A monitoring app shows lower energy than it did during a recent period. Dust is visible from the ground, so cleaning looks like the obvious answer. Someone else suspects aging modules, while the installer mentions inverter limits. Each explanation can be plausible. None is proven by one weekly total.
Solar panel loss of efficiency is often used as a broad label for several different changes. Module efficiency is a conversion characteristic under stated conditions. Power is output at one moment. Energy is accumulated production over time. Lower daily energy can come from weaker sunlight, higher cell temperature, shade, soiling, operating limits, downtime or missing monitoring records even when module conversion efficiency has not suddenly changed.
The useful response is to locate the missing energy before paying for cleaning, service or replacement. A short evidence record can prevent an unnecessary visit, but it can also expose a persistent fault before more production is lost. The sequence is practical: establish a fair baseline, document the pattern, choose a safe check and then select an intervention that matches the evidence.
Separate Efficiency Power and Energy
Nameplate power is not a promise that an array will reach its rated value every sunny afternoon. It is measured under defined test conditions. Field conditions change continuously, and the alternating current energy shown in a portal also reflects the inverter, wiring, controls and system availability.
The United States Department of Energy performance guidance states that module ratings use 1000 watts per square metre of sunlight and a cell temperature of 25 degrees Celsius. It also explains that actual output should be compared with expected output when identifying underperformance.
That changes the first question. Do not ask why the array failed to match its rated direct current power at a random moment. Ask whether it produced less energy than a realistic expectation for the same season, solar conditions and operating state. A hot afternoon should not be compared with a cool morning. A cloudy day should not be compared with a clear day. A period with a full battery or active export limit may not have the same operating boundary as a period when the system could accept all available solar power.
A useful baseline can come from the original production model, a weather adjusted portal estimate or the same month in a previous year. Each has limits. The purpose is not laboratory certainty. The purpose is to decide whether the difference is persistent, financially relevant and unexplained enough to justify deeper diagnosis.
Build a Seven Day Evidence Record
Record seven consecutive days before changing the site, unless there is a safety warning, visible electrical damage, unusual heat, a burning smell or a persistent equipment fault. The record should follow the energy path from sunlight to the portal. That prevents every lower alternating current value from being blamed on the modules.
Match Weather and Time
Use the same terms each day for sky condition, smoke, dust, rain and new shade. Compare matched hours on days with similar conditions. Daily energy is usually more useful than one peak because a brief clear interval can create a high momentary value without representing the full day. Keep the portal time zone and missing intervals visible in the record.
Season matters as well. A month to month decline can follow sun angle and day length, while the same month across years is a better starting comparison. For a small commercial site, note operating hours and export restrictions. The same energy loss can have different financial impact depending on whether the site would have consumed the power or exported it at low value.
Record Surface Conditions
Heat, soiling and shade often overlap. Record visible buildup, new shade and the hours when output changed. Add a photograph from a safe location with a timestamp. A general statement that the panels look dirty gives little diagnostic value unless the condition aligns with the affected area and time.
The IEA PVPS fact sheet on soiling losses reports that soiling is responsible for an average 4 to 7 percent of global photovoltaic energy losses. This is industry context, not a forecast for one roof. Rain pattern, dust type, tilt, nearby activity, access and cleaning cost all influence the local decision.
A cleaning decision needs a before and after comparison under similar conditions. If the production pattern does not improve, repeated cleaning is not a diagnosis. Shade work also needs evidence that the obstruction affects the same hours as the measured loss and that any removal is permitted and safe.
Track Conversion and Availability
The array can produce direct current while the inverter, protection system, grid connection, battery state or export control limits usable energy. Record inverter status, alarm codes, restart events, battery state, power ceilings and any period when the monitoring platform was offline. A flat top can reflect an operating limit. A blank graph can reflect lost communication rather than zero production.
The official page for IEC 61724 Part 1 outlines terminology, equipment and methods for photovoltaic performance monitoring and analysis. The practical buyer lesson is consistency: use the same measurement point, time interval and operating definition when comparing results.
Record field | What to capture | Why it matters |
Date and weather | Sky condition rain smoke dust and temperature context | Supports a fair period comparison |
Shade and surface | Affected area time and safe photograph | Tests whether the visible condition matches the loss |
Peak power | Value and timestamp | Shows ceilings and short operating changes |
Daily energy | Complete total and missing intervals | Supports matched day comparison |
Equipment status | Alarm code restart and inverter state | Separates module questions from conversion downtime |
Operating constraint | Battery export or grid limit | Explains intentional power restriction |
Classify What the Curve Is Telling You
After the record is complete, place the pattern in a decision lane. These lanes are not universal fault thresholds. They are a way to keep evidence, safety and financial consequence connected.
