Calculate Solar Performance Ratio: Formula, Examples

Last updated: November 2026

A performance ratio tells you whether a plant delivers what its size and the sunlight it received should produce. It is a ratio, so it has no unit, and it only means something when the energy, capacity and irradiation inputs are matched.

This guide gives the formula, the inputs to collect, three worked monthly worksheets and the mistakes that distort PR. The formula is:

PR = (E_AC / P_DC,STC) / (H_POA / G_STC), with G_STC = 1 kW/m². STC means standard test conditions, the lab conditions used to rate modules.

Where:

• E_AC is the AC energy delivered at the named meter, in kWh, for the period.
• P_DC,STC is the installed DC module rating at standard test conditions, in kWp (kilowatts of peak power).
• H_POA is the plane-of-array (POA) irradiation for the same period, in kWh/m². This is the sunlight that lands on the tilted panel surface.
• G_STC is the reference irradiance (sun strength), 1 kW/m².

For annual PR, divide the summed energy by the summed reference energy. Do not average the monthly percentages. The worksheets below show the maths and the input checks. For the definition, see performance ratio. To read your results, use the PR interpretation guide.

What Performance Ratio Actually Measures (and What It Doesn’t)

First, define the final yield Y_f = E_AC / P_DC,STC, in kWh/kW. Then define the reference yield Y_r = H_POA / G_STC, in kWh/m² divided by kW/m². Both have units of hours. Their ratio has no unit.

PR = Y_f / Y_r
   = E_AC / [P_DC,STC × (H_POA / G_STC)]
G_STC = 1 kW/m²
PR percentage = 100 × PR

H_POA is the total sunlight over a period, not a reading at one moment. Keep G_STC in the formula even though its value is one. This avoids the missing-unit mistake in short spreadsheet formulas.

The 2013 weather-corrected PR research explains that PR depends on weather as well as plant performance. Dividing by sunlight does not remove the effect of location. Do not turn a seasonal change into a fixed regional fix or a diagnosis without data to back it up.

The Four Time Intervals: Instantaneous, Daily, Monthly, Annual

Do not average ratios that have zero or tiny denominators. Set your valid-data rules before you add up the data. Apply them the same way to energy and irradiation. A filter made for one project is not automatically a standard that applies everywhere.

Step 1: Gather the Inputs You Actually Need

Keep a record of your inputs next to the result:

Take a 250 kWp array on a 220 kW inverter. In this worksheet, use the stated 250 kWp DC STC capacity. The inverter limit can affect output. It does not change the installed module rating used here.

The IEC 61724-1:2021 publication page lists the second edition (published 21 July 2021), which eliminates Class C monitoring systems. IEC 61724-1 is the standard for monitoring PV system performance. Take sensor requirements from the edition your contract names.

For measured performance, use data from the same time the plant ran. A typical meteorological year is a model dataset built to represent a normal year. It is not the weather seen during that operating year. If you estimate POA from another channel or source, state how you converted it and how uncertain it is.

Worked Example 1: 6 kWp Residential Worksheet

All three datasets below are hypothetical inputs chosen to show the maths. They are not measured systems or regional benchmarks.

Assume a fixed installed capacity of 6 kWp. Assume the AC and front-POA boundaries are the same throughout, and every monthly period is complete and accepted.

For June, reference energy is 6 kW × (160 kWh/m² ÷ 1 kW/m²) = 960 kWh. PR is 710 ÷ 960 = 74.0%, rounded. The annual result is 5,595 ÷ (1,255 × 6) = 74.3%. In this worksheet, December has the lowest PR (71.4%). July has the highest energy (720 kWh).

Worked Example 2: 250 kWp Commercial Worksheet

The same calculation works for a larger array. The inputs are invented. So the monthly pattern tells you nothing about any climate.

July gives 33,763 ÷ (185 × 250) = 73.0%, rounded. Annual energy is 270,428 kWh. Annual irradiation is 1,430 kWh/m². Use those sums. Do not use a copied annual row or the plain average of the monthly PR.

Worked Example 3: 5 MWp Utility-Scale Worksheet

Use 5 MWp with MWh output, or convert both to kW and kWh. The reference yield stays H_POA / (1 kW/m²) in hours. Do not mix MWh output with a kWp figure you have not converted.

January gives 668 ÷ (165 × 5) = 81.0%, rounded. Annual PR is 8,490 ÷ (2,190 × 5) = 77.5%. The highest monthly ratios here occur in November and December (82.0%), not January (81.0%).

Temperature is one driver of seasonal PR change. The Sandia PV Performance Modeling Collaborative notes that PR is sensitive to temperature. It typically runs lower in warm months and higher in cold months (Sandia PVPMC). These inputs are made up. On a real plant, also check soiling, clipping and availability before you blame temperature.

What Is a Typical Solar Performance Ratio?

IEA PVPS Task 13 reviewed long-term monitoring data from PV systems in its November 2014 report. It found that typical PR rose over the decades, as the table shows. Task 13 is a research task of the IEA Photovoltaic Power Systems Programme (IEA PVPS).

Figures above come from IEA PVPS Task 13 (report T13-05:2014) and the IEC 61724-1:2021 publication page, both read in October 2026.

Source: IEA PVPS Task 13, Report T13-05:2014, Analysis of Long-Term Performance of PV Systems. The report does not split PR by home, commercial or utility-scale plants. So this page gives no benchmark by plant type. Compare your result with a design-stage PR from the same method for the same plant.

