Availability and performance ratio answer related but different questions. Availability isolates whether the solar plant was capable of producing during the period that counts. Conventional performance ratio normalizes the plant’s total AC energy against rated capacity and measured solar irradiation, so downtime can lower it along with losses that occur while equipment is operating. Owners need both because availability helps explain how much of a performance result may be tied to lost operating opportunity.
Three metrics, three owner questions
Terms and exclusions vary across monitoring platforms, standards, warranties, and service agreements. The calculation written into the applicable contract is controlling. Still, the framework used in the U.S. Department of Energy's FEMP performance guidance and its assessment of federal PV systems is useful for separating the questions.
Swipe to compare
| KPI | Owner question | What a weak result suggests |
|---|---|---|
| Availability | Was the plant capable of delivering power when the defined conditions were present? | Downtime, trips, outages, delayed restoration, or a classification/data problem |
| Conventional performance ratio | How much total AC energy was delivered per rated capacity relative to measured solar irradiation? | Downtime, operating losses, capacity error, irradiance-sensor error, or inconsistent calculation rules |
| Measured-to-expected energy ratio | How much energy was produced compared with the modeled total for the period? | Combined variance from downtime, operating underperformance, model inputs, or data quality |
NREL's assessment of 75 federal PV systems reports multiple metrics because availability supplies context that a performance percentage cannot provide alone. Conventional performance ratio and a total measured-to-expected energy ratio can both decline during outages; event records and an explicitly availability-adjusted calculation are needed to isolate performance during available intervals.
What availability can—and cannot—tell you
Availability is commonly time-based or energy-based. A time-based calculation measures the share of eligible time that equipment or the plant was available. An energy-based calculation weights outages by the production opportunity that was lost. One midday inverter outage may therefore matter much more than several nighttime hours, even if a simple time count treats every hour equally.
The denominator and exclusions matter. A contract may exclude nighttime, grid outages, curtailment, force majeure, planned work, tracker wind stow, or periods with insufficient irradiance. It may define whether partial derating counts as unavailable and how missing data are handled. The IEA PVPS technical and economic KPI guideline emphasizes that both time-based and energy-based availability are used and that treatment of exclusions and missing data affects the result.
Low availability therefore establishes that opportunity was lost under the selected rules; it does not identify the failed component. An inverter trip, utility interruption, station-service loss, communications outage misclassified as downtime, safety shutdown, or delayed response can produce similar dashboard results. Review alarms, independent revenue-meter data, work orders, and event timelines before assigning cause.
What performance ratio can—and cannot—tell you
The traditional IEC performance ratio normalizes AC energy for the plant's rated capacity and measured plane-of-array irradiation. Sandia's PV Performance Modeling Collaborative describes it as final AC yield divided by reference yield. Some owner reports instead use measured production divided by weather-corrected model production during available intervals. These approaches are related, but values from different methods should not be mixed without reconciliation.
A falling performance ratio can be consistent with:
- Plant or equipment downtime included in the reporting period.
- Soiling, shading, snow, or another change in the solar resource reaching the modules.
- High module temperature, tracker or racking position problems, mismatch, degradation, or DC losses.
- Inverter clipping, derating, conversion loss, transformer loss, or an operating constraint.
- An incorrect plant configuration, capacity value, degradation assumption, or expected-energy model.
- Drift, failure, shading, or misalignment in an irradiance or temperature sensor.
Performance ratio is not a root-cause code. It combines many losses into one normalized value, and the conventional metric remains sensitive to temperature. Sandia notes that annual plots commonly show lower values in warm periods and higher values in cooler periods. For Central Valley assets, a month-to-month change should therefore be compared with a consistent model and prior-year conditions—not interpreted from a generic target alone.
How the two KPIs combine
Swipe to compare
| Availability | Performance ratio | Practical interpretation |
|---|---|---|
| High | High | The plant is generally online and converting the available resource near the established expectation; still review total energy, data quality, and emerging trends. |
| Low | High | Outages may have occurred during low-resource periods, may be excluded from the performance-ratio calculation, or may be diluted over the reporting interval. Read the definitions before concluding that available-period performance was strong. |
| High | Low | The plant stays online but produces less than expected. Focus on weather inputs, sensors, environmental losses, derating, and DC/AC performance. |
| Low | Low | Both downtime and operating loss may be present—or data and calculation rules may be corrupting both results. Validate the dataset before dispatching broadly. |
NREL's PV O&M best-practices guide treats availability and weather-adjusted performance as separate but related evidence. If a report presents performance only during available intervals, it should label that adjustment explicitly; otherwise, owners should assume the conventional period-energy calculation may include outage effects.
