Solar PPAs, leases, and EPC purchase compared

Choosing how to pay for a solar system is often treated as a financing afterthought, decided late in a project once the technical design is fixed. That ordering is backwards. The contract structure - power purchase agreement, lease, or direct EPC purchase - changes who bears performance risk, who claims depreciation and incentives, and how much the electricity actually costs over 20-25 years. Getting it wrong locks in a bad deal for the life of the asset.

This article is for commercial building owners, community energy groups, and technically literate homeowners evaluating quotes that arrive with unfamiliar acronyms attached. It assumes basic familiarity with system sizing (see /calculators/system-sizing) but not with project finance.

We compare the three dominant ownership models on a like-for-like basis: contract mechanics, typical term length, risk allocation, and a worked cash flow example, so you can read a proposal and know what you are actually being asked to sign.

The three ownership models

Direct EPC purchase

The customer pays an engineering, procurement and construction (EPC) contractor to design, supply, and install the system outright, either in cash or via a bank loan secured against the property or the equipment. The customer owns the asset from day one, claims any capital incentives or accelerated depreciation, and bears full responsibility for operation and maintenance (O&M) once the workmanship warranty period (typically 2-5 years) expires.

Power purchase agreement (PPA)

A third-party developer finances, owns, and operates the system on the customer's roof or land, selling the electricity generated back to the customer at a contracted rate per kWh, usually with a fixed annual escalator of 1-3%. The developer retains ownership, tax benefits, and O&M responsibility for the contract term, typically 15-25 years. At the end of term the customer usually has options to renew, have the system removed, or purchase it at fair market value.

Solar lease

Similar to a PPA in that a third party owns and finances the system, but the customer pays a fixed or step-up monthly fee for use of the equipment rather than a per-kWh rate for energy produced. The lease payment is largely decoupled from actual generation, which shifts production risk back toward the customer in years of poor irradiance, while a PPA keeps that risk with the developer.

Risk allocation compared

The single most consequential difference between these models is who holds performance risk - the exposure to a system underproducing relative to the design estimate because of soiling, degradation faster than modeled, inverter downtime, or a run of poor-irradiance years.

The balance-sheet line matters more than it looks. Under IFRS 16 and similar lease accounting standards, many long-term leases with an implicit right-of-use asset must now be capitalized on the customer's balance sheet even though no cash purchase occurred, eroding a key historical advantage of leasing for corporates. PPAs structured as a service contract for energy delivery rather than equipment use more often remain off-balance-sheet, though this depends on jurisdiction and specific contract language, and should be confirmed with an accountant rather than assumed.

Worked example: 250 kWp commercial rooftop, three structures

Assume a 250 kWp system on a warehouse roof, specific yield 1,050 kWh/kWp/year (temperate European site), annual generation 262,500 kWh in year one, degrading at 0.5%/year. Grid electricity price is EUR 0.22/kWh, escalating 3%/year. Installed cost via EPC purchase is EUR 1,050/kWp, i.e. EUR 262,500 total.

EPC purchase (cash): Year 1 savings = 262,500 kWh x EUR 0.22/kWh = EUR 57,750, assuming full self-consumption. Simple payback = EUR 262,500 / EUR 57,750/year = approximately 4.5 years. Over a 25-year design life, undiscounted savings net of an estimated EUR 45,000 in lifetime O&M and one inverter replacement (~EUR 18,000 in year 12) run well over EUR 900,000, before accounting for electricity price escalation, which improves the figure further.

PPA at EUR 0.15/kWh, 2%/year escalator: Year 1 cost to customer = 262,500 kWh x EUR 0.15/kWh = EUR 39,375, versus a grid bill of EUR 57,750 for the same energy - an immediate saving of EUR 18,375 with zero capital outlay. Because the PPA rate escalates slower (2%) than the assumed grid price (3%), the savings margin widens every year, though this depends entirely on how the specific contract is structured and whether the escalator is capped or indexed to a market reference.

Lease at EUR 3,200/month fixed: Annual cost = EUR 38,400, essentially flat regardless of generation. In a below-average irradiance year producing only 230,000 kWh, the customer still pays EUR 38,400, effectively raising their per-kWh cost to EUR 0.167/kWh, still below grid price but with no benefit passed through when production disappoints, and no benefit captured, either, when a strong year overproduces - that upside stays with the lessor structurally in some contracts, or is credited back in others.