Pattern | Evidence clue | First safe action | Escalation point |
Normal variation | Output follows weather season or a known limit | Continue matched monitoring | Decline remains unexplained |
Likely reversible loss | Visible soiling or shade aligns with affected hours | Document and assess safe professional removal | No improvement after a comparable check |
Monitoring problem | Portal is offline while local status appears normal | Preserve timestamps and check communication status | Data remains absent or status is unclear |
Possible system fault | Sudden site or local drop repeated alarms or abnormal heat | Stop physical intervention and request qualified service | Immediate professional diagnosis |
The normal lane still requires attention. It means the system remains within an explainable operating pattern, not that monitoring should stop. The reversible lane requires proof after the intervention. The fault lane has a different boundary. Repeated protective trips, persistent fault status, damaged cable, unusual heat or a burning smell are not cleaning questions. Follow the approved isolation procedure and contact qualified service.
Follow the Symptom Before Choosing the Tool
A missing graph and zero production are different symptoms. If the app has not updated, check the time of the last valid record and whether the network, account or gateway changed. If communication is live and the system reports zero production during good daylight, record the display, indicator state and alarm code without opening equipment. Persistent zero output with live communications belongs in a service case.
When the whole system follows the same lower curve, review weather, temperature, system limits and inverter availability. Look for flat ceilings, missing intervals and a change that began after a grid event, settings update or restart. A gradual decline should be compared across the same season and against a weather aware expectation. One hot week does not establish long term module degradation.
A low string or roof area points away from a whole site weather explanation. Record the affected input or area exactly as the monitoring platform labels it. Capture the period when the difference is clearest and photograph visible shade or debris from a safe location. This gives the technician a test target instead of a general low output complaint.
Use Installed Product Data to Test Heat
Heat is often blamed when sunny afternoon output is lower than a cool morning peak. The useful test starts with the coefficient from the installed module model. A generic percentage for solar panels cannot establish the expected response of a particular array.
Our current SNADI/SNAT Solar panel lists 200W, 300W, 430W and 590W N Type monocrystalline options with model specific electrical and temperature values.
For the 430W and 590W options, a maximum power temperature coefficient of minus 0.350 percent per degree Celsius and a nominal operating cell temperature of 45 plus or minus 2 degrees Celsius. It’s standard condition voltage, current, dimensions and weight for each option. Those figures are inputs for an engineering comparison. They do not prove that heat caused a measured site loss.
A defensible heat check still needs operating temperature or a reasonable estimate, solar resource context, array configuration, inverter limits and a clear measurement point. Lower alternating current energy can combine module temperature with conversion loss and intentional operating constraints. We use product data to define expected behaviour, then use monitoring evidence to decide whether the measured result is normal.
Prepare the Evidence Handoff
A focused service request should let the reviewer see the operating boundary before requesting another visit. Include these items.
1. Seven consecutive days of complete production records with the same time zone and measurement point.
2. Weather and shade notes for the hours when the change is visible.
3. Safe photographs of visible soiling shade damage or site changes with dates.
4. Inverter gateway and monitoring timestamps with alarm codes and restart history.
5. Exact module inverter battery and monitoring device models.
6. Original production estimate commissioning record and relevant warranty documents where available.
7. Recent cleaning network grid settings construction or equipment changes.
Do not delay qualified service to complete the record when there is heat, odor, electrical damage, exposed wiring or a persistent protection event. In every other case, a consistent handoff helps SNADI/SNAT Solar engineering support define whether the next task is data recovery, system review, cleaning verification or professional fault testing.
Conclusion
Solar panel loss of efficiency is a diagnosis question, not an immediate reason to clean or replace equipment. Separate module efficiency from power and accumulated energy. Build a comparable seven day record, classify the pattern and follow the symptom shown by the data. Normal variation needs continued comparison. Visible reversible loss needs a measured intervention. Persistent alarms, local imbalance or unexplained downtime needs qualified testing.
Replacement belongs at the end of the path, after the installed model, operating boundary, warranty and system compatibility are verified. Send our SNADI/SNAT Solar engineering team the production record, screenshots, alarm codes, module and inverter models, system capacity and recent changes. We can help define the review scope .
✉️Email: marketing@snadi.com.cn
Website:
☎️WhatsApp / WeChat: +86 1803929353
FAQ
The observed loss may come from lower sunlight, heat, soiling, shade, inverter limits, downtime, monitoring gaps or gradual module change. Compare energy with a realistic baseline before deciding that the modules lost conversion efficiency.
How can solar panel performance be checked safely?
How can I tell whether panels are dirty or faulty?
Does lower daily energy prove module degradation?
Can the inverter make solar output look low?
When should a warranty case be opened?