days and years.
This article explains each of the three markets at the minimum necessary depth — without energy jargon, but with the orders of magnitude relevant to investment analysis. How high a storage system's total revenues can be is covered in Battery storage returns: where the revenue comes from — and what is realistic; here the focus is on how they arise. At the end stands the central question for direct investments: which revenue streams are realistic in a concrete project — and how reliable are the numbers you are calculating with?
Three markets, three logics
The day-ahead market is the daily auction at the EPEX SPOT power exchange: by midday, electricity is traded for every quarter-hour of the following day — this is where the price most people know as the “exchange electricity price” is set. The intraday market is the continuous trading that follows: positions can still be bought and sold until five minutes before delivery within a system's own control zone. And balancing power is not a trading market in the narrow sense but a tender run by the transmission system operators: whoever wins the auction holds capacity ready to stabilise the grid frequency.
The revenue logic differs fundamentally: in the day-ahead market a storage system earns from predictable price differences, in the intraday market from short-term forecast errors and price swings, and in balancing power from holding capacity ready — partly regardless of whether it is ever called. For a battery storage system, access to all three markets practically always runs through a direct marketer (Direktvermarkter): a specialised trading house that steers the system across the markets algorithmically, around the clock.
How does a battery storage system earn money in the day-ahead market?
In the day-ahead market, a battery storage system earns the price difference between the cheapest and the most expensive hours of the following day: it charges in the midday hours when high solar feed-in depresses prices, and discharges in the evening hours when demand is high and the sun is gone. This difference — the “spread” — is the storage system's base compensation. The average daily spread in the German day-ahead market was around €117/MWh in 2024 and rose to around €130/MWh in 2025; as a typical range over recent years, €50–150/MWh is a robust order of magnitude.
The growing share of negative prices is remarkable: in 2024 Germany saw 457 hours of negative day-ahead prices, in 2025 already 573. In these hours a storage system is paid to charge — it absorbs surplus solar and wind power that nobody else will take. What is a problem for the power system is a driver for storage economics: the more renewables push into the grid without flexibility, the wider the gap between cheap and expensive hours opens. The drivers of the spread are accordingly weather (solar and wind feed-in), load and the availability of the remaining power plant fleet.
What happens in the intraday market — and why do batteries have an edge there?
The intraday market corrects the previous day's forecast errors: if the wind blows weaker than expected or a power plant fails, traders have to adjust their day-ahead positions at short notice — and pay for it. Trading is continuous, until five minutes before delivery within a system's own control zone, in hourly and quarter-hourly contracts. Volatility is structurally higher than in the day-ahead market: in 2024 the average maximum daily spread in the quarter-hourly products was around €184/MWh — a good half more than in day-ahead trading.
For batteries this market is structurally ideal: no other asset can switch between charging and discharging within seconds and thereby capture short-lived price spikes in individual quarter-hours. Quarter-hourly prices also follow a recurring sawtooth pattern, because conventional power plants run in full-hour blocks — precisely these small, predictable swings are the playing field of algorithmic storage marketing. And the market is growing: intraday volume at EPEX SPOT rose from 176 TWh (2023) via 215 TWh (2024) to 241 TWh (2025) — more liquidity, more trading opportunities.
What is balancing power — and why are battery storage systems ideal for it?
Balancing power is the reserve with which the transmission system operators keep the grid frequency stable at 50 hertz: if generation deviates from consumption, capacity has to be added or removed within seconds to minutes. There are three tiers — frequency containment reserve (FCR) responds within seconds, automatic frequency restoration reserve (aFRR) within five minutes, and the manual reserve (mFRR) in quarter-hours. Battery storage systems respond in milliseconds, making them technically the ideal balancing asset. Compensation is two-tiered: the capacity price already pays for holding the capacity ready, the energy price additionally pays for the actual call.
The flip side: these markets are small — and they are filling up fast. German FCR demand is only around 584 MW, while the battery capacity prequalified for it, at about 1.35 GW, is more than double that. In the aFRR market (demand around 2 GW), prequalified battery capacity more than doubled between January 2025 and January 2026. Anyone investing in a storage system today should therefore understand balancing power for what it is: a currently high-yielding but limited revenue source whose prices come under pressure with every newly connected storage system. The entry barrier is real at the same time: without prequalification with the transmission system operator — a technical verification process — no storage system participates in these markets.
