How to calculate solar payback period properly
The payback number you see in ads is usually the output of the most favorable assumptions available. Here is the full method, step by step, with every assumption exposed. The math is arithmetic — you can do it in a spreadsheet in twenty minutes.
Step 1: Net installed cost
Start from the gross contract price, subtract incentives you will actually receive. Important for 2026: the 30% federal residential credit (Section 25D) ended for expenditures after December 31, 2025 under the law enacted in July 2025 — see what changed. State and utility incentives vary; confirm each one is still funded before counting it.
net cost C = gross price − incentives actually received
Step 2: Year-one production
Use a physics-based model such as NREL’s free PVWatts with your actual address, roof tilt, and azimuth, and be honest about shading. Production per installed kilowatt varies a lot by location and orientation — roughly from around 1,100 kWh/kW/year in cloudy northern locations to around 1,700 in the sunny Southwest for well-oriented roofs. Get your number; don’t use a national average.
Step 3: Value of production — the step everyone botches
Split production into what you consume on-site (valued at your retail rate, including the rate tier or time-of-use period it offsets) and what you export (valued at your utility’s export rate, which in many states is now below retail):
year-1 savings S₁ = E₁ × [ f × r_retail + (1 − f) × r_export ]
E₁ = year-1 production (kWh)
f = self-consumption fraction (typically ~20–50% without a battery,
depending on your daytime usage)
Your self-consumption fraction depends on whether anyone is home during the day, electric vehicle charging, and appliance schedules. If you don’t know, 30–40% is a more defensible starting range than the 100% a retail-rate calculator implicitly assumes.
Step 4: Degradation and escalation
Savings in year t:
S_t = S₁ × (1 − d)^(t−1) × (1 + g)^(t−1)
d = degradation rate (median ≈ 0.5%/yr per NREL fleet studies)
g = electricity price escalation (test 0%, 2%, 4% — don’t assume one)
Step 5: Simple payback, then discounted payback
Simple payback is the smallest T where cumulative savings exceed net cost:
find smallest T such that Σ S_t (t = 1..T) ≥ C
Then repeat with each year’s savings discounted at a rate representing what your money could earn elsewhere (a conservative choice is a long-term bond or diversified-portfolio return):
discounted savings = S_t / (1 + i)^t
If the project only looks good on simple payback but not discounted, it is a marginal project being flattered by ignoring the time value of money.
Worked example (illustrative assumptions, stated)
All numbers below are example inputs to demonstrate the method — substitute your own.
| Input | Value |
|---|---|
| System size | 7 kW |
| Gross price ($2.80/W) | $19,600 |
| Federal credit (2026 purchase) | $0 |
| Year-1 production (1,400 kWh/kW) | 9,800 kWh |
| Self-consumption fraction f | 40% |
| Retail rate | $0.20/kWh |
| Export rate | $0.08/kWh |
S₁ = 9,800 × (0.40 × 0.20 + 0.60 × 0.08)
= 9,800 × 0.128 = $1,254/yr
simple payback ≈ 19,600 / 1,254 ≈ 15.6 years (before degradation)
With 0.5%/yr degradation and 2%/yr escalation the two effects roughly offset in the first decade, so the simple payback stays in the mid-15s. A retail-rate calculator on the same house would have claimed 9,800 × $0.20 = $1,960/yr and a headline payback of exactly 10 years — five and a half years rosier, from the export-rate assumption alone.
What “worth it” means
A payback comfortably inside the panel warranty period, robust to pessimistic escalation and realistic self-consumption, is a genuinely good project. A payback that only beats the warranty under the sunniest assumptions is a coin flip — and you deserve to know which one you’re buying. That sensitivity analysis, run with your utility’s actual tariff and your local weather, is exactly what a SolarVerdict report contains.
Get your home’s verdict first
SolarVerdict is launching soon: a weather-normalized, no-strings answer to “is solar (or a heat pump) actually worth it for my house?” One honest report with the math shown — planned price $29–49 one-time. We are not installers and we don’t sell leads, so “no, it’s not worth it for you” is an answer we’re allowed to give.
Privacy: your email is used only to notify you at launch. No marketing list, no sharing with anyone, deleted on request.