Solar panel degradation calculator: capacity and energy over a PV plant's life

Current n-type module warranties guarantee 99% of rated power after year 1 and 87.4% to 88.85% after year 30. Measured US plants lose a median 0.75% of output a year, a whole-system rate. This solar panel degradation calculator turns either assumption into capacity by year, a lifetime energy factor and the year output falls below 80%.

What does this solar panel degradation calculator compute?

It estimates how much of a photovoltaic (PV) plant's starting direct current (DC) capacity remains each year, and how much lifetime energy that costs. Four inputs set the case: first-year degradation (0 to 3%), annual degradation from year 2 (0 to 1.5%/year), lifetime (10 to 40 years) and the model, linear or compound.

The arithmetic is fully stated. Capacity starts at 100% and is a share of the day-one value:

End of year 1 = 100 - first-year degradation
Linear:   end of year y = end of year 1 - rate × (y - 1)
Compound: end of year y = end of year 1 × (1 - rate/100)^(y - 1)
Yearly average = (start-of-year capacity + end-of-year capacity) / 2
Lifetime energy factor = mean of the yearly averages

The tool shows three results and a chart. The results are the capacity left after the chosen lifetime (%), the lifetime energy compared with no degradation (%), and the first year in which end-of-year capacity drops below 80%, or a note that it stays at or above 80%. The chart shows capacity lost by the end of each year. Heat is a separate, reversible loss with its own temperature loss calculator.

What is the difference between module degradation and system performance loss rate?

Module degradation is the loss of a module's own power over time. System performance loss rate (PLR) is the decline in a whole plant's output, so it also captures availability, soiling and snow losses. Published numbers mix the two, so the sample matters as much as the median.

The Jordan and Kurtz review comes from NREL (now the National Laboratory of the Rockies). Its rates exclude short-term light-induced loss. It surveys the literature rather than sampling at random, and modules that degrade fast are less likely to stay in the field to be reported.

The 2022 figures come from the PV Fleet Performance Data Initiative, which collected data from more than 7.2 GW, 1,700 sites and 19,000 inverters.

What causes modules to degrade?

Four mechanisms matter most for crystalline silicon. Kiwa PVEL, an independent module test lab, reports all four in its 2026 Scorecard (figures as of October 2026). The cell designs compared are passivated emitter and rear cell (PERC) and tunnel oxide passivated contact (TOPCon).

UVID is the least settled. Fraunhofer ISE finds that standard UV tests overstate field loss unless modules are stabilised under light first, leaving up to 5% after a 60 kWh/m² dose. Kern et al. measured 2.3% to 3.2% non-recoverable loss. Soiling and hail are losses, not degradation: soiling is recoverable, and hail damage is an event, not a rate.

What do manufacturer performance warranties promise?

Datasheets give a first-year loss, a yearly loss after that and a guaranteed power at the end of the term. Current n-type modules promise 1% then 0.35% to 0.4% a year over 30 years. The p-type and PERC datasheets below promise 2% then 0.55% a year over 25 years.

A linear warranty is a straight line from the first-year floor to the end-of-term floor. LONGi's limited warranty writes one tier as a formula for power in year N, from 2 to 30:

Power(N) >= Nominal power × (1 - (1% + 0.4% × (N - 1)))

That tier guarantees 99% in year 1 and 87.4% in year 30, and the calculator's linear model draws the same line. The figures are floors, not forecasts. Terms vary by product, glass type and market, so the warranty text for the specific module governs. Kiwa's median LID losses (0.4% and 0.6%, as of October 2026) sit well inside the first-year allowances of 1% and 2%.

What do the worked examples show?

Two cases run through the model for 30 years with the linear option. The warranty case uses the 1% and 0.4%/year printed above, which are module rates. The fleet case uses the 0.75%/year median, a whole-system performance-loss rate that already includes soiling, availability and inverter ageing. Entering 0.75% as the first-year loss too is an analytical choice, so that every year loses the same amount.

The warranty case reproduces the 87.4% on the Trina datasheet. The fleet case is an extrapolation, because the initiative asks participants for at least 3 years of data, far short of 30. With the compound option, the fleet case ends year 30 at 79.78% instead of 77.5%, because each year's loss is a share of a shrinking base.

Small changes in inputs matter. In the warranty case, raising the annual rate from 0.4% to 0.5% lowers the lifetime energy factor by 1.4 points. One more point of first-year loss lowers it by 0.98 points.

Why does the calculator open on 1%, 0.4%, 30 years and linear?

These are the values printed on current n-type datasheets (Trina) and in the warranty tier above. Year 30 therefore returns 87.4%, a figure anyone can check against a datasheet. They describe a contractual floor, not a measurement. For a whole-system case, enter 0.75 and 0.75, the fleet median; the first-year 0.75 is an analytical choice.

How does degradation feed into yield and LCOE?

Lifetime energy equals the undegraded annual energy, times the years, times the lifetime energy factor. A plant that would make 1,000 MWh a year without degradation makes 28,023 MWh over 30 years in the warranty case and 26,625 MWh in the fleet case. First-year energy itself depends on irradiation, capacity and the performance ratio (PR), explained in capacity factor vs performance ratio.

The levelised cost of electricity (LCOE) divides lifetime cost by lifetime energy, so faster degradation raises it. Later years count for less once discounted, so the effect on LCOE is smaller than the effect on undiscounted energy. The LCOE calculator works with that ratio of cost to energy.

What this means for owners and investors

The two cases show the spread: 87.4% against 77.5% at year 30, and 93.41% against 88.75% in lifetime energy. The degradation assumption behind an energy forecast therefore deserves its own line, stated as module or system. Test results also differ by design, as Kiwa's 3.6% to 14.1% UVID range for one cell supplier shows (as of October 2026), so product-specific data says more than an average. Sources and conventions for this site are on the methods and data sources page.

Limits of the calculator

• Constant rates. It uses one first-year loss and one annual rate. LID, LeTID and UVID depend on time and partly recover, and fleet rates vary with climate (0.48%/year cooler, 0.88%/year hotter).
• No wear-out phase. Jordan and Kurtz note that linearity and the effect of climate were not settled. Late-life acceleration is not modelled.
• No repowering. Module swaps and inverter replacement are outside the model.
• Module rates for modules, system rates for systems. Warranty values describe module power under standard test conditions. A fleet performance-loss rate already includes soiling, availability and inverter ageing, so do not add those again. With a module rate, they are outside the model.
• Yearly averaging. Averaging start and end capacity approximates each year's energy.

Results are estimates, not a performance guarantee or engineering advice.