Solar payback period in Arizona: how long until they pay for themselves

On current Arizona numbers, simple payback for a residential solar system runs roughly 10 to 24 years. That is a wide range, and the width is the useful part: almost none of it comes from Arizona’s sunlight, which is excellent and about the same for everybody. It comes from two variables you control or can at least measure, installed cost per watt and the share of your own production you consume rather than export.

Two changes in the last year moved the answer. The 30 percent federal tax credit ended December 31, 2025, which removed roughly $8,400 from a median 8 kW system. And APS has a rate increase request pending, which if approved makes every avoided kilowatt-hour worth more.

This walks the arithmetic from first input to final number. Verified figures are cited. Assumptions are labeled as assumptions.

How long does it take for solar panels to pay for themselves in Arizona?

Between about 10 and 24 years on the figures below, with roughly 15 to 18 years being the realistic middle for a cash purchase at the median 2024 price with average daytime load. Financed purchases sit at the long end. The three things that move your number most, in order, are what you paid per watt, whether you financed, and how much of your production you use in the house instead of exporting.

What the federal credit repeal did to the number

An identical 8 kW system at the LBNL 2024 median cash price paid back in 12.3 years if it was finished in 2025 and 17.8 years if it was finished in 2026. The 30 percent credit was worth roughly $8,400 on that system, and Section 25D no longer applies to expenditures after December 31, 2025.

The five inputs

Payback is one division problem: net cost divided by annual value. Everything else is working out those two numbers honestly.

Step 1: what the system will produce

Modeled with PVWatts (opens in a new tab) v8 for Phoenix (33.4484, -112.0740): an 8 kW-dc array, 20 degree tilt, south-facing, fixed roof mount, 14 percent system losses, against the NSRDB PSM V3 GOES tmy-2020 weather file.

For scale, Arizona’s average residential consumption is 1,075 kWh a month, or 12,900 kWh a year, per EIA 2024 data. So 14,041 kWh is a 109 percent annual energy offset. That is energy matching, not bill matching, and the distinction is the rest of this article.

Arizona’s installed residential fleet averages 7.49 kW per customer, derived from EIA Form EIA-861M net metering data for May 2026. An 8 kW model sits slightly above that.

Step 2: what is a solar kilowatt-hour actually worth?

Two prices, not one, and they are far apart. On APS, a kilowatt-hour you use inside the house avoids 15.418 cents you would otherwise have paid. The same kilowatt-hour exported to the grid earns 6.171 cents. That 2.5 to 1 ratio, not the panel spec sheet, separates a 10-year payback from a 20-year one on identical hardware.

Arizona ended retail net metering for new solar customers in ACC Decision No. 75859, issued January 3, 2017, in Docket No. E-00000J-14-0023. Exports have been bought at a set rate ever since.

Step 3: how much does solar cost in Arizona?

Published figures for 2024 and 2025 run from $2.32 to $4.70 per watt, which on an 8 kW system is a spread of about $19,000. The single largest driver inside that range is not equipment or roof complexity. It is whether the system was bought with cash or with a loan.

Two credible sources disagree, and the gap is the most useful thing in this section.

LBNL is explicit that loan-financed systems were “considerably higher priced than cash-purchase systems in 2024 (median prices of $4.7/W vs. $3.5/W), due partly to the loan fees rolled into the up-front price.”

That $1.20/W spread is $9,600 on an 8 kW system. It is larger than any incentive still available in Arizona.

Incentives that apply to a 2026 purchase

• Arizona state credit: 25 percent of the cost of the device, capped at $1,000, per A.R.S. § 43-1083. Nonrefundable, with a five-year carryforward.
• Federal credit: zero. Section 25D was repealed by Public Law 119-21 § 70506 for expenditures after December 31, 2025.

Net cost below is installed cost minus the $1,000 state credit.

Step 4: the arithmetic

Annual value at two self-consumption levels, on 14,041 kWh of production.

