Performance Ratio Guarantee: What a Solar EPC Should Commit To and How to Test It
A performance ratio guarantee commits the solar EPC contractor to a minimum PR, which is the energy the plant delivers divided by the energy its rated size should deliver for the sunlight that fell on the modules. To be worth anything, the clause must also fix the PR figure, the instruments, the test period, the weather conditions, the exclusions and the remedy if the plant falls short.
What performance ratio measures
PR strips out the weather. A plant in a sunny month and a plant in a cloudy month will produce different units, but their PR should be similar if both are healthy. The Ministry of New and Renewable Energy's draft quality control manual for rooftop plants defines it as the ratio of plant output to installed plant capacity, at any instant, with respect to the radiation measured. In its instantaneous form, PR is the inverter output in kW divided by the installed DC capacity in kW multiplied by the measured irradiance and divided by 1,000 W/m².
What PR does not measure is the roof. It does not tell you how many units you will get in a year, because that depends on the sunlight at your site. For that, use the solar generation calculator and a simulation from the designer, then use PR to hold the contractor to the quality of the plant.
What pulls PR down
The design PR is a loss budget. Module temperature, dust and soiling, shading, mismatch between modules, DC and AC cable losses, inverter conversion loss and any transformer loss each take a share. Ask the designer for the loss budget line by line, and for the PR the design predicts. A guarantee set at the design value leaves no margin; a guarantee set well below it gives away protection you paid for. Ask also whether the figure is corrected for module temperature. An uncorrected PR is lower in hot months, and a test in May is not the same test as one in January.
How the acceptance test runs
Public specifications show a workable pattern. A state renewable energy agency's published specification for on-grid plants (ANERT, Kerala) assesses PR for plants above 25 kWp using the monitoring data, under IEC 61724. It requires a pyranometer with a calibration certificate, mounted in the plane of the array, of Class II or better for plants of 100 kWp and above, plus module, inverter and ambient temperature sensors. Its acceptance test lasts five days, at least three of which must have irradiance above 600 W/m² for three continuous hours and a daily total above 3,000 Wh/m². If five qualifying days do not occur, the test is extended, and no delay damages arise from that weather delay.
The MNRE manual adds two further calculation modes. The short-duration PR is the energy delivered by the inverter divided by the number of days multiplied by the average daily peak sun hours and the DC capacity. The long-duration PR uses the same form over a longer period. Both are useful: the short test proves the plant at handover, and the long period shows how it behaves across seasons.
- Fix the measuring points: inverter AC output, or the revenue meter, and say which.
- Calibrate and install the irradiance sensor in the plane of the array before the test, and keep the certificate.
- Log at a fixed interval, and agree which data are valid. Remove grid-outage periods.
- Run the test for the agreed days under the agreed irradiance conditions.
- Calculate PR for the test period, and record the result in a signed commissioning report.
A worked example
This is an illustration, not a prediction. Take a plant with 100 kW of DC modules and a five-day test. The pyranometer records a total of 30 kWh/m² over the five days, so the average is 6 peak sun hours a day. The inverter delivers 2,250 kWh. Short-duration PR is 2,250 divided by (100 kW multiplied by 30 kWh/m² on the 1 kW/m² reference), which is 0.75 or 75%. If the contract guarantee were 78%, the plant would be 3 points short and the remedy clause would apply. Change the sensor reading by 10% and the PR moves by seven to eight points, which is why sensor calibration and placement matter.
Writing the clause
| Element | What to specify |
|---|---|
| Guaranteed figure | A PR in per cent, with year-wise values if module degradation is to be allowed for |
| Standard | The IEC 61724 series, or an equivalent BIS standard, named in the clause |
| Instruments | Plane-of-array pyranometer with calibration certificate, module temperature sensor, revenue-grade meter, logger interval |
| Test period and conditions | Number of days, minimum irradiance, extension rule if the weather is poor |
| Temperature treatment | Whether PR is temperature corrected, and the coefficient used |
| Exclusions | Grid outage, curtailment by the utility, owner-caused shading, force majeure, soiling only if cleaning is the owner's duty |
| Who witnesses | Both parties, with the data exported and signed |
| Remedy | Rectification first; then compensation per PR point of shortfall; a cap; a source of payment such as retention or a bank guarantee |
| Repeat tests | Annual PR test during the guarantee period, with the same method |
For reference, the model vendor agreement in the MNRE's office memorandum of 10 June 2022 for the residential rooftop programme has the vendor guarantee a minimum system PR of 75%, tested in line with IEC 61724 or an equivalent BIS standard, for five years. The same document expects modules to carry a performance warranty of more than 97% in the first year and a degradation of less than 1% per year after that. A commercial plant is not bound by a residential programme, so set your own figure, but the numbers show how an official floor is framed. Our guide to performance guarantees and the list of documents an EPC contractor should provide follow on from this.
