Spanish installers commonly advertise commercial solar payback of roughly three to seven years. Those are market claims rather than independently verified figures, and the range exists because there is no standard answer. Payback for a specific business depends on its consumption profile, the tariff periods its generation displaces, the proportion of output consumed on site, the installed cost, what happens to surplus and the financial assumptions applied.
Two businesses in the same industrial estate, buying identical systems at identical prices, can produce genuinely different returns. One runs a continuous process six days a week and consumes most of what it generates. The other operates weekday office hours and exports a third of its output at a fraction of the value. The equipment is the same. The investment is not.
This article sets out how to calculate the return properly, works through a transparent example with every assumption visible, and identifies the modelling choices that make a projection unreliable.
The five measures, and what each one tells you
Different metrics answer different questions, and proposals often present only the most flattering one.
Simple payback is the number of years until cumulative net benefits equal the net installed cost.
Simple payback = Net installed cost ÷ Annual net benefit
It is easy to communicate and it ignores the time value of money, degradation, price changes and everything that happens after the payback point. Useful as a first filter, inadequate as a decision basis.
Annual return on investment expresses the yearly benefit as a percentage of the capital committed.
Annual ROI (%) = (Annual net benefit ÷ Net installed cost) × 100
Where the annual benefit is treated as constant, this is the reciprocal of simple payback, so it carries the same limitations. A 5-year payback is then equivalent to a 20% annual ROI.
Cumulative return is total net benefit over the asset life, less the initial cost.
Cumulative return = (Σ annual net benefits over the period) − Net installed cost
This is where solar looks strongest, because a system paying back in five years continues producing for another twenty. It is also where optimistic assumptions compound most, so it should never be presented without them.
Net present value discounts future benefits to what they are worth today.
NPV = Σ [Ct ÷ (1 + r)^t] − C0
where Ct is the net cash flow in year t, r is the discount rate and C0 is the initial investment. A positive NPV means the project beats the discount rate. The discount rate should reflect the business's actual cost of capital rather than a convention borrowed from a brochure.
Internal rate of return is the discount rate at which NPV equals zero. It allows the project to be compared against other uses of the same capital, which is the comparison a finance director actually makes.
For a capital investment of this size, NPV and IRR are the measures that matter. Simple payback is what gets quoted.
Self-consumed and exported electricity are not worth the same
This distinction drives more of the answer than any other single factor.
A kilowatt-hour consumed on site as it is generated reduces the metered electricity imported at that moment. Its value is the cost the business avoids, which includes the energy price agreed with the supplier plus the variable regulated components applying to imported energy.
A kilowatt-hour exported is valued under a different mechanism entirely. Where the installation qualifies for simplified compensation, the surplus is valued and applied within the billing process, typically linked to wholesale market prices. Where the business sells surplus as a producer, the value is the market price less representation, metering and administrative costs. In either case the figure is materially lower than the avoided import cost, and in a zero-export configuration surplus generation has no value at all because it is curtailed.
Any model that applies a single price to total annual generation is wrong before it starts. The two streams must be separated and valued differently.
Valuing generation against P1 to P6
Spanish commercial tariffs divide the year into six periods with different prices for energy and for contracted capacity. Solar output should be valued according to the periods it actually displaces.
Weekday daylight hours fall mainly in the higher-priced periods, which works in solar's favour. But the assignment shifts by season, and it differs between the peninsula and the island systems, so the weighting has to be built from the site's own interval data rather than assumed.
The practical method is to overlay modelled hourly generation onto actual quarter-hourly consumption, allocate each hour to its tariff period, and calculate a weighted average displaced value. That single figure can then be used in the payback calculation, and it is defensible in a way that a blended annual price is not.
What solar does not reduce
A commercial electricity bill contains elements that generation does not touch.
Contracted capacity is billed in each period regardless of what the panels produce, and a standard photovoltaic installation cannot be relied upon to lower recorded peak demand. A peak occurring under heavy cloud or after dark is unaffected by the presence of panels. Any projection that credits solar with a reduction in the capacity term is claiming something the technology does not reliably deliver.
Fixed charges and taxes on the electricity still imported also continue. Reactive energy is a more nuanced case: the presence of an inverter and its control settings can affect reactive behaviour at the connection point, in either direction, so any reduction in reactive charges must be demonstrated through site-specific electrical analysis rather than assumed as a benefit of installing solar.
Reviewing contracted capacity in each period is a worthwhile exercise, but it is a separate one, and its benefits should not be folded into the solar business case.
Weekend generation
Every Saturday and Sunday falls entirely within P6, the cheapest period, in every season.
