Consider a solar park with a co-located battery in Germany. In spring 2025, the investment case is strong. The battery's revenue plan combines shifting solar output to higher-priced hours with posi-tive aFRR (automatic Frequency Restoration Reserve) balancing capacity, which pays well at the time. The marketing contract assumes that the battery charges only from the solar park next door, maintaining eligibility for support under the Renewable Energy Sources Act (Erneuerbare-Energien-Gesetz, EEG). However, this also excludes negative aFRR, which would require charging from the grid.
Eighteen months later, the same asset faces a different reality. Prices for aFRR capacity have dropped significantly: enspired reports that aFRR up-capacity prices in September 2026 were 67% lower than a year earlier. The European Day-Ahead market now operates in 15-minute intervals. In October 2026, the Federal Network Agency (BNetzA) introduced “MiSpeL”, allowing storage to charge from the grid without losing EEG support. The first full year of operating data reveals curtailment frequency and actual battery cycling.
None of this invalidates the original business case. The focus has shifted, though. What matters now is less whether the investment decision was right, and more whether the asset is managed as effectively as current conditions allow.
In our last post, we argued that flexibility is now a key value driver for energy infrastructure, and that a project's ability to leverage flexibility depends on three factors: the market environment, the grid connection point, and the financing and offtake or commercialization structure. Since all three factors are dynamic and keep moving, a business case begins to age as soon as it is approved.
For us, the final investment decision (FID) is therefore no longer the end of the valuation process. Instead, it marks the beginning of an ongoing one.
The illusion of the one-time decision
For a long time, energy infrastructure followed a straightforward sequence: develop the business case, make the investment decision, secure financing, build the asset, and operate it. This approach rested on one premise: that key assumptions about revenues, costs, and operations would remain stable over long periods. Under the guaranteed tariffs discussed in our last post, typically paid for 20 years, they often did. The implicit understanding was that valuation was complete by financial close or, at the latest, commissioning, after which operations focused primarily on maintaining plant availability.
This premise no longer applies, as demonstrated by the evolution of power purchase agreements (PPAs). In a sense, the EEG feed-in tariff functioned as a PPA, offering stable payments for each kilowatt-hour fed into the grid, typically over 20 years, with the grid operator as the legally obligated counterparty. Market PPAs aim to provide similar long-term certainty, and policymakers have sought to simplify their use. For example, the EU's 2024 electricity market reform even asks the European Union Agency for the Cooperation of Energy Regulators (ACER) to consider voluntary PPA templates. Yet the market has moved towards more variety, not less. At the RE-Source conference in November 2025, participants noted increasing complexity in the PPA market. Contracts now must address cannibalization and regulatory uncertainty; simple pay-as-produced structures are often inadequate, and hybrid PPAs that combine generation and storage are the main trend, even though structuring is still in its early stages.
Meanwhile, deal volumes have fallen. Pexapark reports that PPA capacity in Europe dropped from 15.3 GW in 2024 to 13.1 GW in 2025, with Germany experiencing the sharpest decline among major markets. In contrast, nearly 12 GW and 23 GWh of battery storage were contracted under so-called flexibility purchase agreements and optimization deals in 2025, roughly three times the 2024 volume.
Why profitability has become a moving target
Revenues now come in from several markets at once, and their relative importance shifts more rapidly than a business plan anticipates. Batteries illustrate this clearly. In Germany, balancing capacity has historically provided most of their income, but since it is procured in fixed volumes, every additional competing battery pushes prices down and shifts value toward day-ahead and intraday trading. enspired's data shows that two-hour batteries earned about 20% less in September 2026 than a year earlier, despite higher wholesale prices. The company identifies falling aFRR prices as "the main structural headwind". Batteries financed on balancing revenues require new trading strategies within a few years. For solar, the challenge is timing: as ever more new solar parks generate power during the same hours, prices for these hours decline. In 2025, a kilowatt-hour of solar power in Germany was worth only about half the average spot price. A park’s earnings increasingly depend on whether it can shift its output, which is why there’s a battery next to our example solar park.