At 50 percent self-consumption:

Self-consumed  7,020 kWh x $0.15418 = $1,082
Exported       7,020 kWh x $0.06171 =   $433
                        Annual value = $1,516

At 70 percent self-consumption:

Self-consumed  9,829 kWh x $0.15418 = $1,515
Exported       4,212 kWh x $0.06171 =   $260
                        Annual value = $1,775

Same panels, same roof, same sun. A $259 a year difference from nothing but when the power gets used.

Divide net cost by annual value and you get simple payback.

• 50% self-consumed
• 70% self-consumed

View the numbers as a table

Simple payback in years, by installed cost and self-consumption — full dataset

Category	50% self-consumed ( yrs)	70% self-consumed ( yrs)
$2.32/W	11.6 yrs	9.9 yrs
$3.25/W	16.5 yrs	14.1 yrs
$3.50/W	17.8 yrs	15.2 yrs
$4.70/W	24.1 yrs	20.6 yrs

Simple payback, no discounting. Cost sources: NREL Q1 2025 benchmark ($2.32 and $3.25/W) and LBNL 2025 Data Update, 2024 installations ($3.50 and $4.70/W).

Read across a row and you see the self-consumption effect, worth two to four years. Read down a column and you see something less comfortable: the financing decision moves payback further than the sunshine does.

Here is the same thing as a recovery curve. The line crosses zero at payback.

• 50% self-consumed
• 70% self-consumed

View the numbers as a table

Unrecovered cost of a $27,000 net system, by year — full dataset

Period	50% self-consumed	70% self-consumed
Year 0	27000	27000
5	19420	18125
10	11840	9250
15	4260	375
20	-3320	-8500
25	-10900	-17375

Dollars outstanding. Illustrative: $28,000 installed at the LBNL 2024 median cash price, less the $1,000 Arizona credit.

What simple payback leaves out

Everything above is undiscounted, so it quietly assumes a dollar in year 20 is worth a dollar today. It is not.

The present-value test

Run the same $27,000 system through a present-value test. At a 5 percent discount rate, $1,516 a year for 25 years is worth $21,366 today. At $1,775 a year it is worth $25,017. Both are less than $27,000, which means that at the LBNL median cash price and a 5 percent discount rate, the system does not break even in present-value terms inside 25 years.

Change one assumption and it flips. Those figures hold electricity prices flat forever, which is obviously false. If retail rates rise 3 percent a year, the effective discount narrows to about 2 percent real, and the same $1,516 a year is worth $29,597 today, comfortably above the $27,000 cost.

So the honest framing is not “how many years.” It is: does electricity get more expensive faster than your money would have grown elsewhere. In Arizona right now that is a live question rather than a rhetorical one, with a 15.99 percent revenue increase requested in Docket No. E-01345A-25-0105 and no decision issued.

Three more things every table above excludes.

Degradation

The Department of Energy’s position is that “most systems maintain at least 80% of their initial power output for at least 25 years.” Published estimates of the annual rate vary: NREL’s Annual Technology Baseline uses 0.7 percent per year, an NREL review of field data put the median at 0.5 percent, and a DOE consumer page says about 1 percent. At 0.5 percent a year, cumulative output over 20 years averages roughly 95 percent of year one, which adds about 5 percent to any payback figure above.

Inverter replacement

A string inverter is unlikely to last the full array life. There is no reliable Arizona-specific public figure for the replacement cost, so no number is invented here. Take the figure from your own equipment documentation and add it to the cost side at the year you expect it.

Financing interest

The $4.70/W LBNL figure captures fees rolled into the price, not the interest paid over a loan term. Interest is additional, and on a 20-year loan it can exceed the state credit several times over.

Is solar payback different on SRP than on APS?

Substantially, and not in SRP’s favour on energy alone. The same 14,041 kWh system at 70 percent self-consumption pays back in 15.2 years on APS and about 20.4 years on SRP’s E-13 plan, because SRP’s export credit is 3.45 cents against APS’s 6.171. On SRP’s demand plans the deciding variable is not kilowatt-hours at all.

