Performance ratio and availability: reading your solar plant's numbers

"How much did the plant generate last month?" is the first question most owners ask, and on its own it is the least useful one. Generation depends on how sunny the month was. To know whether the plant is actually performing well, owners need two metrics: performance ratio and availability.

Why generation alone misleads

A cloudy month will produce less energy even if every component works perfectly. A sunny month can mask a failed inverter. Comparing generation month-to-month or year-to-year mixes weather effects with plant performance. Performance ratio and availability separate the two.

Performance ratio (PR)

Performance ratio compares the energy a plant actually delivered with the energy it should theoretically have produced, given the irradiance measured at the site and the plant's installed capacity.

In simple terms:

PR = actual energy output ÷ (installed capacity × irradiation received ÷ reference irradiance)

A PR below the design expectation means energy is being lost somewhere — soiling, shading, equipment faults, degraded components, or losses in cables and inverters.

Things that affect PR:

• Soiling (module cleaning)
• High module temperatures
• Inverter efficiency and clipping
• Cable and transformer losses
• Equipment faults and downtime
• Module degradation over time (degradation and warranties)

Temperature-corrected PR adjusts for the effect of module temperature, making comparison across seasons fairer. Contracts should state which version is used.

Availability

Availability measures how much of the time (or how much of the potential energy) the plant was able to generate. If an inverter is down for three days, the portion of the plant behind it is unavailable for that period.

Availability can be defined in different ways:

• Time-based — the share of daylight hours the plant (or each part) was operational
• Energy-based — the share of potential energy that could have been produced, accounting for when outages happened

Energy-based definitions treat an outage at noon as worse than one at dawn, which better reflects actual loss.

Why definitions matter

In an O&M contract, small definitional differences change the result:

• Which irradiance sensor is used, and how it is maintained
• Whether PR is temperature-corrected
• Which events are excluded (grid outages, force majeure, curtailment)
• Whether availability is time- or energy-based
• How partial outages (one string, one inverter) are counted

These definitions should be written precisely — see structuring an O&M contract.

Using the numbers

A practical monthly review asks:

1. Was availability on target? If not, which outages caused the shortfall and were they excluded events?
2. Was PR on target? If not, which losses explain the gap?
3. Are there trends — a slow PR decline suggests soiling or degradation; sudden drops suggest faults.
4. Are recurring faults being fixed at root cause?

Sensor quality

Both metrics depend on accurate irradiance measurement. Pyranometers or reference cells must be correctly installed, clean and calibrated. A dirty or misaligned sensor produces a misleading PR.

An illustrative example

The figures below are purely illustrative, to show how the two metrics work together.

Suppose a month's monitoring shows that one plant generated less energy than the previous month. Looking only at generation, it seems the plant got worse. But the irradiance sensors show the month was also much cloudier. Calculating PR shows the plant converted the available sunlight just as efficiently as before — so there is no plant problem, only weather.

Now suppose a second month is sunny and generation looks healthy. Availability, however, shows that one inverter was offline for several days, and PR is lower than expected. The good weather masked a real equipment problem that cost energy.

That is the purpose of these metrics: they separate the weather from the plant.

Questions to ask your O&M provider

• How is PR calculated — temperature-corrected or not?
• Which irradiance sensors are used, and when were they last calibrated?
• How is availability defined — by time or by energy?
• Which events are excluded from availability, and how are they documented?
• How are partial outages, such as a single string or tracker row, counted?
• Can we see the raw data behind the monthly figures?

Looking at trends

Single months can mislead. Plotting PR and availability over many months reveals patterns:

• A gradual PR decline between cleanings, recovering after each cycle, points to soiling.
• A slow PR decline that does not recover may indicate degradation.
• Repeated availability dips at the same equipment suggest a root cause that has not been fixed.
• Seasonal patterns help separate climate effects from equipment issues.

How we help owners

As part of solar O&M coordination, we review O&M reports against these metrics, question unexplained losses, and help owners — including investors who have acquired operating plants — understand whether a plant is performing as it should.