What an asset may be able to do is determined by its grid connection and regulatory framework, both of which can change after the investment decision. BNetzA's MiSpeL determination in October 2026 illustrates this. Until then, a battery co-located with a solar park retained its EEG support only if it charged exclusively from renewable power sources. Now, it may also charge from the grid without affecting the solar share’s eligibility, although until the end of September 2027, this requires approval from grid and metering operators. For batteries like the one in our example, that opens access to markets it was previously shut out of, such as negative aFRR, and turns a fixed design choice into an ongoing decision for the owner. The connection itself has become a scarce resource. In 2025, German grid operators received battery storage connection requests totaling, on paper, more than 150 times the operational capacity. Batteries connecting under flexible agreements accept, as discussed in our last post, caps on import, export, and ramp rates, and those caps set part of the revenue ceiling in the contract.
This creates a practical challenge during project development: technical design and revenue assumptions must be modeled under uncertainty and with limited data, making it difficult to compare scenarios built on different assumptions. Contract and marketing structures have also multiplied. As outlined in our last post, a battery can operate fully on the market with a guaranteed minimum revenue or under a tolling agreement, where a third party pays fixed fees and manages dispatch. Tolling agreements typically last several years and don’t need to cover the whole battery. For example, a bankable offer in early 2026 runs for five to seven years after commissioning and lets the owners toll all or a part of the capacity, such as 90/10 or 80/20 splits. Each structure shifts risk but also locks in decisions for years, including dispatch control, market participation and potential upside.
A second practical challenge is translating market, operating and contract risks into a business case that lenders will support. Banks are increasingly insisting on tolling agreements or revenue floors, as noted at the Battery Business & Development Forum 2026. Owners now secure most of their revenue upfront; enspired sees them aiming to secure 50 to 85% of projected merchant revenues. However, the contracted part of an asset can no longer adapt to market or regulatory changes. Owners trade potential upside for bankability, often making these decisions years in advance, when they know the least about how the asset will perform.
How well is my asset performing against its potential?
For owners, investors and lenders, this shift changes what they need to know: how the asset performs relative to its theoretical potential, which decision would add value under current conditions, and whether the marketing and trading strategy is optimal. The challenge lies in finding answers to those questions.
The process starts with scenario analysis rather than relying on a single case. Technical, commercial, and regulatory assumptions are mapped across multiple scenarios and translated into actionable recommendations, such as renegotiating a marketing contract, utilizing grid charging under MiSpeL, or adding capacity. These scenarios are only effective if investment, financing, hedging, and marketing decisions are based on the same numbers, rather than on separate models that each tell a different story.
Benchmarks are also important. Comparing targets to actual results, as well as performance across marketers and traders, shows whether deviations stem from market conditions or strategy, and whether the asset is being marketed as effectively as possible within its technical and market constraints.
Public reference points are starting to appear. Some optimizers now publish the realized portfolio revenues; for example, enspired reports monthly on the on-average best-performing asset, and the range across its German batteries. This gives owners a first outside reference to compare their own asset against.
Portfolio benchmarks are designed for comparability. For instance, enspired structures its figures to be "neutral to asset availability" and excludes days on which an asset was not operational for more than four hours. This approach enables comparisons across assets. However, owners assessing a single project also need to assess actual battery availability and the extent to the grid operator restricts it. Benchmarks neutral to availability cannot show the former, and we assume cross-portfolio benchmarks cannot reflect this because connection terms differ from site to site. Therefore, a fair assessment combines the asset's operating data with the market view.
The catch: managing this process manually is hardly doable. Static expert reports, spreadsheets, and one-off analyses only capture a single point in time, while underlying inputs keep changing. In the previous example alone, balancing prices, day-ahead market design, and rules for co-located storage all changed within eighteen months.
Outlook
A continuous evaluation and analysis layer is therefore needed. This layer should connect project development, financing, marketing, trading, and operations across the entire asset lifecycle, enabling ongoing reevaluations against real data and relevant benchmarks, and allowing for timely adjustments.
The value of an energy asset is increasingly set by the quality of the decisions made over its lifetime, long after its technical design is fixed.
That is why we seek solutions that turn valuation from a one-time exercise into an ongoing discipline, connecting decisions across the entire lifecycle of energy assets in increasingly dynamic energy systems.
© Header picture: EnBW / Uli Deck