Have string inverter systems just gotten a lot more expensive?

[DaveDE2]

Hello, I'm new to the forum, first post.

I am about to put up a 13kW system on my due south facing garage roof. I bought 51ea 255W 60cell Phono Solar panels at $0.59/W. The array physical configuration is 3 rows of 17 modules. I have been baselining Solaredge P300 optimizers and an SE-11400 11.4kW inverter, 3-strings, 17 modules per string. I am now having second thoughts about using Solaredge for this system. Here's why:

1. The location is Colorado, roof with 18 degree pitch, 175 degree azimuth, no shade until maybe 4 o'clock in the afternoon (big trees on west side). Shade will then gradually sweep across the entire array as the afternoon/evening progresses. SAM simulations indicate that shading losses are ~2.8% if I use Solaredge but I'm thinking of using a simple string inverter instead.

2. I'm concerned about reliability and maintenance of the system. If I go Solaredge then I'm locked into using their components for the life of the system. If the inverter goes out it will have to be replaced with a Solaredge inverter (unless I want to eliminate 51 power optimizers that cost me $3k). I'm also somewhat worried that Solaredge may not be around in 10-20 years and I'll be out of luck or their inverters 10-20 years from now will not be compatible with old optimizers. On the other hand, if my system uses sting inverter(s), I can just replace it with almost anybody's product. This is not to mention the hassles associated with possible power optimizer failures on the roof.

3. I'd think that using two 6kW string inveters, each having dual MPPTs running 12-13 modules each would give me similar energy yield and be less risky, easier to maintain. Cost of the SE optimizers and inverter is $5050. Cost of two SMA 6kW inverter is about $4900, a wash. So I'm thinking simple string inverters is the way to go.

but...Now I find out that my local juristiction is requiring the NEC 2014 rapid shutdown. Going the string inverter route, this would require me to add an esitmated $1700 for remote contactors located within 10' of the array and electronics to control them (two each Midnite Solar contactors and Birdhouse remote control). That's a lot of added installation hassle and has it's own reliability concerns. Going the Solaredge route, rapid shutdown requires no extra equipment.

Any suggestions/comments? What would you do?

thanks,

Dave

[sensij]

I have SolarEdge on my roof, but understand your concerns. Mine is a 12 panel system with easy access to all of the panels, I'm not as sure I'd have made the same choice for a 51 panel system.

A third option to consider would be using Fronius equipment. Like the SMA inverters, the Primo inverter has two MPPT inputs, with an MPPT voltage range of 240 V to 480 V. They include communication and datalogging out of the box... that is an add-on for SMA.

You could save a couple hundred $ if you ran something like strings of 11 and 12 into a 5 kW inverter and then a pair of 14 panel strings into a 6 kW inverter. The DC:AC ratio for each inverter would be less than 1.2, costing you very little in "clipped" production.

Using prices from AltEStore:
5000 W Primo inverter - $1700
6000 W Primo inverter - $1965
Fronius Rapid Shutdown box (1 per string) - $275 / Ea * 4 = $1100

Even with shipping, probably under $5k.

[solarix]

Dave, I think your analysis is right on. There is a real reliability risk with distributed inverter systems, but the rapid shutdown regs are a major cost adder for central inverters. While I really like the SolarEdge design, I have been really burned by their reliability. SMA is coming our with their low-cost solution to the rapid shutdown any day now but I haven't been able to find details about it yet. See if you can find that option, I think it would be worth waiting for.

[DaveDE2]

Thanks for your input sensij and solarix. At this point I have acquired all of the PV system components except for the inverter(s). I'm leaning toward using a pair of the new SMA SB6.0-1ST-US-40 6kw units. I can get them for about $4k. The SMA rapid shutdown box is purportedly dual string, so I'd need two of them. Prices are not available until later this month. If they come in around $500ea then SMA vs SolarEdge is a wash price and performance wise for this system.

