MicroInverters Problems So Far ?

[LucMan]

My 5.2 KW system is now 7 years old. When first installed I had 2 PowerOne Aurora 3000 inverters, one failed imediately and was replaced with an SMA.
After watching the production for several weeks I realized that Aurora daily production was significantly less per day than the SMA. I showed the results to the installer and asked him to replace the Aurora with another SMA, which they did. Power output on both strings has been equal since then.

[DanS26]

My 5.2 KW system is now 7 years old. When first installed I had 2 PowerOne Aurora 3000 inverters, one failed imediately and was replaced with an SMA.
After watching the production for several weeks I realized that Aurora daily production was significantly less per day than the SMA. I showed the results to the installer and asked him to replace the Aurora with another SMA, which they did. Power output on both strings has been equal since then.

Perfect example that if you are going to design with string inverters, then it is best to design with at least two inverters.

Yes, slightly more cost but you get that feedback and power comparison that is lacking in a single inverter. Easy to know and then track down a bad module or inverter, for that matter, as your experience attests.

[paul65k]

In a single string, partial shading on one or just a few panels would probably cause those panels to be bypassed, and not hurt the rest of the string. With micros/optimizers, those shaded panels might still be able to produce a little bit of power, but in most cases, it isn't going to affect the overall energy production very much.If you have strings in parallel or short strings, the micros make a better case. In the parallel situation, you risk dragging the voltage of the unshaded parallel string down to that of the partially shaded string, or in the case of a short string, dropping the voltage below the minimum necessary for the inverter to function.

Follow-up question on this........so in the example I posted above in this scenario you describe would the 3 shaded panels (1 in one string and 2 in the other) still contribute the ~3Kw for the day or would these be eliminated from the production.......with the micro inverters I can see the individual production but all that really matters is the total to the grid of course......just trying to understand what the differences would be....thanks!!

[sensij]

Follow-up question on this........so in the example I posted above in this scenario you describe would the 3 shaded panels (1 in one string and 2 in the other) still contribute the ~3Kw for the day or would these be eliminated from the production.......with the micro inverters I can see the individual production but all that really matters is the total to the grid of course......just trying to understand what the differences would be....thanks!!

Most of the 3 kWh contribution from those shaded panels probably occurred while they were unshaded, so that portion of the production would be the same on a string inverter. Looking at the 5 min data in Enlighten will give you some idea of how much energy was produced during the shaded portion of the day. Depending on exactly how the shade was striking the panel, you would lose some or all of that energy contribution. I'm guessing the daily total would still be close to 3 kWh, but you would be able to detect a small drop.

If you had used a string inverter with dual MPPT, and had each string on its own input, the production of the rest of the panels should be within measurement error of what they produced with microinverters, essentially unaffected.

I don't really disagree with Solarix about string inverters vs micros, my own desire for SolarEdge is mostly because I was/am *very* interested in the panel data and detailed reports, and I get enough shade during parts of the year that there is at least a chance that the production difference would make up for the cost difference. On paper, the optimizers should last longer than microinverters, but the distribution of failure times will probably have some overlap. Again, for me, I am comfortable going up and replacing an optimizer if needed, and have good access to my array, so the risk is less for me than for someone who would need to call a solar installer out to troubleshoot / replace. The bigger risk to me is that since SolarEdge is a proprietary system, no one else can make replacement optimizers... I'm counting on SolarEdge being around as long as I own the array. If things change and they disappear, I intend to buy a couple optimizers to put on the shelf before they are no longer available. That shouldn't really be a problem with micros, as long as the powerline communication systems don't conflict between manufacturers.

[albert436]

Well, you can look at the published financial statements for the public (ie traded) solar equipment companies. Lots of good information there if you take the time to read and understand what the accountants have to say.

For example, for years the rate of increase in warranty accruals far exceeded the rate of sales growth for Enphase. Big red flag.

Wow now that is an interesting bit of information.

There are very few roofs that can't be efficiently handled with a dual input inverter. The main application for distributed inverters is when you have serious shading issues, but even there you are better off solving it by eliminating the shade somehow than by band-aiding the design with module level optimization.

OK got it. Not sure how the neighbors will react when I cut down their trees, but I hope they will be understanding. OK all kidding aside, the dual input inverter would work for two sections of roof facing different directions, right ?

And -- another very elementary question -- well it's off topic but here goes. What approximate percentage of daily energy on a clear day, all else being equal, is produced in the middle 6 hours of the day, or in the 3 hours to each side of the panels' orientation I guess I should say. What I'm really trying to figure out about shade is if it is only early morning or late afternoon, say after peak consumption time (after 6pm) how much will it affect overall power generated ? What got me thinking about it was your comment about the shading issues. I wonder if we make too much of being perfectly shade free but if it is after the peak production hours it may not matter that much. I'll have to look at some of these guys' production curves again to answer my own question. Not talking random shade here but to one side or the other, morning/late afternoon.

[J.P.M.]

Wow now that is an interesting bit of information.



OK got it. Not sure how the neighbors will react when I cut down their trees, but I hope they will be understanding. OK all kidding aside, the dual input inverter would work for two sections of roof facing different directions, right ?

And -- another very elementary question -- well it's off topic but here goes. What approximate percentage of daily energy on a clear day, all else being equal, is produced in the middle 6 hours of the day, or in the 3 hours to each side of the panels' orientation I guess I should say. What I'm really trying to figure out about shade is if it is only early morning or late afternoon, say after peak consumption time (after 6pm) how much will it affect overall power generated ? What got me thinking about it was your comment about the shading issues. I wonder if we make too much of being perfectly shade free but if it is after the peak production hours it may not matter that much. I'll have to look at some of these guys' production curves again to answer my own question. Not talking random shade here but to one side or the other, morning/late afternoon.

On the last issue: Run PVWatts, get the hourly output and look for a clear day near your time of interest. Compare the output for various hours and stick a fork in it.

[solarix]

Ya, you can check with PV Watts for your area. My software shows the middle 6 hours for a south facing, 20deg tilted array to be 85% of your production

[albert436]

Thanks guys, I had looked at this stuff before and that's about what I had figured, but it has been a while and so much on my plate recently, only so much room in my small brain.

Will take a look for myself in any case, can't trust these internet forum guys.

[albert436]

Yes it will be interesting to see what it looks like for other orientations, or if the orientation is the predominate factor.

Need to run myself off for another working day at this point, hope everyone has a loverly day !