A solar park after 20 years: continued operation, repowering or decommissioning?

Anyone reviewing a direct investment in a solar park or battery storage facility soon runs into the number 20: 20 years of EEG remuneration, a 20-year tax depreciation period for photovoltaics, often a 20-year financial model. From there it is a short step to the mistaken idea that everything stops after 20 years. In reality, several clocks run in parallel in a solar park, and they stop at different times. Keeping those clocks apart makes project calculations far easier to judge: what is secured, what is an assumption, and what is a hidden reserve?

How long does a solar park really run?

Longer than 20 years. The 20 years are a subsidy and tax deadline, not a technical one. In a project calculation, four different clocks are running, and they should not be confused: tax depreciation, the subsidy or privilege window, technical lifetime, and the contract terms of the land lease and the participation itself.

The technical clock runs longest. Wafer-based solar modules age slowly: a Fraunhofer ISE measurement series across 44 quality-assured plants in Germany found an average degradation of nominal power of around 0.15 percent per year, while a broad NREL review of almost 2,000 published measurements puts the median at 0.5 percent per year. Sound calculations use the more conservative figure. Even then, a module still delivers close to 90 percent of its initial output after 25 years, and manufacturers typically warrant no more than 10 to 15 percent performance loss over 25 to 30 years. Battery storage ages faster and depends far more on how it is operated; how degradation, warranties and augmentation work there is covered in The battery as a real asset: lifespan, degradation and warranties of a grid-scale storage system.

What ends after 20 years, and what continues?

The subsidy ends, not the operation. Under Sec. 25 of the German Renewable Energy Sources Act (EEG), market premiums and feed-in payments run for 20 years; for plants with a statutory applicable value the period extends to 31 December of the twentieth year. After that, the payment claim expires and nothing else: the plant remains permitted, the grid connection stays in place, and the electricity may still be sold, just without a state-secured minimum price.

For small installations the legislator created a transitional solution: so-called post-subsidy installations still receive a follow-on feed-in payment until the end of 2032. Under the statutory definition, however, only installations up to 100 kilowatts qualify (Sec. 3 no. 3a EEG), which means the typical rooftop system. A megawatt-scale solar park does not fall under it. After the subsidy ends it moves into merchant marketing under Sec. 21a EEG: the electricity is sold on the exchange via a direct marketer or secured through a power purchase agreement; how such contracts are built is explained in Understanding PPAs: how power purchase agreements make solar park revenues predictable.

For battery storage the question presents itself differently, because it lives on market revenues rather than EEG payments from day one. Its most important privilege is the exemption of charging power from grid fees, and that too is designed to run for 20 years from commissioning (Sec. 118 (6) EnWG). Since the cells are designed for 15 to 20 operating years, this window covers the usual analysis period; it only becomes critical for models that reach well beyond it. The details are in Grid fees for battery storage: the exemption until 2029 and the AgNeS reform.

What are the three paths at the end of the subsidy period?

Continued operation, repowering or decommissioning. Which path wins economically depends on the condition of the plant, on power prices and above all on what the land lease and the permit allow.

In practice this is not a decision taken in year 20 but a gradual process: operators keep comparing whether the next major repair (say, an inverter replacement) is still worth investing into continued operation or becomes the occasion for repowering. For investors in a direct participation, the point is less which path the model assumes and more whether the provider keeps all three open.

What does repowering mean in practice?

Repowering means the plant technology is replaced with current equipment while the site stays. The value lies less in the old modules than in what exists around them: secured land, a grid connection with feed-in capacity, access rights, an established operation. Because modern modules deliver considerably more output per square metre than the technology of 20 years ago, significantly more capacity can be installed on the same site; the grid connection, the scarcest factor in many regions, is already there.

Two points are often misunderstood. First, swapping modules does not automatically trigger a new subsidy: under Sec. 3 no. 30 EEG, replacing the generator or other technical components does not change the plant's commissioning date. New EEG remuneration is only available through the regular route, meaning a new award in the auctions or a power purchase agreement at new-build terms. Second, the permitting situation is not trivial: whether repowering counts as a mere modification or needs a new procedure is not regulated uniformly in German building law; a 2025 study commissioned by the Federal Environment Agency explicitly calls for clearer administrative rules. A provider who prices repowering in as a certain value should be able to answer both points.

For tax purposes, repowering is at its core a new investment: the new components are capitalised and depreciated over their useful life, and under the conditions of Sec. 7g of the German Income Tax Act a new investment deduction amount may come into play. How that mechanism works is explained in The Investitionsabzugsbetrag explained simply: how the IAB works; the specifics belong with your tax advisor.

Who pays for decommissioning, and what is left of the site?

