Skip to content
RayValueSolar math for your home

How long solar panels last and how fast they fade

Published

Written and fact-checked by the RayValue team · How we research · Sources

Old, weathered solar panels on the roof of a metal shed
On this page
  1. Key takeaways
  2. What “degradation” means
  3. What the research found
  4. What that means over 25 years
  5. How long the equipment lasts
  6. Putting it into your payback math
  7. Next steps

A rooftop solar system is a 25-year bet, so it is fair to ask what the panels will actually be doing in year 20. The short version: panels do not stop working on a schedule. They lose a small slice of output each year, and the part most likely to need replacing along the way is the inverter, not the panels.

This post walks through what national lab studies measured, what that means for payback, and what to check in your paperwork.

Key takeaways

  • Large studies from the National Renewable Energy Laboratory (NREL, now called NLR) put typical output loss at roughly 0.5% to 0.9% per year (medians), depending on the study and climate. Averages run higher, up to about 1.1% in the hottest zone.
  • Hotter climates showed faster loss in the largest U.S. fleet study, and the average loss in the hottest zone was higher than its median. For a hot-climate roof, it is wise to plan toward the higher end of the range.
  • Panels in homes are commonly described as lasting 20 to 25 years or more. String inverters usually last 10 to 15 years.
  • Inverter replacement is a real cost that most simple payback math leaves out. Ask for the warranty terms in writing.
  • Use ranges, not a single number. Our calculator shows low, likely and high cases for this reason.

What “degradation” means

Degradation is the slow, permanent loss of a panel’s ability to turn sunlight into electricity. It is different from dirt on the glass or a few cloudy weeks, which come and go. NREL researchers describe the permanent kind as a gradual loss that is small compared with daily and seasonal swings in sunshine, but that adds up across a system’s life.

Because it is measured per year, it is easy to misjudge. A loss of 0.5% a year sounds negligible. Over 25 years it adds up, so it is worth doing the arithmetic.

What the research found

The long-run review: about 0.5% per year

In a 2012 review, NREL researchers Dirk Jordan and Sarah Kurtz collected nearly 2,000 degradation rates from published field studies covering about 40 years of outdoor testing. The median was 0.5% per year. The average was higher, at 0.8% per year, because a minority of modules lost output much faster. About 78% of the reported rates were below 1% per year.

Those numbers include a lot of older panel designs, so they describe the history of the technology, not a guarantee for a panel you buy today.

The recent U.S. fleet study: about 0.75% per year

A 2022 paper by Jordan and colleagues looked at real commercial and utility-scale systems in the U.S.: more than 7.2 gigawatts of capacity, 1,700 sites and 19,000 inverters, roughly 6% to 7% of the U.S. PV market at the time. The overall median performance loss was 0.75% per year, which the authors said is in line with earlier findings. The study’s rates can include temporary losses such as dirt that is not fully cleaned off, so they may overstate permanent fading.

A February 2024 NREL update expanded the data to about 8.5 gigawatts and 24,000 inverter data channels, and described its degradation results as consistent with the earlier publications, while noting that the answer depends on which systems are in the sample and on soiling. When the authors used a method that explicitly removes recoverable soiling, they found average degradation around 0.5% per year, and they say part of the measured loss may come from soiling that is not fully recovered. Note that these are commercial and utility-scale systems, not houses. They are the best large public dataset available, but your roof may behave differently.

Heat matters

The same 2022 study sorted systems into temperature zones. The coolest zone it reports (T3) had a median loss of 0.48% per year (mean 0.63%). The middle zone (T4) had a median of 0.78% (mean 0.91%), and the hottest zone (T5) had a median of 0.88% (mean 1.14%). The mean sits above the median because a minority of systems lost output much faster.

The study also grouped systems by humidity zone, but the authors state that a trend with respect to humidity zones could not be determined with confidence. It does not give a Florida-only figure, and we have not confirmed which zone a given Florida or California address falls into, so treat the zones as a guide to direction, not a local forecast. A hotter roof should lean toward the higher rates, and the T5 mean of 1.14% is a reasonable cautious case.

What that means over 25 years

Here is an illustrative calculation, not a forecast for your roof. It assumes output falls by the same amount each year, which is a simplification, and uses the rates above.

Yearly loss Source of the rate Output after 25 years (about)
0.5% 2012 long-run median 87.5% of original
0.75% 2022 fleet median (2024 report consistent) 81% of original
0.88% 2022 hottest-zone (T5) median 78% of original
1.14% 2022 hottest-zone (T5) mean 71.5% of original

(Arithmetic: 100% minus 25 years times the yearly rate, rounded.)

So on these rates, a system could be producing somewhere around 72% to 88% of its first-year output by year 25, and more than that in the middle years. For payback, the early years matter most, because that is when the system is producing the most. A payback that only works in the last years of the system’s life is more exposed to degradation than one that pays back earlier.

The calculator reduces output a little each year, as described on our methodology page, where you can see how payback is built. It compounds a yearly decline rather than subtracting a straight line, so its numbers will differ slightly from the table above. It does not include inverter or battery replacement, which is the next topic.

How long the equipment lasts

Panels

The U.S. Department of Energy says residential panels usually last 20 to 25 years. A separate DOE article from 2018 said panels on the market can last 20 to 30 years. Both describe expected lifetimes, not guarantees.

Inverters

The inverter converts the panels’ direct current into the alternating current your home uses, and it has electronics that wear out. DOE says string inverters usually last 10 to 15 years and may need replacing during the life of the panels. Microinverters, which sit under each panel, tend to have lifetimes similar to the panels. The 2018 DOE article put the typical inverter lifetime at about 10 years and described longer-lasting designs as a research goal, not a current norm.

That means a 25-year payback on a string-inverter system should reasonably assume at least one inverter replacement. We do not have a verified price for it, so ask your installer to quote one in writing, and compare against a microinverter option.

Warranties

Warranty terms differ by product, so we will not quote a standard. For reference, a 2009 DOE consumer guide said most solar systems came with a 25-year warranty, with inverter warranties usually around 10 years. That is old guidance, and current products vary, so read the actual documents. Ask for these separately:

  • Panel product warranty: covers defects. Ask how many years.
  • Panel performance warranty: promises a minimum output level over time. Ask what percentage it guarantees, and by which year.
  • Inverter warranty: ask the length, and whether labor and shipping for a swap are included.
  • Workmanship or roof warranty: covers the installation itself, including roof penetrations.

Also ask who honors the warranty if the installer goes out of business. Our guide on how to compare solar quotes lists more questions to put to installers.

Putting it into your payback math

  1. Start from your own bill. Your utility’s rates decide what each kilowatt-hour is worth. See your utility page, for example FPL, Duke Energy Florida, TECO, PG&E, SCE or SDG&E.
  2. Use the range. Treat 0.5% as a low-loss case and 0.9% to 1.1% as a cautious case for a hot roof.
  3. Add a line for the inverter if your quote uses string inverters.
  4. Remember there is no federal credit. The IRS says the residential clean energy credit is not available for property placed in service after December 31, 2025, so the system cost you pay is the cost you need to earn back. See is solar worth it without the tax credit.

Next steps

  • Run the Florida calculator or the California calculator to see a low, likely and high payback for your home.
  • If you already have quotes, check the inverter type and every warranty line before comparing prices.
  • If you want quotes from installers, you can request them here. There is no obligation, and degradation and inverter questions are fair to ask every installer.

None of these figures are promises. Real output depends on your roof, shade, equipment and weather, and a good installer should be willing to explain how they estimated your numbers.

Related reading: what residential solar costs per watt in 2026.

Sources