One source of loss with a solar power system is the conversion of the DC power produced by the panels to the AC power used by the electric grid. This is done by a piece of equipment called an inverter, in my case an Eltek THEIA 3.8 HE-t UL. The manufacturer claims it is "up to 97.3% efficient," meaning that I'll lose at least 2.7% of the power produced by the panels. The specification for the inverter includes an efficiency curve showing how the efficiency varies depending on the amount of power being produced.
Spec efficiency curve
I was curious to see how my setup compared to the spec, so I pulled data for December 2013 through October 2014. I don't have the instantaneous numbers that are used in the spec (it would be quite a lot of work to collect and process all of that) and I don't have access to the direct data on the inverter, so I compared the DC power output from the Tigo power optimizers on the panels to the inverter output measure by my eGauge. I used daily totals for both numbers, then created a scatter plot of the efficiency versus the total power output for the day.
Actual efficiency curve
- Minimum: 58.45%
- Maximum: 95.08%
- Mean: 91.76%
- Median: 93.90%
The numbers are generally in line with with the spec though they tend to be a few percent lower. The low efficiency for low production days is one of the reasons that the winter is such a tough time for solar, since anywhere from 15 - 40% of the power coming from the panels can be lost due to conversion. Luckily that only works out to a handful of kilowatt-hours, and isn't a problem in the summer.