For a business operating Monday to Friday this has two effects that both reduce return. On-site demand at weekends is lower, so a larger share of weekend generation is exported or curtailed. And whatever is consumed on site displaces the cheapest electricity of the week, so it is worth less than the same output on a weekday.
A model built on annual totals will not show this. A model built on interval data will.
A worked example
The following uses a 200 kWp rooftop installation on an industrial building. Every input is stated so that a reader can substitute their own.
Assumptions
| Input | Value | Basis |
|---|---|---|
| Installed capacity | 200 kWp | Example system |
| Installed cost, excluding IVA | €150,000 | €750 per kWp, mid-range for this size band |
| Annual generation | 300,000 kWh | 1,500 kWh per kWp; illustrative yield assumption, verify through PVGIS for the site coordinates |
| Self-consumption rate | 65% | From overlaying generation on interval consumption data |
| Electricity self-consumed | 195,000 kWh | 300,000 × 0.65 |
| Surplus exported | 105,000 kWh | 300,000 × 0.35 |
| Weighted value of self-consumed electricity | €0.155 per kWh | Weighted across the tariff periods displaced |
| Value of exported surplus | €0.050 per kWh | Net of applicable costs |
| Annual maintenance and monitoring | €2,500 | Approximately €12.50 per kWp per year |
The calculation
| Line | Calculation | Amount |
|---|---|---|
| Value of self-consumed electricity | 195,000 × €0.155 | €30,225 |
| Value of exported surplus | 105,000 × €0.050 | €5,250 |
| Gross annual benefit | €35,475 | |
| Less maintenance and monitoring | −€2,500 | |
| Net annual benefit | €32,975 |
Simple payback = €150,000 ÷ €32,975 = 4.5 years
Annual ROI = (€32,975 ÷ €150,000) × 100 = 22.0%
Reading the result
The 4.5 year figure is entirely a function of the nine inputs above. Change the self-consumption rate from 65% to 50%, holding everything else constant, and the net annual benefit falls to €28,250, giving a payback of approximately 5.3 years. Change the weighted displaced value from €0.155 to €0.130 instead, and the net annual benefit falls to €28,100, again giving approximately 5.3 years. Neither change touches the equipment.
This is why a payback figure without its assumptions is not information. It is a claim.
Low, central and high scenarios
A single projection implies a precision that does not exist. Three scenarios show the shape of the uncertainty.
| Input | Low | Central | High |
|---|---|---|---|
| Installed cost, excluding IVA | €165,000 | €150,000 | €135,000 |
| Self-consumption rate | 55% | 65% | 75% |
| Weighted value of self-consumed electricity | €0.130 | €0.155 | €0.175 |
| Value of exported surplus | €0.035 | €0.050 | €0.060 |
| Annual maintenance | €3,000 | €2,500 | €2,200 |
| Net annual benefit | €23,175 | €32,975 | €41,675 |
| Simple payback | 7.1 years | 4.5 years | 3.2 years |
The spread runs from just over three years to just over seven, which is roughly the range Spanish installers advertise. The difference between the ends of that range is not a difference of opinion about solar. It is a difference in six specific inputs, all of which can be established for a real site rather than guessed.
A business should ask which scenario its own data supports, and should treat a proposal presenting only the high case as incomplete rather than optimistic.
Beyond simple payback
Simple payback ignores what happens after the payback point, which for a twenty-five year asset is most of the story. It also ignores degradation, inverter replacement and the time value of money.
For the central case above, extending to a full lifetime model requires four further assumptions:
| Input | Value |
|---|---|
| Asset life modelled | 25 years |
| Annual panel degradation | 0.5% |
| Inverter replacement | Year 13, approximately €30,000 |
| Discount rate | 6%, being the business's cost of capital |
An important detail in the mechanics: the generation-related benefits degrade year on year while maintenance does not. Both the self-consumption saving and the export revenue fall by 0.5% annually as output declines, but the €2,500 maintenance allowance is held constant in real terms across the full period. Net benefit therefore erodes slightly faster than gross generation does, and by year twenty-five the annual net figure is around 12% below year one rather than the 11.3% suggested by degradation alone.
Holding electricity prices constant in real terms, the central case produces a net present value of approximately €237,600 and an internal rate of return of approximately 20.9%. The low case, run on the same lifetime assumptions, produces an IRR of approximately 12.3%.
Those figures are outputs of the stated assumptions and nothing more. Their value is that they can be compared directly against other uses of the same capital, which a payback period cannot.
Why holding today's price constant for 25 years is weak modelling
Two modelling errors are common and they pull in opposite directions.
