English is currently available.
Energy & Renewable Energy Engineering DISCUSSION
My new 400 W panels never reach their rated power. On a clear summer day at solar noon, with an air temperature of 35 °C, the monitoring shows about 290 W per panel. The panels are clean, unshaded and tilted toward the sun.
Is 72 percent of the nameplate normal, or is something faulty? I would like to know how the rating is defined and how to estimate what a healthy panel should produce in these conditions, so I can tell a real fault from expected behaviour.
The nameplate power is measured at standard test conditions: an irradiance of 1000 W/m², a cell temperature of 25 °C and a defined spectrum. Those conditions almost never occur together outdoors. When the sun is delivering 1000 W/m², the cells are typically 25 to 30 °C hotter than the air, so a 25 °C cell would need freezing weather with full sun.
Output is proportional to irradiance and falls with cell temperature. Both effects, plus the losses between the panel and the point where you measure, explain most of the gap you are seeing. Reaching the nameplate figure is unusual, and a brief excursion above it, on a cold clear day or when bright cloud edges add reflected light, is more likely than reaching it in summer heat.
Estimate the cell temperature from the NOCT value on the datasheet: T_cell = T_air + (NOCT - 20) × G / 800, with G in W/m². For NOCT = 45 °C, 35 °C air and G = 900 W/m²: T_cell = 35 + 25 × 900 / 800 = 63 °C.
That is 38 °C above the rating temperature. With a power coefficient of -0.35 percent per °C, the loss is 38 × 0.35 = 13.3 percent. Expected panel output: 400 × (900 / 1000) × (1 - 0.133) = 312 W at the panel terminals. Even a clear summer noon is often 900 to 950 W/m² rather than 1000, because of haze and the angle of the sun on the panel.
Then apply the losses downstream of the panel. Typical values are 2 to 3 percent for dirt even on panels that look clean, 1 to 2 percent in DC wiring, and 3 to 5 percent in the inverter or charge controller, with a further small loss from mismatch between panels in a string. Taking 3, 2 and 4 percent: 312 × 0.97 × 0.98 × 0.96 = 285 W.
Your 290 W is right on that estimate, so the panels are behaving normally. For comparison, a fault usually shows as one panel or one string well below its neighbours, not as the whole array sitting uniformly at 70 to 75 percent in hot weather.
If the monitoring is on a battery system, check one more thing before judging the array: a charge controller only draws what the battery and loads can accept. Once the battery reaches absorption or float, the controller moves away from the maximum power point on purpose and the reported PV power drops, however good the sun is. Measure in the morning with a partly discharged battery and a load running.
The angle matters too. Output follows the cosine of the angle between the sun and the panel normal, so a panel 30 degrees off loses about 13 percent of the direct beam. A fixed tilt is only square to the sun for a short time on a few days of the year.
For a real check of panel health, compare against the datasheet rather than the nameplate. Open-circuit voltage is the easy measurement: with the string isolated by its DC isolator, the voltage at the string terminals should be the rated Voc times the number of panels, reduced by the temperature coefficient, which means a few volts lower per hot panel. Treat the conductors as live, since a string in sunlight cannot be switched off, and leave short-circuit current tests to someone with DC-rated test equipment.
A panel with a shorted bypass diode or a cracked cell shows up as a voltage roughly one third lower than its neighbours, or as a hot spot in a thermal image. Shade from a cable, leaf or antenna on a single cell has a similar effect under load, because the shaded cell limits the current of its whole sub-string.
Back to Energy & Renewable Energy Engineering community
