The principle: compare like with like
A string has no absolute “correct” value, but it should behave like its neighbors if it has the same modules, the same orientation and the same conditions. Locating the problem relies on that comparison: look for the string that deviates from the group, then find out why.
Step 1: start from the data
If the system monitors current per string or per inverter input, compare the daily profiles of equivalent strings, normalized for irradiance. A string that always stays below the others, or that drops at certain hours, already gives direction: a drop at the same time every day is usually a shadow. If measurements are taken only at inverter level, the analysis is less detailed and you have to go down to the field.
Step 2: field measurements
- Open-circuit voltage: a voltage lower than that of neighboring strings suggests missing modules, conducting bypass diodes or damaged cells.
- Current: lower current with normal voltage suggests soiling, shading or uniform degradation.
- I-V curve with irradiance and temperature logged: steps indicate mismatch or active diodes; a rounded knee, high series resistance.
Step 3: inspection
Thermography under sufficient irradiance reveals hot cells or modules, high-resistance connections and conducting bypass diodes. Visual inspection looks for broken glass, discoloration, soiling, shadows from new objects, burned connectors and animal damage.
Frequent causes
- Damaged connectors or cables, with high resistance or intermittent contact.
- Degraded modules or modules with microcracks.
- Short-circuited bypass diodes.
- Localized soiling or shading.
- Ground fault or leakage.
- A problem with the inverter input or string fuse.
What to check first
The cheapest things first: the string’s protection devices and connectors, then shading and soiling, then measurements. Field measurements require working with DC voltage present as long as there is light, so they must be carried out by qualified, authorized personnel under the applicable safety regulations. All comparisons must be made under stable irradiance conditions.