Solar inverter efficiency, on a datasheet, is alternating-current power out of the inverter divided by direct-current power into it, at a stated condition. A 1 percent gap between two brochures is meaningless until you know whether both numbers are peak, whether both are weighted across a load range, and whether either one includes the night. Module efficiency is a different ratio, measured on the panel. Do not add the two percentages and call the sum the system.
Peak, weighted, and the load you actually run
Peak efficiency is the best point on that inverter’s curve. Inverters spend much of the year away from that point, because the array is not at full power in the morning, in cloud, or in heat. A weighted figure, when the sheet has one, samples several load levels and sums them with weights someone published. A European weighted figure and a California Energy Commission weighted figure are not the same recipe. If the sheet does not name the weighting, do not invent one.
The rest of the lines around that percentage, including the MPPT window and the maximum input, are in how to read an inverter datasheet. Clipping, which throws away array power above the AC rating, is not an efficiency failure. It is a sizing outcome, written up in inverter clipping and the DC-to-AC ratio.
| Figure | What it is | What it is not |
|---|---|---|
| Peak efficiency | The highest point of output over input on that test | The efficiency at 8 a.m. |
| Weighted efficiency | A stated mix of part-load points | A prediction of your annual bill |
| Night or standby consumption | Watts drawn when the array is not producing | A reason, by itself, to switch the inverter off each evening |
| Module efficiency | Module power over light power at standard test | Something you add to the inverter percentage |
Heat and light move the operating point
A hot inverter may derate before it reaches the load where the peak efficiency was measured. The curve for that is on the manual, and the idea is in heat, altitude, and inverter derating. A module that is hot produces less DC, so the inverter sits at a lower load, where its efficiency may be worse than the peak. Both effects are real. Neither one is a license to subtract a made-up percent from the nameplate and call it the year’s energy.
Night consumption belongs on the datasheet as a wattage, not as a rumor. Whether that wattage justifies a nightly shutdown is already answered in should you turn a solar inverter off at night. For a listed grid-tied inverter the energy argument is usually small. The safety argument for leaving it in its normal state is not.
A percentage that cannot close a purchase
Efficiency does not choose the waveform, the battery voltage, the backup transfer, or the service the inverter may connect to. A modified-sine inverter can print a flattering efficiency because the extra heat showed up in the motor, not in the inverter’s own meter. Waveform is in pure sine versus modified sine. If two inverters are both suitable and both honestly specified, the weighted figure is one input to the choice. It is not the choice.
