Operations guide

Performance ratio (PR): how to judge a plant after commissioning

How PR is calculated, why it beats raw generation as a yardstick, and what to write into the O&M contract.

Short answer

Performance ratio (PR) is the share of the energy your modules could have produced, from the sunlight that fell on them, that reached your meter. It corrects for weather, so it shows plant health where raw generation cannot. Judge PR against the plant’s own design model and trend, and have the O&M (operations and maintenance) contract fix how it is measured.

Key takeaways

  • PR compares what the plant delivered with what its modules could have produced from the sunlight it received.
  • Because it corrects for weather, PR shows plant health in a way raw generation cannot: a cloudy month lowers generation and leaves PR almost unchanged.
  • There is no universal “good” PR. Judge it against the design model for that month and against its own trend.
  • A PR guarantee only works if the contract fixes the formula, the irradiance sensor, the exclusions and what is paid on a shortfall.

Performance ratio shows how much available sunlight became power

Performance ratio (PR) answers one question: of the energy your modules could have produced from the sunlight that fell on them, how much reached your meter? Everything that goes wrong between the sun and the meter, from dust and heat to a tripped string or an inverter fault, shows up as a lower PR. That is why it is the number to watch once a plant is commissioned.

How PR is calculated, with one worked month

The definition used by the IEC 61724-1 monitoring standard divides two yields, both expressed per kWp of installed DC capacity:

  • Final yield = AC energy delivered (kWh) ÷ DC rating of the plant (kWp).
  • Reference yield = solar irradiation in the plane of the modules (kWh/m²) ÷ the 1 kW/m² reference irradiance. It works out as the number of hours the modules would have spent at full sun.
  • PR = final yield ÷ reference yield.
Illustrative exampleOne month on a 1,000 kWp plant

Assumptions

  • 160 kWh/m² of irradiation measured in the plane of the modules
  • 1,25,000 kWh delivered at the meter
Reference yield
160 hours
Final yield
125 kWh per kWp
Performance ratio
78.1%

Assumed figures to show the method. Not a quote and not a result from any Adhvanik plant.

Why PR is a better health check than units generated

Generation moves with the weather. Suppose the next month is cloudier: 120 kWh/m² of irradiation and 93,600 kWh delivered. Generation is down by 25%, which looks alarming, but PR is 78.0%, almost unchanged. The plant is healthy; the month was not sunny. The reverse also happens: a sunny month can hide a fault behind a good generation figure, while PR drops. That is why an O&M team watches PR alongside units.

Seven things that pull PR down

Common causes. Your monitoring data shows which apply to your plant.
CauseWhat you seeUsual fix
SoilingGradual PR decline between cleanings, faster in dry, dusty monthsCleaning schedule matched to site dust and rainfall
Module temperatureLower PR in the hottest months on a healthy plantDesign: ventilation gap and layout; contract: a temperature-corrected PR
String or module faultsA step down in PR, often visible on one inverter or stringString-level monitoring, thermography and IV-curve testing
Inverter downtimeSharp drop while the inverter is offFast fault response and spares
ShadingLoss at the same hours each day, changing with the seasonLayout review; trim or move the obstruction
Export limits or grid outagesGeneration clipped while the plant is healthyRecord separately so it is not counted against the plant
Measurement errorPR too high or too low with no physical causeClean, align and recalibrate the irradiance sensor

There is no universal good PR: compare against three things

No single PR suits every plant. It depends on climate, module temperature, the DC-to-AC ratio and how irradiance is measured. Three comparisons are more useful than any benchmark figure:

  • Against the design model. The yield simulation predicts a PR for each month. A plant that runs below its own prediction month after month has a problem.
  • Against its own trend. PR should repeat its seasonal pattern year on year, with a slow decline as modules age. A sudden step down is a fault.
  • Across inverters on the same site. Inverters on the same roof see the same sun. If one lags the others, look there first.

Uptime and PR catch different losses

Uptime, or availability, measures whether the plant was switched on and able to generate. PR measures how well it generated while it was. A plant can show near-perfect uptime and still lose output every day to dirty modules, a string of failed connectors or an inverter running derated in the heat. That is why the two belong together in an O&M contract: uptime covers outages, PR covers the slow losses that never trip an alarm. If a contract guarantees only uptime, the losses that cost the most over 25 years sit outside it.

Why PR slips slowly as the modules age

Modules lose a little output every year, at a rate set out in the manufacturer’s performance warranty, so a well-run plant’s PR drifts down over its life. Three things follow. A guarantee should state how the expected PR changes year by year, rather than holding one figure for 25 years. Year-on-year comparisons should use the same months, because the seasonal swing is larger than a year of ageing. And a decline faster than the module warranty allows is a reason to test the modules, using IV-curve tracing (a current-voltage test of a string of modules) and thermography (infrared imaging), while the warranty can still be claimed.

What a PR guarantee must define in the contract

A PR guarantee is only as good as its definitions. Agree the formula (and whether it is temperature-corrected), the irradiance sensor and its calibration, the measurement period, the events excluded (grid outages, export limits, force majeure) and the remedy when the guarantee is missed.

For reference, our standard O&M contract carries a 95% uptime guarantee plus a performance-ratio guarantee. Lost generation is credited at ₹3.00/kWh. Plants we maintain report generation, PR and irradiance on Asset Connect, our monitoring platform. Contracts run up to 25 years.

Four questions to put to any O&M vendor

  1. Can you show measured PR from operating plants, month by month?

    A promised PR means little without evidence from plants already running.

  2. Which PR formula does the guarantee use, and is it temperature-corrected?

    The same plant can pass or fail depending on the formula.

  3. Where is the irradiance sensor, and who cleans and calibrates it?

    A dirty sensor makes a poor plant look good.

  4. What is excluded from the guarantee, and what is paid on a shortfall?

    Exclusions and remedies decide whether the guarantee has teeth.

Sources

Policy and tax rules change. Facts below were checked in Nov 2026; confirm the version in force for your state and tax year before you sign.

  1. Hukseflux: How to calculate PV performance ratio and performance index (application note) (opens in a new tab). yield and PR definitions after IEC 61724-1
  2. Marion et al., NREL: Performance parameters for grid-connected PV systems (opens in a new tab)
  3. IEC 61724-1: Photovoltaic system performance, Part 1: Monitoring (IEC webstore) (opens in a new tab)

Common questions

Seasonal PR, CUF and irradiance sensors

Short answers. The working is in the guide above.

Get an O&M assessment
Why is PR lower in summer?

Modules lose output as they heat up, so the hottest months show a lower PR even on a healthy plant. Temperature-corrected methods exist to separate this effect from real faults; agree in the contract which one applies.

Is PR the same as CUF?

No. CUF (capacity utilisation factor) compares output with what the plant would produce at full rating every hour of the year, so it depends on location and weather. PR corrects for the sunlight received, which makes it the better health check.

Can PR be above 100%?

Not on a correctly measured plant. A PR near or above 100% usually means the irradiance sensor is dirty, misaligned or miscalibrated.

Who should measure irradiance?

A sensor in the plane of the modules, kept clean and calibrated, is the usual basis. Satellite irradiance is useful as a cross-check but adds uncertainty.

Free feasibility study

Talk to us about O&M for your plant

The feasibility study is free, takes 7 days and carries no commitment. An engineer replies within 48 hours.

What the study covers

  • Financial ROI sheet based on a preliminary design
  • Policy check for your state and connection
  • Roof feasibility and the area available for installation
Call us+91 22 4618 3484Get a free feasibility study