Modeling solar payback: the variables that matter and the ones that don't

A 6 kW rooftop system at $3 a watt pays for itself in 8.8 years at average US electricity prices — but 5.0 years in California and 15.0 years where power is cheap. Same panels, same sun, triple the payback range. The panels are not the variable. Your electricity price is.

Solar payback is usually sold as a single number — "pays for itself in seven years!" — as if it were a property of the panels. It is not. Payback is a ratio between what the system costs and what each kilowatt-hour it produces is worth to you, and both halves of that ratio vary wildly by house. The model itself is small enough to fit in one function. The interesting part is which inputs move the answer and which barely touch it.

The model 

function paybackYears({ systemKw, costPerWatt, taxCredit, kwhPerKw, electricityPrice }) { const grossCost = systemKw * 1000 * costPerWatt; const netCost = grossCost * (1 - taxCredit); const annualProduction = systemKw * kwhPerKw; // kWh per year const annualSaving = annualProduction * electricityPrice; return netCost / annualSaving; }

Four inputs, one division. But they do not matter equally — and the one most installers talk about least is the one that matters most.

Worked example: 6 kW system, $3/W, 30% tax credit

  • Gross cost: 6,000 W × $3.00 = $18,000
  • After the 30% federal tax credit: $18,000 × 0.70 = $12,600 net
  • Annual production: 6 kW × 1,400 kWh per kW = 8,400 kWh a year (a reasonable US average; sunnier states do better, cloudier ones worse)
  • Annual saving at $0.17 per kWh: 8,400 × $0.17 = $1,428 a year
  • Simple payback: $12,600 / $1,428 = 8.82 years

Electricity prices rise, and panels degrade slightly, so the real payback is a bit shorter: with prices rising 3% a year and production falling 0.5% a year, cumulative savings pass $12,600 about 8.1 years in. Over a 25-year panel life the same assumptions give roughly $48,683 in total savings against the $12,600 net cost — about $36,083 ahead after payback. The panels keep producing long after they have paid for themselves, and that tail is where most of the money is.

The variable that matters: your electricity price

Hold everything else fixed and change only the price you pay per kilowatt-hour:

  • At $0.10/kWh: $840 a year saved → payback 15.0 years
  • At $0.17/kWh: $1,428 a year saved → payback 8.8 years
  • At $0.25/kWh: $2,100 a year saved → payback 6.0 years
  • At $0.30/kWh: $2,520 a year saved → payback 5.0 years

Same system, same roof, same sun. The payback triples between cheap-power and expensive-power regions. This is why national payback averages are nearly useless: a system that is a no-brainer in California can be a 15-year slog two states over, and no change to the panels fixes that.

The variables that barely matter

System size. Double the system to 12 kW and both the cost and the savings double — payback stays at 8.8 years. Size decides how much money you save in total, not how fast you break even. (The exception is if a bigger system produces more than you use and the surplus is only worth a low export rate — then the marginal panels pay back slowly. Size matters at the edge, not in the middle.)

Panel orientation. Point the array east-west instead of due south and production drops roughly 12%: 8,400 kWh becomes about 7,392 kWh, savings fall to about $1,257 a year, and payback stretches from 8.8 to 10.0 years. Noticeable, but not decisive — a bad roof direction costs you a year, not a decade. A cheap electricity tariff costs you six.

Cost per watt. This one does matter, but less than people think. At $2.50/W the net cost is $10,500 and payback is 7.4 years; at $3.50/W the net cost is $14,700 and payback is 10.3 years. Shopping around for a better installer quote is worth roughly three years of payback — significant, and worth doing, but it still cannot rescue a system in a cheap-electricity region.

The figure that surprised me: the tax credit is worth more than a year of sunshine

The 30% credit takes the system from $18,000 to $12,600 — a $5,400 discount that cuts payback from 12.6 years to 8.8 years, nearly four years. Compare that with squeezing 12% more production out of a perfect south-facing roof: that saves about one year. Policy moves the answer more than physics does. Anyone modeling payback without the credit — or after it expires — is modeling a different purchase.

Honest caveats

  • This model assumes every kilowatt-hour you produce offsets a kilowatt-hour you would have bought at full retail price. If your utility pays a low export rate for surplus power you send to the grid, the marginal kilowatt-hours are worth far less. Homes that use most of their solar during the day get the full $0.17; homes that export half of it do not.
  • Production varies by location and roof. 1,400 kWh per kW per year is a US average. Arizona beats it, Seattle does not, and shade from a single tree can cost more than a wrong-facing roof. Get a site-specific production estimate before trusting any payback number.
  • Electricity price escalation is assumed, not guaranteed. 3% a year is a reasonable planning figure, but prices can flatten for years. At 0% escalation the payback is the simple 8.8 years, not 8.1 — the conclusion survives, the exact year does not.
  • This model excludes financing, inverter replacement, and maintenance. Paying cash is assumed. A loan adds interest that stretches payback; an inverter replacement around year 12–15 costs roughly $1,000–$2,000 and pushes the 25-year profit down, not the payback year much. A battery is a separate purchase with its own (usually much longer) payback — do not fold it into the panel math.
  • The 30% federal tax credit is a US policy figure. It phases down on a schedule and does not exist everywhere. Check what applies in your country and year before running these numbers.

Run your own numbers

Put in your system size, your installer's price per watt, your electricity rate, and your roof's production — and see which variable actually decides your payback:

Solar panel payback calculator — free, no signup, runs entirely in your browser.

The panels are the constant. Your electricity price, your tax credit, and how much of the power you actually use are the variables — and one of them matters three times as much as the rest.