The arithmetic does not change on SRP. The two prices in it change enormously, because SRP’s solar plans price energy very differently from APS.

Take the same 14,041 kWh at 70 percent self-consumption and price it on E-13, assuming an illustrative 12 cent blended avoided price across seasons and time periods:

Self-consumed  9,829 kWh x $0.12   = $1,179
Exported       4,212 kWh x $0.0345 =   $145
                       Annual value = $1,324

That is $27,000 divided by $1,324, or 20.4 years, against 15.2 years on APS with the same hardware.

On E-27 it looks worse still on energy alone, because both the avoided price and the netting credit sit near 6.5 cents. But E-27 carries demand charges of up to $36.05 per kW in July and August, and that is where an SRP customer’s real lever is. On that plan the question is not how many kilowatt-hours the array offsets, it is whether storage can hold down a 30-minute peak.

The general point: which utility and which rate plan you are on changes payback by years, and it does so before anyone looks at your roof.

What shortens payback

• Higher retail rates. Every approved increase makes avoided kilowatt-hours worth more.
• Daytime load. Working from home, a pool pump on a midday schedule, pre-cooling before the on-peak window.
• Cash purchase over a loan. $1.20/W at the 2024 medians.
• Claiming the state credit. Only useful if you owe Arizona income tax; nonrefundable with a five-year carryforward.

What lengthens it

• The falling export rate. APS has cut the export rate exactly 10 percent a year since 2017, from 12.9 cents to 6.171 cents. Your own rate locks for 10 years at interconnection, but everyone deciding today starts from a much lower base.
• An empty house from 8 to 6. Production peaks at midday and leaves at 6.171 cents, then comes back in the evening at 15.418 cents.
• Demand charges. On APS R-3 and both SRP demand plans, solar barely touches the charge.
• Fixed charges. APS’s daily basic service charge, the Grid Access Charge on nameplate kW-dc, and SRP’s $20 to $40 monthly service charge all survive a perfect solar month.

A worksheet you can run in an evening

1. Pull twelve months of hourly usage from your utility account. This is the single biggest source of error if you guess it.
2. Run your own roof through PVWatts. Free, government-run, no phone number required, about two minutes.
3. Compute your self-consumption share by overlaying hourly production on hourly usage. If you cannot do that, at least count what fraction of your consumption happens between 9 a.m. and 4 p.m.
4. Pull the two prices from your own tariff, not from a national average.
5. Divide. Net cost over annual value. Then repeat it with your installed cost moved 20 percent in each direction and see how much the answer moves.
6. Discount it. Apply whatever return you would otherwise expect, and separately try it with an electricity escalation rate. If the answer only works at one specific escalation assumption, you have found the real risk.

If your payback lands past 15 years, that is not automatically a no. It is a signal to look at why, and the reason is almost always one of three things: cost per watt, financing structure, or a load shape that exports too much.

Rates, incentives and consumption figures were verified against published APS and SRP tariffs, EIA data, Arizona Revised Statutes and the enrolled text of Public Law 119-21 on August 9, 2026. Production figures come from a PVWatts v8 run for Phoenix on the same date. Installed cost figures come from LBNL’s 2025 Data Update covering 2024 installations and NREL’s Q1 2025 benchmark. Every discount-rate and blended-price figure is labeled illustrative and depends on assumptions stated inline.

Related reading

• Is solar worth it in Arizona? is the wider decision guide this arithmetic sits inside.
• APS solar buyback rates explained covers the 6.171 cent export rate and the 10-year lock that anchors the export half of the math.
• Arizona net metering and net billing explains why exports are worth less than imports in the first place.
• SRP’s E-27 rate plan explained covers the demand charge that dominates payback on SRP’s demand plans.
• Arizona solar incentives and tax credits covers the state credit in the net-cost line and what the federal repeal removed.

• Solar Basics
• APS
• SRP
• Incentives
• Arizona