How to Interpret Any PR Result

Check the data first. Then look into performance:

1. Check units, installed capacity, which meter or channel you used, and the time covered.
2. Check that sensor and modeled irradiation describe the same plane and reference.
3. Compare with a fair baseline from the same method, including the weather and operating conditions.
4. Compare downtime, curtailment (output cut on purpose), clipping, shading and maintenance records with the interval data.
5. Work out the uncertainty and any open inputs before you blame equipment or ageing.

A high ratio can come from a low denominator. A low ratio can come from a missing numerator. So check the inputs before you suspect equipment. Use loss analysis to split the overall gap into its causes.

When the energy boundary changes

PVsyst’s PR documentation uses E_Grid (energy sent to the grid) for a plain grid system. It uses E_Grid plus E_Solar (energy used on site) when the system uses some of its own power or has storage. Counting exports alone understates the solar energy used. With storage, PVsyst notes that hourly or sub-hourly PR can exceed 1 because consumption can be delayed. So add up the data over a longer period. Do not count battery charge and discharge twice.

Bifacial modules collect light on both sides. For these systems, rear-side energy measured against a front-only reference can push PR above 100%. IEC 61724-1 edition 2 adds a bifacial PR. It adds the rear-side irradiance to the denominator. PVsyst’s documentation says versions before 8.0.12 left the bifaciality factor out of that calculation. State the method and software version when you compare reports.

Temperature-corrected PR

PR swings with array temperature. A weather-corrected PR removes the seasonal part of that swing. NREL proposed it in technical report NREL/TP-5200-57991 (Dierauf et al., 2013). PVsyst says the method is now part of IEC 61724-1. It writes the formula as:

PR_Temp = E_Grid / ( P_nom,PV × Σ [ (GlobInc / G_Ref) × (1 + μ_Pmpp × (T_array − T_array,weighted)) ] )

• GlobInc is the sunlight reaching the array at each time step, in W/m².
• G_Ref is 1000 W/m².
• μ_Pmpp is the module’s power temperature coefficient, a negative number. It shows how much power drops as the module heats up.
• T_array is the array temperature at the time step.
• T_array,weighted is the year’s average array temperature, weighted by GlobInc. Sunnier times count more.

Source: PVsyst documentation, Weather-corrected PR. PVsyst adds two limits. The correction over-corrects seasonal behaviour in shed systems because the rows shade each other. It is also almost unusable for tracking systems. Other seasonal effects, such as soiling, are not corrected.

PR vs. Specific Yield vs. Capacity Factor: Which Metric to Report

Specific yield is E_AC divided by the stated installed capacity. Capacity factor divides the period energy by the rated power times the hours in the period. State whether the rating is DC or AC. Use the real length of the period. Do not assume every year has 8,760 hours.

For the hypothetical 5 MWp worksheet, these metrics answer different questions from PR. A financial review also needs energy value, costs and timing. See solar installation cost per kWp for the cost side.

The Three Most Common PR Calculation Mistakes

• Replacing POA with GHI: name the plane and any conversion you used. GHI is the sunlight on a flat, horizontal surface. No single percentage error fits every tilt and location.
• Using inverter capacity: use the installed DC STC module rating in this method. For how inverter and transformer limits affect a commercial plant, see commercial solar transformer sizing.
• Mixing accepted periods: use the same timestamps and valid-data rules for energy and irradiation. Then add up the totals.

Keep the raw readings, exclusions, calculation version and output boundary with the worksheet. Before you compare a new result with an old one, list what changed in that record.

For the yield and financial side of a project, see the generation and financial tool, or book a SurgePV demo to model a real site.

Frequently Asked Questions

How do you calculate solar performance ratio?

Calculate solar performance ratio (PR) by dividing the AC energy read at the meter by the energy the plant would make at its STC rating under the measured sunlight: PR = (E_AC / P_DC,STC) / (H_POA / G_STC). STC means standard test conditions. IEA PVPS Task 13 reports a typical PR near 0.8 for systems installed from 2005 to 2012.

Should I use GHI or POA irradiation?

Use plane-of-array (POA) irradiation, the sunlight on the array surface. GHI is horizontal irradiation. It cannot replace POA for a tilted array. If you convert GHI to POA with a model, name the model and use the same time period.

Is annual PR the average of monthly percentages?

Not usually. First add up the monthly energy and the monthly reference energy. Then divide. A plain average gives a low-sun month the same weight as a high-sun month.

Can PR exceed 100%?

Yes, in some cases. A bifacial system measured against a front-only reference can exceed 100%. PVsyst also notes that systems with storage can show hourly PR above 1. If you see it on an ordinary system, check the units, time alignment, meter and capacity inputs first.

What is a good solar performance ratio?

IEA PVPS Task 13 reports typical PR near 0.8 for systems installed from 2005 to 2012. Its 2014 report shows a spread of 70% to 90% for newly installed systems. Compare your plant with the design-stage PR from the same method. Weather, temperature, shading and availability all move the result.

How often should PR be calculated?

Monthly PR shows trends. Daily or finer data helps you look into an event. Add up energy and irradiation to combine periods, keep data-quality flags, and follow the monitoring rules in your contract.

Sources and methodology

Sources, all read in October 2026: IEC 61724-1:2021 publication page, PVsyst PR documentation, NREL/TP-5200-57991, IEA PVPS T13-05:2014 and Sandia PVPMC. All worksheet data are hypothetical, and each monthly and annual PR was recomputed from its own inputs.