Trap guide · the four assumptions

Solar payback mistakes — what's inside a 7-year number.

Updated May 16, 2026. EIA and PVWatts re-verified the same day; we re-pull both quarterly.

Most residential solar proposals report a single payback figure with arithmetic confidence: "Your system pays for itself in 7.2 years." Underneath that one number sit four assumptions, each silently moving the result by a year or more, and three of them are usually wrong by enough to matter. The payback figure isn't arithmetic. It's a forecast in a suit.

Below: the four assumption errors that most often hide inside residential-solar payback math, what realistic ranges look like for each, and what the headline year does when you swap the defaults for grounded numbers.

Mistake 1 — Using a fixed utility-rate escalator

Most proposals assume utility rates rise 3–5% per year, every year, for 25 years. That one assumption compounds: at 5%, a bill that's $200/month today is $677/month by year 25. At 2%, it's only $328. The payback year shifts dramatically depending which number you plug in.

EIA data¹ shows U.S. residential rates rose 7.4% year over year through February 2026 — meaningfully above the 3–5% range. But the 10-year residential rate trend has been closer to 2–3%/yr. Both directions of error are real, depending on which window you pick. A proposal that picks 4% and runs it for 25 years is making a specific forecast — not splitting the difference, not citing a source, just compounding.

How to test: rerun the payback math with a conservative escalator (2.5%) and an aggressive one (5%). If the year shifts by more than two, the original figure was fragile. If it shifts by three or four, the escalator was load-bearing on the entire savings case.

Mistake 2 — Inflating the annual production estimate

The second-largest assumption is the system's annual production. The proposal lists it in kWh/yr. NREL PVWatts² is the canonical estimator — it uses NSRDB irradiance data and a default 14% total-system-losses figure to model annual AC output for any ZIP, tilt, azimuth, and DC kW. Most U.S. residential installs land between 1,000 and 1,600 kWh per installed kilowatt annually.

Where proposals inflate: lower-than-default loss assumptions (proposals sometimes use 10% or 12% instead of 14%), best-case shading models that don't see your specific trees, or equipment specs that assume perfect-condition output from year one. A 10% production overestimate compounds into a roughly proportional revenue overestimate, which shortens the payback year by 7–10%.

How to test:divide the proposal's annual production by your DC kW. If the result is above 1,500 kWh/kW outside the Sunbelt, or above 1,300 kWh/kW with significant shading, run your ZIP through PVWatts directly and see what NREL expects. A 10–15% gap deserves an explanation.

Mistake 3 — Ignoring panel degradation

Solar panels lose roughly 0.5% of rated output per year — the industry-standard linear degradation curve specified in most 25-year performance warranties. A panel rated at 400W in year one produces roughly 380W in year 10, 360W in year 20, and about 350W by the end of the 25-year warranty period.

Some proposals quietly ignore the degradation curve in the 25-year savings chart — flat production for 25 years. Others use the manufacturer's warranty floor (typically 80–88% of rated output at year 25) without modeling the linear path to get there. Cumulative production over 25 years is roughly 7–11% lower than the flat-line assumption depending on which degradation rate the panels actually deliver, and that compounds against the rising-rates-saved-from-the-grid story.

How to test:ask what annual degradation rate the proposal's savings model assumes. An honest answer is "0.5%/yr linear" or "0.55%/yr per the manufacturer's spec sheet." If the answer is "no degradation modeled" or "we used the year-25 warranty floor," the proposal's 25-year savings figure is high by ~5–10%.

Mistake 4 — Assuming the wrong compensation model

The fourth mistake is the one most homeowners don't know is a question. When your system produces more than your home uses at 2pm on a July afternoon, the excess gets exported to the grid. How your utility credits that export determines whether 1 kWh exported is worth 1 kWh imported (net metering) or worth ~20–30% of that (net billing, or wholesale rate).

Proposals frequently assume retail-rate net metering even in utility territories that have moved to net billing or a wholesale-rate compensation structure. California's transition from NEM 2.0 to the Net Billing Tariff (effective April 2023) is the most-cited example, but utilities in many other states have shifted or are shifting. Compensation-model differences alone can change a 7-year payback into a 12-year payback for an identical system.

How to test: read the net metering vs net billing guide, then ask your utility directly: "What compensation model applies to new residential solar interconnections starting today?" If the proposal assumes retail-rate net metering and your utility has moved to net billing, the payback figure is built on a number that's 50–80% too generous.

What changes when you re-run honestly

Take an 8.4 kW system with a salesperson's 7.2-year payback. Plug in realistic numbers for each of the four mistakes above:

• Escalator: 4% → 2.5% (matching the 10-year EIA trend)
• Production: 10,900 kWh → 9,800 kWh (matching PVWatts default 14% losses)
• Degradation: flat → 0.5%/yr linear
• Export credit: retail → 75% of retail (modest net-billing-style haircut)

Result for that hypothetical system: payback shifts from 7.2 years to roughly 11–12 years. The system still pays back inside the 25-year warranty window — it's not a bad financial decision. But the headline number was a forecast with rosy inputs across the board, and the real number tells a different story about the size of the win.

The 10-minute redo

The flagship quote calculator flags payback claims under 7 years as "optimistic" and runs the production estimate against PVWatts' 1,000–1,600 kWh/kW range. A dedicated payback calculator that exposes every assumption — escalator, degradation, compensation model, financing — is in the queue — building next. Until then, the four mistakes above are runnable on a back of a napkin if you have the system size and your utility rate handy.

What this guide doesn't cover

Lease and PPA payback math, which is a different problem entirely (you don't own the system; the comparison is monthly lease payment against pre-solar bill). Battery payback, which separates "savings" from "resilience" value. Federal tax-credit timing — covered in the canonical how to read a solar quote guide. Time-of-use rate-design specifics that can change the value of midday production substantially.

1. 1. EIA Electric Power Monthly, Table 5.6.A — Average Price of Electricity to Ultimate Customers by End-Use Sector. February 2026 residential average: 17.65 ¢/kWh, +7.4% year over year vs February 2025. 10-year residential rate trend: ~2–3% per year. Released April 23, 2026. Verified 2026-05-16. eia.gov/electricity/monthly ↩
2. 2. NREL PVWatts (v8) — annual production estimator using NSRDB irradiance data with default 14% total system losses (soiling, shading, snow, mismatch, wiring, connections, light-induced degradation, nameplate rating, age, availability). Typical US residential output: 1,000–1,600 kWh/kW annually. Verified 2026-05-16. pvwatts.nrel.gov · API docs ↩

Next: Net metering vs net billing — two compensation models, two different paybacks.