What to exclude, and what to refuse to exclude
Reasonable exclusions are grid outages, utility curtailment, force majeure and damage by others. Be wary of exclusions for soiling when the contractor controls the cleaning, for shading that the design should have foreseen, or for inverter downtime caused by a poor design. Ask also for availability, meaning the share of daylight hours the plant is able to run, as a separate figure if the contractor also operates the plant. A dead string lowers PR only if the formula still counts the full installed capacity, so insist that the capacity in the PR formula is the installed DC capacity in the contract, not the capacity that happens to be running.
How Secured Engineers helps
We design and build C&I solar plants and we are willing to put a measured performance clause into the agreement. Design is done in-house, so the loss budget and the design PR come from the team that will build the plant. Fire NOC, CEIG and DISCOM net-metering applications are filed in-house, and mechanical, electrical, plumbing, fire and solar work can sit under one contract. We work from offices in Noida (Delhi NCR) and Ludhiana, with 535+ projects across 18+ states and 89.77% repeat clients.
If you are comparing offers, send us the performance clauses and the loss budgets you have received and we will tell you which ones can be tested. See our solar EPC service, message us on WhatsApp, call +91 70099 85817, or ask for a free proposal.
Frequently asked
What performance ratio should a solar EPC contractor guarantee?
There is no single correct figure. The design loss budget sets what the plant can achieve at your site. The MNRE model vendor agreement for residential rooftops uses 75%, and a commercial plant should be guaranteed at a figure supported by its own loss budget, with year-wise values if degradation is allowed for.
How is performance ratio tested at acceptance?
Typically over a defined run of days with a calibrated plane-of-array pyranometer, logged inverter or meter output and a minimum irradiance rule. One public specification uses five days, with at least three days above 600 W/m² for three continuous hours and a daily total above 3,000 Wh/m², extending the test if the weather does not qualify.
Is a generation guarantee better than a PR guarantee?
Not unless it corrects for weather. A guarantee in units per year depends on the sunlight of that year, which neither party controls. A PR guarantee removes the weather from the test. Many owners ask for a PR guarantee and treat a units-per-year figure as a forecast unless a correction method is written down.
What can a contractor reasonably exclude from the PR test?
Grid outages, utility curtailment, force majeure and damage caused by others are fair exclusions. Be careful with exclusions for soiling when the contractor controls cleaning, or for shading and inverter downtime that follow from the design. Ask for each exclusion to be listed in the clause, with how it is measured.
Does the PR guarantee last for the life of the plant?
Only for the period written in the contract. The MNRE model residential agreement runs its 75% guarantee for five years. For a commercial plant, ask for an annual test during the guarantee period, using the same method and instruments, and tie any compensation to a retention or bank guarantee that is still in force.
Related services
Guides that reference this page
- Solar Plant Commissioning Checklist: Tests Your EPC Must Run Before You Sign Off
- Solar EPC vs Developer vs Installer: Who Does What and Who You Should Contract
- Liquidated Damages in a Solar EPC: Delay LD, Performance LD and Realistic Caps
- Solar EPC Bid Evaluation Sheet: Scoring Modules, Inverters, Scope and Warranties
- Monitoring a C&I Solar Plant: String-Level Data, Alerts and What Reports You Should Get
- Solar Module Cleaning for a 1 MW Plant: Water, Frequency, Robots and the Soiling Loss Maths
Related on this site
- termALMM
- termNet Metering
- termOpen Access (Solar)
- serviceMounting Structures & BOS (Solar EPC)
- serviceO&M & Performance (Solar EPC)
- calculatorSolar Roof Load Calculator
- calculatorSolar Panel Calculator
- buyer's guideHow to Select a Solar EPC Contractor
- guideCLU for industrial land
- guidePESO licence for diesel storage
More on Solar & Renewables · all insights