This all leads me think about system monitoring. I'm wondering how the two systems would compare in that regard. Seems like SolarEdge would have the advantage there since output of each module can be analyzed (I think), whist with SMA, only each inverter's array performance can be analyzed. My understanding is that the SMA tools don't even allow for string monitoring unless Modbus is used along with some programming, something that is not up my alley and I'd like to avoid. I could be wrong about that though. My head is spinning with what all tools I would need to get a good feel for how well the system is working. I'd like to at least be able to compare the four string outputs of the array as time goes on. Does anybody know how Sunny Portal/WebConnect compares to the SolarEdge tools? Are they comparable, is one a lot better than the other?

[sensij]

That SMA inverter is very new, you'd be the first forum member (who has posted about it) with one installed. Having 3 MPPT's is a big step forward. It has different communication options than the SBxxxx-TL-22-US series, which required a Webconnect card for ethernet capability. The new one has built in ethernet and Wifi capability, and you can log directly into the inverter with a browser to see your data... it no longer looks like Sunny Explorer software is required. I can't tell from the documentation if individual string performance is available without resorting to Modbus, or if only overall output can be see through the web based application.

With SolarEdge, there is not actually local ethernet access to the data, it all goes to the cloud. The only local access would be using an RS485 connection with a Modbus protocol. When you log in to SolarEdge's cloud server, you can see normal production data, as well as individual optimizer data, create your own graphs, and download reports. There are aspects of it that are nice, but because the optimizers can't communicate simultaneously, it is difficult to detect differences between them live. It would be easy to see a total panel failure, but that would be true if you can see the string data for a non-SolarEdge inverter.

SolarEdge has published an API that gives access to inverter level data (not individual optimizer level), making it very easy to share your data via sites like PVOutput.org. SMA elected to keep their Sunny Portal cloud data inaccessible (except to commercial customers for a fee), and it is hard to see how the change in inverter connectivity of this newest generation would result in a change to this policy. PVoutput.org membership is possible for SMA owners, but you'll probably need a Raspberry Pi or other always-on computer to query the inverter and then transmit the data to the 3rd party site.

[DaveDE2]

The new SMA inverters do look good. More features and lower cost than the SBxxxx-TL-22-US series. I'm not so sure about the utility of 3 MPPTs in these inverter sizes however because I'm not sure how low the PV voltage can be and still have the trackers operate. The data sheet is confusing. One line says "Rated MPP Voltage Range" is 220v-480v, and the next line says "MPPT operating voltage range" is 100v-550v. If the MPPTs can operate down to 100v, then great, 3 strings can be used adequately. But if they only operate down to 220v, then the MPPTs wouldn't even kick in for a 7 module string. I called SMA and inquired about it and the tech support person couldn't even explain what the spec means and how low the MPPTs operate. Regardless, I'd probably just use 2 strings from each inverter since efficiency is ~1% higher with the higher input voltage. I have an inquiry into them about web based string monitoring but it seems that sort of question may be way over thier heads and I've heard that support from them on this sort of stuff is SLOW

That's too bad about SMA and PVOutput. Your PVOutput link looks sweet. Maybe I can do a Raspberry Pi implementation later. Thanks for elaborating.

[sensij]

Another forum member did some heavy lifting trying to track down what the difference between the rated and operating range is, in this thread. In a nutshell, when they advertise the inverter as having 97% CEC efficiency, it is referring to the MPPT rated range. If you design outside of that range, but still in the operating range, the MPPT will do what it should, but at somewhat lower (but not catastrophically low) efficiency.

The utility of three MPPT's is for an array split over more than two roof faces, or when the system size is more than two strings can handle. By adding a 3rd MPPT, you no longer need to make a pair of parallel strings at equal length, which helps make shade handling more efficient. For example, an array built from 23 panels, 315 W each (72 cell panels) would, in some locations, go over the Voc limit if split into just an 11 and 12 panel string. It isn't nearly as significant as the change from a single MPPT to two, but it is still a nice step forward.

[solarix]

Ya, dual input inverters were a big advancement and gave designers much more flexibility, but there were still times when it was not possible to make a single inverter solution. I can't wait for these new SunnyBoys to be available...