SnapEnergyLab
Solar

Why solar panels never produce their rated power

The rating on the panel is measured in a laboratory at conditions your roof almost never sees.

Published 23 August 2026 · Published by SnapEnergyLab. Methodology and default assumptions are documented separately.

Rooftop solar panels on a suburban house with the shadow of a passing cloud crossing part of the array

Quick answer

A module’s rating is measured under standard test conditions: 1,000 W/m² irradiance and a cell temperature of 25 °C. Those two conditions rarely occur together on a roof — bright enough for full irradiance usually means the cells are far hotter than 25 °C.

Losses then stack multiplicatively through temperature, inverter, soiling, wiring, mismatch and shading. The combined result is the performance ratio, typically 0.80–0.85 for a good unshaded installation.

How to read the numbersYour inputSite defaultCalculated result
Rated array
6.0 kWp
Your input
Realistic summer peak
4.6 – 5.2 kW
Calculated result
Performance ratio
0.80 – 0.85
Site default

Standard test conditions are a laboratory, not a roof

STC exists so that modules from different manufacturers can be compared on identical terms. It was never intended as a prediction of field output, and treating the nameplate as one is the single most common source of disappointment with a new solar installation.

The temperature term is the biggest routine offender. Silicon PV output falls by roughly 0.3–0.4% for every degree the cell sits above 25 °C. On a still, sunny July afternoon, cell temperature of 55–65 °C is normal, which costs 10–15% on its own — at the exact moment irradiance is highest.

Some manufacturers also publish NOCT or NMOT ratings, measured at more realistic operating temperatures. Where a datasheet gives one, it is a better basis for expectation than the STC figure.

The loss stack

Typical photovoltaic system loss categories and magnitudes
LossTypical rangeNotes
Cell temperature above 25 °C5 – 12%Output falls roughly 0.3–0.4% per °C above the test condition. A dark roof in summer can put cells 30 °C above air temperature.
Inverter conversion2 – 5%DC to AC conversion, worst at very low input power.
Soiling1 – 5%Dust, pollen, salt and bird droppings. Higher in dry dusty climates and near trees.
DC and AC wiring1 – 3%Resistive losses, larger on long cable runs.
Module mismatch and tolerance1 – 3%Strings run at the current of the weakest module.
Shading0 – 30%+Highly site-specific. Optimisers or microinverters reduce, but do not remove, the penalty.
Degradation0.3 – 0.7% per yearCumulative. A 20-year-old array typically retains 85–90% of its original rating.
Snow and downtime0 – 5%Regional; concentrated in a few months rather than spread evenly.
  1. performance ratio = (1 − L₁) × (1 − L₂) × … × (1 − Lₙ)
  3. example: 0.92 × 0.97 × 0.98 × 0.98 × 0.98 = 0.840
  4. AC output (kW) = rated kWp × (irradiance ÷ 1000) × performance ratio

Losses multiply, they do not add. Five 3% losses cost 14.1%, not 15% — a small difference here, but the distinction matters once shading pushes one term high.

Worked example: peak output on a hot clear day

  1. Rated array6.0 kWpYour input
  2. Irradiance on the plane950 W/m²Illustrative model
  3. Irradiance factor950 ÷ 1000 = 0.95Calculated result
  4. Cell temperature58 °C, i.e. 33 °C above STCIllustrative model
  5. Temperature loss at 0.35%/°C33 × 0.0035 = 11.6%Calculated result
  6. Other losses combined≈ 8%Site default
  7. AC output6.0 × 0.95 × 0.884 × 0.92 = 4.64 kWCalculated result
Peak AC output from a 6 kWp array≈ 4.6 kW

Inverter clipping is a design choice, not a fault

Many systems are deliberately built with more DC panel capacity than AC inverter capacity — a 6 kWp array on a 5 kW inverter, for instance. Because the array rarely reaches its rating, the inverter is busy for more hours of the year and the small number of clipped peak hours costs less than a larger inverter would.

If your monitoring shows output flat-topping at exactly the inverter rating on the brightest days, that is the design working, not a failure. The annual energy lost to clipping in a sensibly sized system is usually under 2%.

Set your own loss assumptions

System losses are an editable input, not a hidden constant. Change them and watch the modelled annual figure move, so you can see how much of a quote’s optimism sits in that one number.

Open the solar output calculator

Assumptions and limitations

  • Loss ranges are typical published values used in PV modelling, not measurements of your installation. Shading in particular is entirely site-specific and can dominate everything else.
  • The worked example uses illustrative irradiance and cell temperature values chosen to show the arithmetic, not observed data from a specific location or day.
  • Temperature coefficients vary by module technology. Check the datasheet value for your modules rather than assuming 0.35% per °C.

Sources

  1. IEC 61215 — Terrestrial photovoltaic modules, design qualification and type approval — International Electrotechnical Commission

    Supports: Standard test conditions of 1,000 W/m², 25 °C cell temperature and AM1.5 spectrum used for module ratings.

  2. PVWatts Calculator — system loss defaults — U.S. National Renewable Energy Laboratory

    Supports: Typical loss categories and default magnitudes used in PV production modelling.

Related guides