Validate the KPI before diagnosing the array
A trustworthy KPI requires aligned timestamps, correct plant capacity and topology, reliable production metering, representative irradiance and temperature data, and documented treatment of curtailment, outages, clipping, missing records, and equipment changes. IEC 61724-1 defines monitoring-system classes, measurements, and analysis methods because data quality is part of performance analysis—not a housekeeping detail.
- Confirm the reporting boundary. Identify the meter, capacity, equipment, time zone, interval, and period included.
- Read the calculation definition. Record exclusions, thresholds, expected-energy method, degradation, and missing-data treatment.
- Test data completeness. Look for gaps, flat lines, duplicate timestamps, implausible sensor values, and meter resets.
- Compare independent evidence. Reconcile inverter totals, revenue meter, utility data, alarms, weather, and work orders where available.
- Segment the loss. Separate full outages, partial outages, operating underperformance, curtailment, environmental effects, and data uncertainty.
- Target field diagnostics. Use the evidence to select the right inspection, thermal scan, I-V curve tracing, inverter troubleshooting, or metering work.
Turn the dashboard into an owner-ready report
A decision-ready report should show the KPI definition beside the result. It should also show measured and expected energy, lost-energy estimates by cause, excluded intervals, unresolved data gaps, open work orders, and the financial or operational priority. Trend the same calculation over time and preserve changes to sensors, equipment, models, and plant capacity.
For solar asset management, the most useful question is not “Did the site pass?” It is “Which controllable loss deserves action, how certain is the diagnosis, and how will restored performance be verified?” TerraSun's performance optimization process connects analysis to field evidence and repair closeout rather than stopping at a monthly percentage.
Safety and operational boundary
Owners can review reports, confirm timestamps, gather bills and work orders, and ask how the KPI was calculated. Do not open electrical enclosures, defeat interlocks, reset recurring faults, change inverter controls, or perform energized testing to investigate a score. Those tasks belong to personnel qualified for the equipment and hazards, working under approved site procedures and manufacturer instructions.
Owner takeaway
Availability tells you how much eligible operating opportunity was retained. Conventional performance ratio shows total AC energy relative to rated capacity and measured irradiation, which means outages and operating losses can both influence it. A measured-to-expected energy ratio provides another whole-period comparison, while an explicitly availability-adjusted metric can isolate available-period performance. Review the definitions, validate the inputs, and segment the loss before assigning field work.
Official technical references
- U.S. DOE FEMP: Operate and Maintain an Existing Photovoltaic System
- NREL: Understanding Solar Photovoltaic System Performance—An Assessment of 75 Federal Photovoltaic Systems
- NREL: Best Practices for Operation and Maintenance of Photovoltaic and Energy Storage Systems
- Sandia PVPMC: Performance Ratio
- IEA PVPS Task 13: Best Practice Guidelines for the Use of Economic and Technical KPIs
- IEA PVPS Task 13: Guidelines for Operation and Maintenance of Photovoltaic Power Plants in Different Climates
- IEC 61724-1:2021: Photovoltaic system performance—Monitoring
- IEC TS 63019:2019: Photovoltaic power systems—Information model for availability
Common questions
What is the difference between solar performance ratio and availability?
Availability describes whether the system was capable of operating during the applicable period. Conventional performance ratio normalizes total AC energy against rated capacity and measured solar irradiation, so it can reflect both outages and losses while the plant is operating. The written calculation and exclusions determine the interpretation.
Can a solar plant have high availability and still underperform?
Yes. A plant can remain online while producing less than expected because of soiling, shading, thermal effects, tracker problems, DC mismatch, inverter derating, sensor error, or other losses.
Does low availability identify the failed component?
No. Low availability shows that operating time or energy opportunity was lost, but alarms, meter data, service records, site conditions, and qualified diagnostics are needed to determine the cause.
What is a good performance ratio for a commercial solar system?
There is no universal target for every site. The useful benchmark is the value expected from the system design, climate, age, equipment, measurement method, and contract definition, compared consistently over time.
Why can performance ratio change with the season?
The conventional performance ratio is sensitive to module temperature and to the quality of irradiance measurements. Hot modules generally operate less efficiently, so seasonal comparisons need consistent methods and appropriate context.
Which KPI best shows the owner's total energy loss?
A total measured-to-expected energy comparison captures both downtime and reduced output, but owners should still review availability and performance ratio separately because they lead to different corrective actions.