Data recording during the maintenance process of solar panels is a key link in evaluating system performance, optimizing maintenance strategies and ensuring safe operation. The following are the precautions for data recording, which should run through the entire maintenance process:
First, the core objective of data recording
Performance tracking
Based on data such as power generation and attenuation rate, determine whether the component has reached its designed service life (for example, the annual attenuation rate within a 25-year service life is ≤0.7%).
For example, the annual power generation of a certain power station has decreased by 5% year-on-year. It is necessary to combine the data to investigate whether it is due to component aging, shading or inverter efficiency issues.
Fault traceability
Record the occurrence time, location and phenomena of faults (such as hot spot temperature and hidden crack location) to provide a basis for subsequent design improvement.
Case: In a certain area, junction boxes have been continuously burned out. Data records show that the temperature in this area has long exceeded 45℃, indicating that the heat dissipation design needs to be strengthened.
Compliance and Audit
Meet the grid connection requirements (such as power factor and harmonic data), or deal with the verification of insurance claims and subsidies.
Second, key data types and recording points
Power generation performance data
Record contents: Daily power generation, monthly power generation, system efficiency (PR value), irradiance, ambient temperature.
Notes:
Meteorological data (such as the calibration time of the irradiance meter) need to be recorded synchronously to avoid performance misjudgment caused by instrument errors.
When the power generation fluctuation exceeds 10%, the maintenance operations of the day (such as cleaning and component replacement) need to be recorded.
Electrical detection data
Record contents: insulation resistance, ground resistance, DC side voltage/current, inverter conversion efficiency.
Notes:
The insulation resistance test should be conducted in a dry environment, and the humidity during the test should be recorded (it is recommended that the humidity be less than 80%).
When the inverter efficiency is abnormal, the operating status of the cooling fan and the temperature of the air inlet need to be recorded.
Component status data
Record contents: EL test hidden crack rate, infrared thermal imaging hot spot temperature, glass light transmittance, and back plate color change condition.
Notes:
The EL test should be conducted at night or in a shaded environment to avoid interference from ambient light.
When the hot spot temperature exceeds the ambient temperature by 20℃, the component model, installation Angle and the status of adjacent components need to be recorded.
Structural and mechanical data
Record contents: Rusted area of the bracket, torque value of the bolt, deviation of the component inclination Angle.
Notes:
The bolt torque should be recorded at 80%-90% of the design value to avoid thread damage caused by excessive tightening.
When the deviation of the inclination Angle exceeds ±1°, the wind speed, wind direction and installation time of the day should be recorded (to determine whether it is caused by foundation settlement).
Third, norms and tools for data recording
Record Specification
Uniform format: Use standardized templates (such as PDF or Excel) to avoid ambiguity caused by free text.
Timestamp: All data must be recorded with precise time (such as “2023-10-15 14:30”) to facilitate the correlation of environmental conditions.
Signature confirmation: The operator must sign to confirm the authenticity of the data to prevent human tampering.
Recording tool
Portable devices: Use GPS positioning detectors (such as infrared thermal imagers, IV testers) to automatically record location and time.
Digital platform: Data is uploaded in real time through the SCADA system or operation and maintenance APP to reduce manual entry errors.
Image retention: Take photos of the faulty components (including the component number and the location of the fault), and file them for future reference.
Fourth, data storage and security
Storage method
Local backup: Use anti-magnetic and moisture-proof storage devices (such as encrypted hard drives) to retain data for at least five years.
Cloud storage: Select a cloud platform that complies with the ISO 27001 standard and set hierarchical access permissions (for example, operation and maintenance personnel can only view, and administrators can edit).
Data security
Encrypted transmission: Upload data via VPN or HTTPS protocol to prevent leakage.
Tamper-proof mechanism: Add digital signatures to key data (such as power generation) to ensure integrity.
Fifth, Data Analysis and Application
Trend analysis
By comparing historical data (such as year-on-year and month-on-month comparisons), potential problems can be identified.
Example: The power generation of a certain component has decreased for three consecutive months. Based on the EL test data, it is judged to be an expansion of hidden cracks.
Predictive maintenance
Predict the lifespan of components by using machine learning models (such as LSTM neural networks) and schedule replacements in advance.
Case: A certain power station shortened the anti-rust treatment cycle from 3 years to 2 years by analyzing the rusting rate of the support frame.
Decision support
Provide a basis for technological transformation (such as whether to add cleaning robots or replace high-efficiency components).
Example: Data shows that dust accumulation in a certain area led to a 15% loss in power generation, and it was decided to install an automatic spray cleaning system.
Sixth, Common Problems and improvement suggestions
Data missing
Reasons: Malfunction of the detection equipment, personnel negligence.
Improvement: Set up data integrity checks (such as the correlation check between power generation and irradiance), and for missing data, make up the measurement or note the reason.
Data conflict
Reason: The accuracy of multiple sets of detection equipment varies.
Improvement: Regularly calibrate the instrument (such as once every six months), record the calibration coefficients and correct the data.
Analysis lag
Reason: Manual statistics are inefficient.
Improvement: Introduce automated reporting tools (such as Power BI) to generate visual reports in real time.
