Solar Monitoring: Reading Production Graphs and Spotting Faults

Solar monitoring is the practice of tracking a photovoltaic (PV) system’s energy production over time. The primary purpose is to verify that the system is performing as designed and to identify faults early. This guide explains what a normal production graph looks like, how to interpret deviations, and how to distinguish between expected variations and actual problems. It is written for system owners and operators who have access to monitoring data, typically through an inverter’s web portal or a third-party monitoring service.

Monitoring data is usually presented as a graph of power output (in kilowatts, kW) over time, or as energy production (in kilowatt-hours, kWh) over days, months, or years. The shape of these graphs carries information about the system’s health and the conditions it operates under.

What a Normal Day Looks Like

A typical daily production curve for a grid-tied system without shading follows a bell-shaped curve, peaking around solar noon. The exact shape depends on your latitude, season, and weather. Here are the key features:

• Start and end times: Production begins shortly after sunrise and ends shortly after sunset. The curve is symmetric around solar noon, which is not necessarily 12:00 PM local time.
• Peak power: The maximum power output occurs when the sun is highest in the sky, assuming clear skies. This peak should roughly match the system’s rated capacity under standard test conditions (STC), but it will be lower due to real-world conditions (e.g., temperature, soiling, inverter efficiency). A typical derate factor is 0.75–0.85, meaning a 5 kW system might peak at 3.75–4.25 kW on a clear day.
• Smoothness: On a clear day, the curve is smooth. Small wiggles may be due to passing clouds or inverter maximum power point tracking (MPPT) adjustments.
• Seasonal variation: In summer, the curve is higher and wider; in winter, it is lower and narrower. This is normal and expected.

Interpreting Production Graphs

Daily Production Graph

A daily graph shows power output throughout the day. Look for:

• Clouds: Passing clouds cause dips in the curve. The dips may be sharp if clouds are thick, or gradual if they are thin. The system will recover quickly after the cloud passes.
• Shading: If a tree or chimney shades part of the array, you will see a distinct step-down in power at the time of day when the shade falls on the panels. The step may be flat or sloped, depending on the shading pattern. Unlike clouds, shading occurs at the same time each day and is predictable.
• Inverter clipping: If the inverter’s maximum AC output is lower than the DC power from the array, the production curve will have a flat top during midday. This is called clipping. It is a design choice, not a fault, and results in a small annual energy loss (usually 1–3%).

Monthly and Annual Production

Monthly production varies with season. In the Northern Hemisphere, production is highest in May–July and lowest in December–January (assuming no snow). Annual production should be within 5–10% of the first year’s production, with a slight decline each year due to panel degradation (typically 0.5% per year). A sudden drop in monthly production compared to the same month last year may indicate a problem.

Normal vs. Faulty Behavior

Common Faults and Their Signatures

String Faults

In a string inverter system, panels are connected in series. If one panel fails (e.g., a bypass diode short), it can bring down the entire string. The production graph will show a significant drop (e.g., 50% if one of two strings is down) that is constant across the day. The drop may be present from the start of the day, or it may appear suddenly.

Microinverter or Optimizer Faults

With microinverters or power optimizers, each panel is monitored individually. A faulty microinverter will show a panel producing zero or significantly less than its neighbors. The production graph for that panel will be flat or low. This is easy to spot if you have panel-level monitoring.

Inverter Faults

Inverter failures often result in zero production. The inverter may display an error code, or the monitoring portal may show a fault. Some inverters have a “derating” mode that reduces output due to high temperature or grid voltage issues. This will show as a lower peak or a curve that is clipped at a lower level.

Soiling

Dust, bird droppings, or leaves on panels reduce output. The effect is more pronounced on clear days and can be seen as a uniform reduction in power across the day. The reduction is proportional to the soiling ratio. For example, 10% soiling means 10% less power. Soiling is often visible in the morning or after a dry spell. Rain usually cleans panels, so a sudden recovery after rain indicates soiling was the cause.

