Performance Ratio is the number that turns nameplate kilowatt-peak into real kilowatt-hours. Get it wrong by five points and every sizing, yield and financial figure downstream is wrong with it.
This guide defines PR, walks through each loss term, shows how the losses combine, and explains where PR differs from efficiency and capacity factor.
Apply it directly in the PV System Size Calculator or measure it from plant data with the Performance Ratio Calculator.
Definition
where H_poa is plane-of-array irradiation (kWh/m²) over the period, G_STC is 1 kW/m², and P_dc,STC is the installed DC capacity at standard test conditions.
PR is dimensionless, usually 0.70–0.85 for well-built plants, and independent of how sunny the site is — which is exactly what makes it a measure of plant quality rather than resource quality.
Why PR Matters
- It converts a resource figure (PSH) into an energy figure.
- It is the contractual performance metric in most EPC and O&M agreements.
- It is climate-normalised, so plants in different locations can be compared fairly.
- A falling PR over time is the earliest indicator of soiling, string failure or inverter faults.
The Loss Terms
Temperature loss
Annual energy-weighted temperature loss is typically 5–7% in temperate climates, 8–12% in hot ones, and near 3% for well-ventilated arrays in cool climates. Ground-mount arrays run cooler than flush-mounted rooftop arrays.
Quantify it for your site with the Temperature Loss Calculator.
Soiling loss
Dust, pollen, bird droppings, salt and industrial fallout on the glass. Typical annual averages:
| Environment | Soiling loss |
|---|---|
| Rainy temperate, regular rainfall | 1–2% |
| Suburban, quarterly cleaning | 2–3% |
| Arid or dusty, monthly cleaning | 3–6% |
| Industrial, agricultural or desert, poor cleaning | 6–15%+ |
Soiling is highly seasonal: a long dry season can push instantaneous loss above 20% before the first rain.
Mismatch loss
No two modules have identical I-V curves, and series strings are limited by their weakest member. Manufacturing tolerance, uneven degradation and small shade patches all contribute. Typical 1–3% with sorted, single-SKU strings; higher when modules from different batches or wattages share a string.
Shading loss
Near shading from chimneys, parapets, poles and adjacent rows, plus far shading from the horizon. Unlike other terms this is highly non-linear — a single shaded cell can throttle a whole string — and it can only be quantified properly with a 3D scene. Values range from under 1% on a clear ground-mount site to 8%+ on a cluttered roof. Estimate it with the Shading Loss Calculator.
DC cable loss
Design targets are usually 1–2% at full load for string cabling. Check yours with the Solar DC Cable Sizing Calculator and the Voltage Drop Calculator.
Inverter loss
Conversion loss between DC input and AC output. Modern string inverters reach 98–98.6% Euro/CEC-weighted efficiency, giving 1.4–2.5% loss. Clipping — energy lost when the DC array exceeds the inverter's AC limit — is a separate item and grows with the DC/AC ratio.
AC cable and transformer loss
Inverter to LV panel, LV panel to transformer, and transformer to point of interconnection. Typically 0.5–1.5% for AC cabling; a distribution transformer adds roughly 1%.
Availability loss
Time the plant is not exporting: inverter faults, grid outages, curtailment, maintenance windows. Well-run plants achieve 99%+ availability, giving 1% or less.
LID, IAM and spectral effects
- LID (light-induced degradation) — an initial 1–2% drop in the first weeks of operation for p-type cells.
- IAM (incidence angle modifier) — reflection losses at oblique angles, roughly 1–3% annually.
- Spectral mismatch — the solar spectrum differs from the AM1.5 reference, worth ±1%.
Degradation
Long-term power decline, typically 0.4–0.6%/year after year one. It is not part of year-one PR, but it reduces PR steadily over the plant's life. Model it with the Panel Degradation Calculator.
How Losses Combine
Worked example:
| Loss | Value | Surviving factor |
|---|---|---|
| Temperature | 8.0% | 0.920 |
| Soiling | 3.0% | 0.970 |
| Mismatch | 2.0% | 0.980 |
| Shading | 2.0% | 0.980 |
| DC cable | 1.5% | 0.985 |
| Inverter | 2.2% | 0.978 |
| AC cable | 1.0% | 0.990 |
| Availability | 1.0% | 0.990 |
| LID / IAM / spectral | 2.5% | 0.975 |
Adding the same losses arithmetically gives 23.2%, implying PR = 0.768 — about 1.6 points pessimistic. The gap widens as losses grow: with 40% of total losses the arithmetic sum is off by more than four points.
Never count a loss twice. If inverter loss is inside PR, do not also divide by inverter efficiency. If DC and AC cable losses are separate terms, do not add a generic "wiring loss" on top.
Typical PR Values
| System type and climate | Realistic year-1 PR |
|---|---|
| Rooftop, hot and dusty climate | 0.72–0.78 |
| Rooftop, temperate climate | 0.78–0.83 |
| Ground-mount, fixed tilt, temperate | 0.80–0.84 |
| Ground-mount with tracker, good O&M | 0.82–0.86 |
| Poorly maintained or heavily shaded | below 0.70 |
A claimed PR above 0.85 in a hot climate should be treated as a red flag unless it is backed by measured data.
PR vs Efficiency vs Capacity Factor
Module efficiency is the fraction of incident irradiance a module converts at STC — around 20–23% today. It describes the module, not the plant, and it is already inside the nameplate kWp.
Performance ratio is the fraction of the nameplate-derived theoretical energy actually delivered. It describes the plant, not the resource.
CUF mixes resource and plant quality, so an excellent plant in a cloudy climate shows a poor CUF (13–15%) while an average plant in a desert shows a good one (20–24%). Use PR to judge quality; use CUF to judge output per installed watt.
How PR Is Used in Sizing
A 20 kWh/day load at 5 PSH needs 4.00 kWp at PR 1.0, 5.13 kWp at PR 0.78 and 5.71 kWp at PR 0.70. The PR assumption alone moves the array size by more than 10%, which is why it should be derived from a loss chain and stated explicitly rather than borrowed from a datasheet.
If you size from specific yield (kWh/kWp/year) instead, PR is already embedded in that figure and must not be applied again.
Measuring PR on an Operating Plant
You need plane-of-array irradiation from a calibrated pyranometer or reference cell, exported energy from the revenue meter, and the as-built DC capacity. Compare monthly PR against the design value: a gradual decline points to soiling or degradation, a step change points to a failed string, tracker or inverter.
Related Reading and Tools
- PV System Size Calculator — full sizing chain with a transparent loss table
- How to calculate PV system size
- How many solar panels do I need?
- Annual Energy Yield Calculator
- Temperature Loss, Shading Loss and Azimuth Deviation Loss calculators