The first is applying compound electricity price inflation across a twenty-five year horizon. A projection assuming 4% or 5% annual increases roughly triples the value of the electricity displaced by year twenty-five, and the resulting cumulative return is driven more by the inflation assumption than by anything about the installation. Spanish wholesale prices have been volatile in both directions, and regulated components have moved separately again. No one can forecast this reliably.
The second is treating today's price as fixed for the entire period, which is equally arbitrary and understates a real cost the business is exposed to.
The defensible approach is to model in real terms, holding electricity prices constant in today's money, and to test sensitivity to real price changes of, say, plus or minus 2% annually. That shows how much of the case rests on the price assumption. If the project only works at 5% annual inflation, the project does not work.
Project return and equity return
Where finance is used, two different returns exist and they are frequently conflated.
Project return measures the performance of the asset itself: the full capital cost against the full stream of benefits, ignoring how it was funded. This is the right measure for deciding whether the installation is worth building.
Equity return measures the performance of the money the business actually put in. If a €150,000 system is funded with €45,000 of equity and €105,000 of debt, the equity return is calculated on the €45,000, against the net benefit after debt service.
Where the cost of debt is below the project return, gearing raises the equity return. A project with a 20.9% project IRR funded at 5% will show a considerably higher equity IRR, which looks impressive and reflects the financing structure rather than the installation.
Both numbers are legitimate. Presenting the equity return while describing it as the project return is not. Ask which is being shown.
VAT, depreciation, grants and municipal reductions
Tax and incentive treatment should be applied carefully and never assumed.
IVA applies at the general rate of 21% to commercial installations. For a VAT-registered business it is normally recoverable subject to the usual deduction rules, so the payback calculation is generally performed on the net figure. It remains a cash-flow item during the recovery period.
Free depreciation. Real Decreto-ley 7/2026 extended free depreciation for company investments in renewable self-consumption installations into 2026. The qualifying investment is capped at €500,000, buildings are excluded, and the business must maintain its total average workforce for the twenty-four months following. This accelerates the tax shield rather than changing the pre-tax return, and its value depends on the company's own tax position. Confirm the treatment with the business's advisers before it appears in a model.
Grants. There is no permanent national scheme for commercial self-consumption. Support has come through IDAE programmes and regional calls that open and close through the year, are often subject to available funds and frequently require application before installation begins. A grant should not be deducted from the capital cost until it has been awarded in writing. Modelling the project both with and without it shows whether the investment stands on its own.
IBI and ICIO reductions are set by each municipality's fiscal ordinance rather than nationally. Where they apply they are worth including, but only after the local ordinance has been checked.
Batteries need their own calculation
Storage should be evaluated as a separate investment with its own return, not folded into the solar figures.
The reason is that a battery earns its money differently. Photovoltaic panels avoid imported electricity during daylight. A battery shifts energy from one period to another, protects against excess capacity charges or provides backup capability, and each of those benefits has a different value and a different reliability.
A battery also has a shorter life than the panels, degrades on a cycling basis rather than a calendar one, and carries replacement cost inside the modelled period. Combining the two into a single payback figure obscures the fact that the solar element may be strongly viable while the storage element is marginal, or the reverse.
Model the solar case. Then model the incremental cost and incremental benefit of adding storage. If the combined figure is the only one presented, ask for the split.
What shortens or lengthens payback
| Variable | Shortens payback | Lengthens payback |
|---|---|---|
| System cost | Lower €/kWp, larger system spreading fixed costs | Difficult roof, access equipment, long cable runs, medium-voltage work |
| Daylight consumption | Continuous operation across the generating window | Concentrated demand outside daylight hours |
| Sizing | Capacity matched to demand, high self-consumption | Oversized array producing surplus of low value |
| Location and irradiation | Higher site-specific yield | Lower yield, verified through PVGIS modelling |
| Shading and orientation | Unobstructed south-facing surface | Obstructions, poor orientation, shallow or adverse pitch |
| Electricity contract | Higher import cost being displaced | Already competitive contract, low displaced value |
| Surplus treatment | Compensation or sale at a workable value | Curtailment under zero export, or low export value |
| Degradation | Modest, typically modelled around 0.5% annually | Understated in the model, overstating later years |
| Maintenance | Accessible site, straightforward cleaning regime | Coastal, dusty or agricultural conditions requiring more frequent work |
| Inverter replacement | Modelled and budgeted | Omitted, then arriving as an unplanned cost |
| Finance cost | Cost of debt below project return | Financing above project return, or a short term |
| Grants and tax treatment | Confirmed and awarded | Assumed, or counted before eligibility is established |
Warning signs in installer ROI projections
Certain modelling choices consistently flatter a projection. Any one of them justifies asking for the underlying calculation.