Degradation

Panels degrade over time, typically 0.5% per year. This is a slow decline, not a sudden drop. If you compare annual production, you should see a gradual decrease. A sudden drop of more than 5% from one year to the next is not normal degradation.

What People Get Wrong About Solar Monitoring

Mistaking Clouds for a Fault

A common mistake is to see a dip in the production graph and assume the system is broken. Clouds are the most frequent cause of dips. The key is to check the shape: clouds cause irregular dips that vary in depth and duration. Shading causes a predictable step at the same time each day. If the dip is not repeatable, it’s likely clouds.

Expecting Peak Power at Noon

Another misconception is that the system should produce its rated capacity at solar noon. In reality, real-world conditions (temperature, soiling, inverter efficiency) reduce output. A 5 kW system might never reach 5 kW. The rated capacity is under standard test conditions (25°C, 1000 W/m² irradiance, air mass 1.5), which rarely occur in practice. So, a lower peak is normal.

Ignoring Seasonal Variation

Some owners compare summer production to winter and think something is wrong. The sun’s path changes, so production is naturally lower in winter. The same system might produce 30% less in December than in June. This is not a fault.

Assuming Monitoring Data is Always Accurate

Monitoring systems can have errors. CTs (current transformers) may drift, communication may fail, or the portal may show incorrect values. If the data looks off, verify with the inverter’s display or a separate meter. Also, some monitoring systems report AC production, while others report DC. Know which you are looking at.

How to Use Monitoring to Verify Performance

Compare to a Baseline

The best way to know if your system is performing is to compare current production to a baseline. This could be the first year’s production, or a modeled estimate. Many monitoring portals provide a “expected” or “estimated” production based on weather data. If actual production is consistently below expected by more than 10%, investigate.

Use the Right Metrics

• Specific yield (kWh/kWp): Total energy produced per kilowatt of installed capacity. This allows comparison across systems. A typical value is 1,200–1,600 kWh/kWp per year in sunny regions, but it varies widely.
• Performance ratio (PR): The ratio of actual output to expected output, accounting for irradiance and temperature. A PR of 0.75–0.85 is typical. If PR drops below 0.7, there may be an issue.

Check for Alerts

Most monitoring systems have alerts for faults. Set up email or app notifications for inverter errors, communication loss, or unusually low production. But be aware that some alerts are false positives (e.g., a brief grid outage).

Commands for Monitoring Data

If you have access to the inverter’s API or a data logger, you might use command-line tools to fetch data. Here are examples:

# Fetch current power from a SolarEdge inverter (using Python)
python -c "import requests; print(requests.get('http://<inverter-ip>/api/status').json())"

# This returns JSON with current power, voltage, etc.
# Look for the 'power' field in the response.

# Fetch daily energy from an Enphase system (using the Enphase API)
curl -H "Authorization: Bearer <token>" "https://api.enphaseenergy.com/api/v2/systems/<system_id>/energy_lifetime"

# This returns a list of daily energy values.
# Compare the latest value to previous days to spot a drop.

# Use a generic tool like `curl` to query a Fronius inverter
curl "http://<inverter-ip>/solar_api/v1/GetInverterRealtimeData.cgi?Scope=System"

# The response includes 'P_AC' (AC power) and 'E_Day' (daily energy).

These commands are examples; the exact endpoints and parameters vary by manufacturer. Always refer to the manufacturer’s documentation.

When to Call a Professional

If you suspect a fault, first check the basics: is the inverter on? Is there a grid outage? Are the breakers tripped? If everything looks fine but production is still low, contact a licensed solar installer. Do not attempt to open the inverter or work on the DC wiring. PV systems produce dangerous DC voltage whenever there is light, and there is no switch to turn it off at the panel. Only qualified personnel should work on the DC side.

Local Regulations and Incentives

Net metering, interconnection, export compensation, and incentives are set by state and utility and change over time. They affect the financial value of your production but not the physical performance. Check with your utility and local authorities for current rules. Monitoring data can help you verify that you are being credited correctly.