Solar Monitoring & Performance Ratio: Expected vs Actual Production

Welcome

Hello, and welcome. This is Super Structures General Contractors — a national general contractor headquartered in Powhatan, Virginia — here to help you and your clients build something that lasts. We're glad you're with us, and we look forward to connecting with you.

A solar system can lose a third of its output and the owner might never notice — the lights still come on. Monitoring is how you catch quiet failures before they cost months of production. This lesson teaches you to read the data like a pro.

Educational content only — not electrical design, a safety program, or hands-on testing and troubleshooting training, and not a substitute for a licensed electrician or a NABCEP-certified professional. Testing energized PV equipment is hazardous. Follow the NEC as adopted by your Authority Having Jurisdiction (AHJ), NFPA 70E and OSHA, your company's electrical safety program, the manufacturers' instructions, and your utility's rules. Code and standard editions change — confirm the ones that apply to your work.

What monitoring platforms show

Almost every modern inverter, microinverter, and optimizer system includes a monitoring platform — a cloud dashboard fed by a gateway or the inverter's communication link. Depending on the equipment, it shows:

• System-level energy (kWh) and power (kW) over time
• Inverter-level data: AC output, DC input, status, and error codes
• Module- or string-level data for microinverters, optimizers, or string monitoring
• Alerts for faults, communication loss, and low production

Commercial systems add a data acquisition system (DAS) with weather stations: irradiance sensors (pyranometers or reference cells), module temperature sensors, and ambient temperature, plus revenue-grade meters.

Expected vs actual production

Production varies daily with weather, so "it made less today" means nothing on its own. The right question: did it produce what it should have for the sunlight it received? Expected production comes from:

• The design model — the annual and monthly estimate from the original design software, which used typical weather data.
• Weather-adjusted expectation — the model corrected for the actual sunlight measured by on-site sensors or satellite data.
• Peer comparison — comparing inverters, strings, or modules in the same array against each other. If ten strings match and one is 20% low, that one has a problem regardless of the weather.

Performance ratio (PR)

Performance ratio is the standard measure of system quality, independent of how sunny it was:

PR = actual AC energy ÷ (rated DC capacity × measured irradiation ÷ reference irradiance)

In plain language: of the energy the array should have produced given its nameplate and the sunlight it saw, how much made it to the meter? Losses from temperature, wiring, inverter conversion, soiling, shading, mismatch, and downtime all reduce PR. A well-built system typically lands somewhere in a broad range around the high 70s to mid-80s percent, varying with climate and technology. What matters most is the trend: a falling PR over weeks signals something is wrong.

Specific yield

Specific yield — kWh produced per kWp installed over a period — lets you compare systems of different sizes in the same region. It does depend on the weather, so compare like periods and nearby sites.

Reading alerts and patterns

• Zero production from one inverter or microinverter: device failure, tripped breaker, or communication loss — check whether the meter shows production before rolling a truck.
• Gradually declining output: soiling, degradation, or vegetation growing into the array.
• Output that flattens at the same value every midday: inverter clipping — normal by design when the DC array is larger than the inverter, but excessive clipping or clipping on a cool day can indicate a setting issue.
• A dip at the same time each day: a new shade source — a growing tree, a new building, or a rooftop unit.
• Low output on hot days only: normal temperature losses, or poor airflow behind modules.

A remote triage routine

Before dispatching a technician, work through the data remotely:

1. Is it real? Confirm the system is reporting at all — rule out a communication outage first.
2. How widespread? One module, one string, one inverter, or the whole system? Whole-system drops often point to the grid, a tripped breaker, or the rapid shutdown initiator.
3. When did it start? Match the start time to weather, a utility event, or work on the building.
4. What does the inverter say? Read the error log and look up codes in the manufacturer's service manual.
5. Can the customer help? Ask them to check breakers and the inverter display safely — never to open equipment. A good triage turns a two-visit repair into one visit with the right parts on the truck.

Going Deeper (Intermediate)

Communication loss is the most common monitoring "fault" and often has nothing to do with the solar — the homeowner changed internet routers or Wi-Fi passwords. Train customers on how to reconnect, and use cellular gateways where reliability matters. A system with no data is a system you can't prove is working.

Advanced / Pro-Level

On commercial fleets, analysts calculate weather-corrected PR, availability (the share of time equipment was able to produce), and lost energy by cause. Contracts often guarantee availability or energy, with liquidated damages for shortfalls. Pros keep sensors clean and calibrated — a dirty irradiance sensor makes the array look better than it is. They also separate expected degradation (a small annual percentage stated in the module warranty) from abnormal loss.

Practice Challenge

A 10 kW system's monthly production is 18% below the design estimate. The month was unusually cloudy, and all 30 microinverters report similar output. Is there a problem? (Answer: probably not with the equipment — uniform output across all units plus a cloudy month points to weather. Compare with weather-adjusted expectation or nearby systems. If the weather explains the gap and peers match, document it and keep monitoring.)

In Practice

Set up alerts for every system you install, check fleet dashboards on a schedule, and contact the customer before they notice a problem. Proactive service turns a failed device into a referral instead of a complaint.

Common Mistakes to Avoid

• Judging performance from one day's production
• Comparing to the annual estimate without adjusting for weather
• Ignoring communication-loss alerts until the customer calls
• Letting irradiance sensors get dirty or drift out of calibration
• Mistaking normal clipping or temperature losses for faults

Takeaway: Judge production against what the sunlight allowed, not against yesterday — use performance ratio, weather-adjusted expectations, and peer comparison to spot real problems early.