Contract terms worth scrutinizing

• Escalator basis: is it a fixed percentage, or indexed to a published tariff or inflation measure? A fixed 3% escalator in a low-inflation environment can become expensive relative to the grid.
• Performance guarantee: does the developer guarantee a minimum annual output, and what is the remedy (bill credit, cash payment) if it is missed?
• Change of ownership / property sale clauses: PPAs and leases are typically structured to transfer to a new building owner; confirm the assignment mechanism and whether it requires the buyer's consent, since this frequently complicates property transactions.
• Buyout pricing: fair-market-value buyouts at year 10 or 15 should have a defined valuation methodology in the contract, not an open-ended appraisal at the developer's discretion.
• Curtailment and export limits: if grid export is capped or curtailment is imposed by the DNO/DSO, clarify who absorbs the resulting revenue loss under a PPA (see /blog/curtailment-and-negative-electricity-prices).
• Decommissioning and removal: at end of term, who pays to remove the system if the customer does not want to renew or buy?

Which model fits which situation

Direct purchase generally delivers the lowest lifetime cost per kWh for an organization with available capital, a tax position that can absorb depreciation or investment incentives, and the internal capability (or budget) to manage O&M. It is straightforward and there is no counterparty risk from a developer defaulting or being acquired.

A PPA suits organizations that want savings from day one without capital outlay or balance-sheet debt, and that value having performance risk sit with a specialized operator. It is common for public-sector bodies, retail chains, and any entity without appetite for tax-equity structuring.

A lease is less common in mature European markets than in the US residential sector, largely because IFRS 16 accounting and consumer protection rules have narrowed its advantages; where it persists, it typically appeals to customers who want a predictable fixed cost line rather than exposure to weather-driven generation variability, in exchange for giving up any generation upside.

Common mistakes

• Comparing a PPA rate only to the current grid tariff without projecting the grid tariff's own escalation over the contract term.
• Ignoring the impact of the customer's own consumption profile - a PPA or lease priced against expected self-consumption can turn expensive if the building's load shifts (e.g. after electrification of heating) and exported surplus is compensated at a much lower rate.
• Failing to check what happens to the contract on sale of the property, which can stall or complicate a transaction.
• Assuming ownership means no risk - a purchased system with no O&M budget set aside will underperform silently as soiling and minor faults accumulate.
• Treating the headline EUR/kWp price as comparable across EPC quotes without checking equipment tier, warranty length, and scope (e.g. whether monitoring and structural work are included).
• Not modeling a below-average irradiance year explicitly when comparing a PPA (variable payment) against a lease (fixed payment).

Frequently asked questions

Is a PPA always cheaper than buying the system outright?

Not necessarily. A PPA avoids capital outlay and shifts performance risk, but the effective cost per kWh over 20 years is usually higher than a purchased system's fully amortized cost, because the developer prices in its financing cost, margin, and risk premium.

Can a residential customer sign a PPA in Europe?

It varies by country. Residential PPAs are well established in parts of the US but less common in the EU residential segment, where leasing and consumer loans are more typical; commercial and industrial PPAs are widespread across Europe.

What happens if the PPA developer goes bankrupt?

The contract and underlying asset are typically assigned to another operator or a secured lender under standard project finance structures, but customers should check the specific step-in rights clause rather than assume continuity.

Does a PPA affect my ability to get other financing?

It can. Some PPAs are structured as an off-balance-sheet service, but lenders reviewing a company's financials will often still consider the long-term payment commitment when assessing overall leverage.

Are incentives like feed-in tariffs available under a PPA?

Any capital grants or accelerated depreciation typically accrue to the asset owner - the developer under a PPA or lease, not the host customer - though this is sometimes reflected indirectly in a lower contracted PPA rate.

Standards and references

Relevant standards: IEC 62446-1 (PV system documentation, commissioning tests, and inspection), IEC 61724-1 (PV system performance monitoring), and IFRS 16 (Leases) for accounting treatment of long-term equipment contracts.

Sources: IEA PVPS Task 1 reports on national PV markets and business models; IRENA "Renewable Power Generation Costs" annual series; NREL reports on solar PPA pricing trends and the NREL/LBNL "Utility-Scale Solar" data series; Fraunhofer ISE "Photovoltaics Report" for cost benchmarking; JRC PVGIS for site-specific yield estimation feeding into any of these cash flow models (see /calculators/payback).

For further reading on sizing the underlying system correctly before evaluating any of these contract structures, see /calculators/system-sizing and /blog/mppt-explained for how inverter choice affects long-term yield assumptions used in these models.