Unrealistic self-consumption. A rate above 70% requires a genuine reason in the consumption data. Where a proposal assumes it without showing interval data, the number has been chosen rather than calculated.
Excessive electricity price inflation. Compound annual increases of 4% or more over twenty-five years do most of the work in a cumulative return figure. Ask what the result looks like at zero real inflation.
No degradation. Output declines over the asset life. A model with flat generation for twenty-five years overstates later years by roughly a tenth.
No maintenance or replacement cost. An inverter may require repair or replacement within a twenty-five-year model. A projection with no operating cost and no replacement provision is not a model.
Grants treated as guaranteed. A projection with a grant already deducted from the capital cost, before any award, is showing a project that may not exist.
Capacity charge savings attributed to solar. Contracted capacity reductions come from reviewing the contract, not from installing panels. Crediting them to the solar project double-counts a separate exercise.
Exported electricity valued like self-consumption. The single most common way to inflate a return. If the model applies one price to total generation, the figures do not hold.
A proposal that survives all seven checks is worth taking seriously. One that fails several is describing a different installation from the one being sold.
Frequently asked questions
What is the payback period for commercial solar in Spain? Spanish installers commonly advertise three to seven years, but there is no standard figure. Payback depends on installed cost, the proportion of generation consumed on site, the tariff periods displaced, the value of surplus and the financial assumptions applied. It should be calculated from the site's own consumption data.
How do you calculate commercial solar payback? Divide the net installed cost by the annual net benefit. The annual net benefit is the value of self-consumed electricity, plus the value of exported surplus, less maintenance and monitoring costs.
Why are self-consumed and exported electricity valued differently? Self-consumed electricity avoids the cost of importing that kilowatt-hour at that moment. Exported electricity is valued under simplified compensation or through a sale as a producer, both of which are typically linked to wholesale prices and materially lower. Under a zero-export configuration, surplus has no value because it is curtailed.
Should I use IRR or simple payback? Simple payback is a useful first filter. IRR and NPV are the appropriate measures for a capital decision, because they account for the full asset life, degradation, replacement costs and the business's cost of capital.
Does commercial solar reduce contracted capacity charges? Not reliably. Capacity is billed in each period regardless of generation, and peaks can occur outside daylight or under cloud. Capacity reduction comes from reviewing the contract, and crediting it to solar overstates the return.
What self-consumption rate is realistic? It depends entirely on the overlap between generation and demand. Rates above 70% require sustained consumption across the generating window and should be demonstrated from interval data rather than assumed.
How should electricity price inflation be modelled? Model in real terms with prices held constant in today's money, then test sensitivity to real changes of plus or minus 2% annually. A case that only works with high assumed inflation is a case built on a forecast rather than on the installation.
Should batteries be included in the solar payback calculation? No. Storage earns its return differently, has a shorter life and carries replacement cost within the modelled period. Model the solar case first, then the incremental cost and benefit of adding storage.
Reviewed 19 August 2026. All figures in the worked example are illustrative inputs, not market averages. Electricity prices, tax treatment and support schemes change, and any model should be built from current figures and the site's own consumption data.
Sources used
Market context and electricity pricing, accessed August 2026:
- Kowiik, Tarifas luz empresa junio 2026, June 2026, giving OMIE averages of approximately €0.065 to €0.08 per kWh and noting adjustment services adding €10 to €38 per MWh
- Revista Negocios, Precio de la luz para pymes en España abril 2026, May 2026, citing a Spanish wholesale market average of €65.52 per MWh for 2025 and an indicative all-in figure of around €0.20 per kWh for SMEs
- Eleia Energía, Tarifas de luz para empresas en 2026, April 2026, on 3.0TD and 6.XTD structure
Regulatory and fiscal, primary sources:
- Real Decreto 244/2019, self-consumption framework and the treatment of surplus
- Real Decreto-ley 7/2026, free depreciation for company renewable investments, the €500,000 cap, the exclusion of buildings and the twenty-four month workforce condition
- CNMC Circular 3/2020, tariff period structure and capacity billing
- IDAE, ayudas y subvenciones, for current programme status
One item to note before publication. The worked example uses 1,500 kWh per kWp as an illustrative yield rather than a modelled figure for any specific location, and the article states this. The weighted displaced values of €0.130 to €0.175 per kWh are constructed from published SME electricity cost data rather than taken from a single source. The NPV and IRR figures have been independently reproduced against the stated assumptions.