I have been looking at this in a little more detail. There is definitely a discrepency between string and micro inverters in a shaded situation. The micro inverters seem to underperfom the string inverter by a clear margin. But interestingly the micro inverters show an increase in yield when run through the software separately as opposed to in an array formation. To do this I set up the array as normal in 3D mode and then deleted all but one panel and assigned 1 micro inverter, then output the results. I did this in turn for each of the 14 panels in the array that I am working on and the results were as follows (shading is a big problem in this instance):
14 no Sharp HIT H250-EO1 250W panels with SMA SB 3300 = output 2,787.6 kWh
14 no Sharp HIT H250-EO1 250W panels with 14 no Enecsys SMI-S240W-72-UK micro inverters = output 2,586.8 kWh
14 no Sharp HIT H250-EO1 250W panels with 14 no Enecsys SMI-S240W-72-UK micro inverters (run individually through PV Sol 1 at a time then output totalled manually) = output 2,776.78 kWh
I have reported this to The Solar Design company and have sent through the project files for them to look at, though at the moment they cannot offer an explanation as to why this is happening. I am left wondering whether the individual totals are correct or whether they are still off the mark?
I had hoped that the micro inverters would be the answer in this situation as the customer is very keen on PV, but the combined output is dissapointing. Does anyone else have a better way of estimating yield from micro inverters? It will be difficult to justify the expense when the yield forecasts are no better than a string inverter. There is little assistance on manufacturer's site and in documentation.