The operator pays, and the site becomes free again. For solar parks permitted as privileged projects in undesignated outlying areas (for instance along motorways and railways or as agrivoltaics under Sec. 35 (1) no. 8b and no. 9 of the German Federal Building Code), the law already requires, as a condition of the permit, a binding declaration to dismantle the installation after final abandonment of use and to remove soil sealing (Sec. 35 (5) sentence 2 BauGB); the authority is to secure compliance, for instance through a land charge. Most large ground-mounted parks, however, are built on the basis of a municipal development plan. There the decommissioning duty is regularly anchored in an urban development contract under Sec. 11 BauGB and backed by security, usually a bank guarantee whose amount the municipality sets.

How high decommissioning costs will actually be cannot be put into euros credibly today; even the study commissioned by the Federal Environment Agency recommends first creating a standardised expert basis so that security amounts can be determined consistently nationwide. Two things can be said, though. First, a solar park is not hazardous waste but mostly a store of raw materials: steel, aluminium, copper and glass whose recovery finances part of the decommissioning. Second, disposal of the modules is regulated: under the German Electrical and Electronic Equipment Act they count as large equipment, manufacturers must take them back, and their category is subject to minimum quotas of 85 percent recovery and 80 percent recycling (Sec. 22 ElektroG). For battery storage, the EU Battery Regulation sets its own rising recovery quotas; what happens at the end of a storage system's life is described in The battery as a real asset: lifespan, degradation and warranties of a grid-scale storage system.

For judging a project, the structure therefore matters more than the exact cost figure: is a decommissioning obligation fixed contractually, is it backed by sufficient security, and does decommissioning appear as a cost item in the calculation? A model that simply stops at zero at the end of the term has a gap.

Why is the land lease the real lever?

Because it decides who owns the options after year 20. Continued operation and repowering both require the site to be secured beyond the end of the subsidy. If the lease ends after 20 years, the operator negotiates at the worst possible moment: the landowner knows that the grid connection and infrastructure are tied to this site, and can price that in. This is why site agreements for solar parks are usually concluded beyond the subsidy period, often with staggered extension options reaching 30 years and more.

German civil law sets a little-known limit here: a lease concluded for longer than 30 years can be terminated extraordinarily by either party after 30 years have elapsed (Sec. 544 of the German Civil Code, applied to leases via Sec. 581 (2)). Very long fixed terms are therefore worth less than they sound; more robust is a base term with extension options plus security in the land register, typically a restricted personal easement that preserves the right of use even if the land changes owner. Well-negotiated contracts also state explicitly whether repowering is covered by the lease and how decommissioning is handled at the end.

How does residual value belong in the return calculation?

Conservatively or not at all. After the subsidy ends, revenues come from the market value of solar power, published monthly by the transmission system operators, or from follow-on PPAs; neither can be forecast credibly two decades in advance, especially as growing solar volumes depress market values in sunny hours. How this already shows up in hours of negative prices is described in Negative electricity prices: what they mean for solar and storage investors. A robust calculation therefore lets the participation pay off within the secured subsidy period and treats the years after that as upside with cautious price assumptions, not as a load-bearing pillar of the return.

Residual value also has a second, tax-related twist: once depreciation is exhausted, hardly any book value stands against the proceeds. If the plant or the participation is sold at that point, the sale price is essentially taxable profit; the mechanics are explained in After the IAB: How the ongoing returns and the sale of a direct investment are taxed. The residual value also plays a role in the profit-intent test, because the total-profit forecast needs a plausible end-of-term recovery scenario; details are in Liebhaberei and profit intention: when the tax office cancels the tax lever. Those who do not want to hold until the end of the term will find the exit routes in Selling a direct investment early: how liquid an energy direct investment really is.

Which questions should you put to any provider?

How a provider handles the end of the term says a lot about the quality of a project. These questions belong in every conversation:

• How long does the land lease run, which extension options exist, and is the right of use secured through an easement in the land register?
• Does the lease explicitly cover a later repowering?
• What residual value and which revenue assumptions does the model use after EEG remuneration ends?
• Does the participation still pay off if the years after the subsidy are set to zero?
• What decommissioning obligation exists, how much security backs it, and is decommissioning a cost item in the model?
• Which inverter replacement cycles and other reinvestments are planned?
• For storage: how far does the grid-fee exemption under Sec. 118 (6) EnWG reach relative to the model horizon, and is augmentation planned?
• Which degradation assumption is used in the yield report, and is it more conservative than the manufacturer's warranty? How to read such a report as a whole is shown in Reading a yield assessment: what P50 and P90 mean for a solar park direct investment.
• Who looks after the plant in year 15 and in year 25, and what does that cost?

How we check end-of-term and residual value in our project review

We review every project before presenting it to investors, and the end of the term is a review block of its own: lease term and options relative to the model horizon, security in the land register, decommissioning arrangements and their backing, reinvestment assumptions for inverters and equipment, and the question of whether the return holds up without residual-value fantasy. What generally distinguishes a diligent provider is summarised in How to tell a trustworthy provider of energy direct investments; what ongoing support looks like over the decades after signing is shown in What happens after closing: reporting, asset management and why a partner is not a broker. If you have a specific offer on the table or want to see what an end-of-term check looks like in practice, talk to us in a no-obligation initial consultation. We do not make return promises.