What is revenue stacking?
Revenue stacking means marketing a storage system across several markets in parallel: the direct marketer decides algorithmically — for every quarter-hour — where the capacity earns the most: hold balancing capacity tonight, charge negatively priced solar power tomorrow at noon, discharge into the evening peak and capture intraday swings in between. This includes the complete operational management: state-of-charge planning, schedule management, prequalification and 24/7 trading. How this ongoing marketing is monitored after the purchase is described in What happens after closing: reporting, asset management and why a partner is not a broker.
The effect is substantial: market analyses and real trading data show that multi-market optimisation earns roughly 40 to 90 percent more, depending on the benchmark and strategy, than a storage system serving only a single market. At the same time, the mix is shifting: in 2026 around 55 percent of storage revenues still come from the balancing markets — by 2030, market analysts expect the lion's share (up to 95 percent) to come from day-ahead and intraday trading. Revenue stacking is therefore not an optional extra but the precondition for a storage system to ride this shift instead of depending on a single market.
| Market | Role in the revenue mix | Volatility | Entry barrier |
|---|---|---|---|
| Day-ahead | backbone of revenues, share growing to 2030 | medium | low — auction access via the direct marketer |
| Intraday | growing; fine-tuning and additional revenues | high | medium–high — 24/7 trading and algorithms required |
| FCR (primary reserve) | former revenue anchor, saturated today | low, prices falling | high — prequalification with the TSO |
| aFRR / mFRR | high-yielding today, saturation foreseeable | medium–high | high — prequalification with the TSO |
How will the revenue markets develop to 2030?
Two opposing forces determine the revenue outlook. The first widens the spreads: continued solar and wind expansion and the coal phase-out remove dispatchable generation from the system and increase price swings — visible in the rise of the average daily spread from €117 to €130/MWh and of negative-price hours from 457 to 573 between 2024 and 2025. The second force works against it: every newly connected storage system smooths exactly the price peaks all storage systems live on (“cannibalisation”). Scenario analyses show that if significantly more storage capacity comes online than expected, day-ahead revenues per megawatt can fall noticeably by 2030; the small balancing markets are already filling up over the next few years.
How fast this happens is the real unknown. The grid-connection queues of the German grid operators contain requests for more than 700 GW of storage capacity — only around 2.5 GW are connected so far, and experience shows only a fraction of requests get built. Expansion forecasts for 2030 range from roughly 100 to 170 GWh of storage capacity depending on the scenario. For investors, a simple rule follows: a serious revenue forecast does not extrapolate today's spreads, but calculates ranges — and discloses which market environment triggers which scenario.
What this means for direct investments
For assessing a concrete investment, this means: what matters is not how high revenues were last quarter, but how the project is positioned across the three markets over its lifetime. A robust calculation shows best, mid and worst case across all revenue streams with named assumptions — and in particular does not simply extrapolate today's still-attractive balancing revenues over 15 years. Which risks stand behind the revenue assumptions and how they are structurally addressed is covered in Risks in BESS direct investments — and how they are structurally addressed. The structure of the project itself also changes the revenue mix — for instance, whether the storage system stands alone or shares its grid connection with a solar plant.
The marketing model also belongs under scrutiny, because it distributes the market risk: under tolling, the storage system receives a fixed fee per megawatt and the marketer bears the market risk; under a floor model, a revenue floor covers the downside while the upside is shared; in the merchant model, the project bears the full market risk in exchange for a revenue share paid to the marketer — with conditions always negotiated project by project. Questions you should ask any provider:
- Who is the direct marketer, and in which markets (day-ahead, intraday, FCR, aFRR) is the storage system actually prequalified and active?
- Which compensation model applies — tolling, floor or merchant — and who bears the market risk, and to what extent?
- How strongly does the revenue forecast depend on today's balancing power prices — and what happens in the model once these markets are saturated?
- Are revenue sensitivities (best/mid/worst) disclosed with named assumptions — or just an average?
These questions double as a seriousness test: a provider who cannot or will not answer them precisely either does not have the marketing under control or shies away from transparency. Further warning signs and due-diligence questions are collected in How to tell a trustworthy provider of energy direct investments.
What the revenue mechanics look like in a concrete project — which direct marketer, which compensation model, which sensitivities — is what we discuss in a no-obligation initial consultation: based on real project figures, with named assumptions instead of a marketing number.