What this calculator does
It builds the full 25-year cashflow of a rooftop PV system — year by year, with panel degradation and tariff escalation compounding against each other — and reads the payback, the lifetime net saving, the levelised cost of energy and the internal rate of return off it.
Every number that varies by country is an input. The tariff you pay, the rate you are paid for export, the installed price, the rebate, the sun hours at your latitude: you supply all of them, in whatever currency you use. Nothing is hardcoded, so the answer is as right in Lagos or Lahore as it is in Leeds. The corollary is blunt and worth stating first: the result is only as good as the sun-hours figure and the tariffs you enter. Get the sun hours wrong by 20% and the payback is wrong by 20%. Guess the tariff and the whole thing is a guess.
The formula
Year-1 generation is the standard first-order derate estimate:
E₁ = P_dc × H × 365 × PR
P_dc— array size at standard test conditions, kW DC (the number on the quote)H— average daily peak sun hours at the site, h/day. A “peak sun hour” is one hour at 1 kW/m², soHis numerically the same as daily plane-of-array irradiation in kWh/m²/day. Look it up for your coordinates, tilt and azimuth on a solar resource map or PV yield dataset. There is no defensible global default; the value ranges from roughly 2.5 in northern Europe to over 6 in the desert belts.PR— performance ratio (derate), as a fraction. Temperature, soiling, mismatch, wiring, inverter efficiency and availability all live here.
The year-1 saving splits generation by where it goes:
S₁ = E₁ × f_sc × t_import + E₁ × (1 − f_sc) × t_export
f_sc is the self-consumption share. A kWh used the instant it is made avoids an import, so it is worth the full retail import tariff. A kWh exported is worth only the feed-in rate.
Then the lifetime cashflow:
CF₀ = −(installed cost − incentive)
CF_t = S₁ × (1 − d)^(t−1) × (1 + e)^(t−1) t = 1 … 25
with d the annual degradation and e the annual tariff escalation. The yearly saving is therefore a geometric series with ratio (1 − d)(1 + e). Payback is the year the cumulative cashflow crosses zero, interpolated inside that year; discounted payback is the same on cashflows discounted at your rate. IRR is the rate that makes NPV zero, solved by bisection. LCOE is the net cost divided by the discounted lifetime generation — discounting the energy as well as the money is what “levelised” means.
Reading the result
Self-consumption versus export is the whole argument. In most markets the export rate has collapsed to a small fraction of the retail tariff — it is common to be paid a fifth or a tenth of what you pay. When that is true, a kWh you use is worth five to ten kWh you sell, and the economics stop being about how big the array is and start being about when you use power. Running the dishwasher, the pool pump, the water heater or the EV charger at midday instead of at night moves the payback further than another 2 kW of panels will. The same logic prices storage: a battery converts cheap exports into avoided imports, and its value is exactly the spread between the two rates you entered.
The tool warns when self-consumption is under 40% and the export rate is under half the import tariff, because that is the combination where most of the generation is being sold at a heavy discount and the quoted payback is fragile.
LCOE is the cleanest comparison number. It is what a kWh from the roof costs you over the system’s life. Set it against the import tariff, not against the export rate: if LCOE is below the import tariff, self-consumed solar beats grid supply, and the gap is your margin.
Treat IRR and the 25-year saving as sensitivities, not forecasts. Both depend on a tariff escalation assumption 25 years out. Run the numbers again at 0% escalation. If the case survives that, it is a real case.
Typical values
- Peak sun hours — roughly 2.5–3.5 h/day in northern Europe and Canada, 3.5–5 across much of the US, China and southern Europe, 5–6.5 in Australia, the Middle East, north and southern Africa. Always look up your own site rather than taking a regional figure.
- Performance ratio — 70–85% for a well-installed, unshaded system. Below 65% implies real shading, a poor orientation or a hot climate with no ventilation gap. Above 90% is not achievable.
- Degradation — 0.25–0.7%/year for modules under a modern linear performance warranty, after a larger first-year settling loss. Anything faster deserves a look at the warranty schedule.
- Self-consumption without storage — 20–35% for a home empty during the day, 40–60% where someone is home or there is a daytime load, 60–85% with a well-sized battery.
Worked example
A 6.6 kW DC system, installed for 9000 with a 1500 upfront rebate — net cost 7500. The site averages 5.0 peak sun hours/day; the installer quotes an 80% performance ratio. The household expects 40% self-consumption. The import tariff is 0.30/kWh, export pays 0.05/kWh. Assume 3%/year tariff escalation, 0.5%/year degradation and a 5% discount rate.
E₁ = 6.6 × 5.0 × 365 × 0.80 = 9636 kWh
self-consumed = 9636 × 0.40 = 3854.4 kWh
exported = 9636 − 3854.4 = 5781.6 kWh
S₁ = 3854.4 × 0.30 + 5781.6 × 0.05
= 1156.32 + 289.08 = 1445.40
Note where the money comes from: 40% of the energy produces 80% of the saving.
The yearly saving grows by 0.995 × 1.03 = 1.02485 per year. Summing that series, the cumulative saving reaches 6000.70 by the end of year 4, and year 5 contributes 1594.52 — so the 7500 is recovered at 4.94 years. Discounted at 5% the same crossing happens at 5.77 years.
Over 25 years the savings total 49 275.83, giving a net saving of 41 775.83. Discounted lifetime generation is 129 557 kWh, so LCOE = 7500 / 129 557 = 0.0579 per kWh — well under the 0.30 import tariff.
The IRR is 21.48%. You can check that by hand without a solver: at 21% the present value of the 25 savings is 7683.80 against a 7500 outlay (NPV +183.80), and at 22% it is 7311.75 (NPV −188.25). The root is bracketed between them, and bisection converges on 21.48%.
FAQ
Where do I get the peak sun hours figure? From a solar resource map or PV yield dataset for your own coordinates — most national energy agencies, and several free global datasets, publish plane-of-array irradiation by location, tilt and azimuth. Use the annual average for the tilt and orientation you are actually installing, not the horizontal figure and not a national average.
Why is my payback so much worse than the installer’s? Usually one of three things: an optimistic self-consumption share, a tariff escalation assumption doing a lot of quiet work, or no discounting at all. Set escalation to 0% and self-consumption to something you can defend from your own half-hourly consumption data, and compare again.
Should I include the inverter replacement? For a 25-year view, yes — most systems need one around year 10–15. This tool models capital cost once at year 0, so subtract the expected replacement cost from the 25-year net saving to see how much of the case it eats.
What if my support scheme pays per kWh rather than upfront? Add it to the export rate (for generation you export) or, if it is paid on all generation regardless of destination, add it to both the import and export rates. Leave the upfront incentive at 0.
Does a bigger system always pay back faster? No. Beyond the point where you can absorb the midday output, extra capacity only produces exports, and exports are valued at the feed-in rate. Once self-consumption is saturated, each additional kW pays back at the export rate alone — often several times slower than the first few kW.
Indicative figures for comparing options. They are not a financial forecast, and they depend entirely on the sun-hours, tariff and cost figures you enter. Confirm export eligibility, connection limits and any incentive conditions with your network operator and supplier before committing.