micro-inverters or optimizers when shading is major issue ?

[scrambler]

Bypass diodes are solid state and if the J-Box is located in an area that it gets sufficient cooling (an even bigger problem for optimizers) then there's no 'stress' when it's activated. There's no reason a bypass diode couldn't be 'activated' for the life of the panel.

I am surprised by that and confused, as many articles I read do speak about the not so great lifespan of bypass diodes and the fact that they can be stressed by repeated shade conditions.
In addition to the optimizer review I mentioned earlier, below is another article that also mentions it.

https://www.positronicsolar.com.au/2019/08/02/panels-shade-and-diodes/

It is hard to know where the reality lies with all this.
The fact that apparently some panel manufacturer like LG appears to have mention in the warranty text such as the one quoted in the optimizer review (below)
"LG, state that their panel warranty does not cover: Improper installation or reinstallation and poor solar system design. (Examples of improper installations and very poor system design are modules installed in conditions which put long term stress on the bypass diodes in the modules, and also reduce the system output for the owner – for example prolonged significant strong shadowing of the modules e.g. via trees, walls, gables, overhangs, valleys, chimneys, satellite dishes etc (In such situations a professional solar designer will suggest a micro-inverter or optimiser solution and with such a proper solar system design solution the module warranty is fully applicable)."

[nwdiver]

The review below seem to contradict that statement, do you disagree with its statement that :
"Before a bypass diode has the time to engage, the Tigo optimisers see the impedance caused by the shade and open a bypass tunnel to match the impeded current. This allows the inverter to work at a higher current without affecting the solar panel bypass diode."

https://mcelectrical.com.au/blog/tig...el-optimisers/

It's possible that module level MPP is able to respond faster than a MPP in the inverter but how much additional generation is that going to get you? From what I've seen modern inverters respond in seconds to a change in conditions.

Any credibility that article had was lost for me when with the graphic that perpetuated the module-level MPP myth. Shading one panel will not reduce the output of unshaded panels. There may be some additional generation you can squeeze out of partially shaded panels since string level MPP will 'delete' partially shaded sections... but would that really be worth the additional cost?

I have a SE system on one of my installs. There is a section that receives shade and would be 'deleted' without optimizers when shaded. Each shaded panel probably produces <50wh/day during the period that it would be removed from the string w/o an optimizer. That's ~18kWh/yr. It it worth spending $30 to harvest an extra ~18kWh/yr? IMO no. And that's not counting the other ~30 panels that don't benefit much at all which would also need optimizers....

[scrambler]

It's possible that module level MPP is able to respond faster than a MPP in the inverter .

Tigos (object of the review) do not include an MPPT, they do not work like SolarEdge Optimizer (That do include an MPPT)

[nwdiver]

Tigos (object of the review) do not include an MPPT, they do not work like SolarEdge Optimizer (That do include an MPPT)

Sure.... I was using MPP as short-hand for module level optimization.... however Tigo does that.

[jflorey2]

"Before a bypass diode has the time to engage, the Tigo optimisers see the impedance caused by the shade and open a bypass tunnel to match the impeded current. This allows the inverter to work at a higher current without affecting the solar panel bypass diode."

Marketing BS. Diodes work within nanoseconds. Optimizers take hundreds of milliseconds to seconds.

[nwdiver]

I found an article from NREL that both TIGO and Solar Edge cite as evidence for the need for module level optimizations. BUT... in the study NREL uses 3 parallel strings. Which as I've state IS one area where shading WILL influence the entire array. Once you get string level MPPT the only benefit to module level optimization is the ability to recover generation from the partially shaded area itself. Unshaded sections ARE NOT effected like they would be with parallel strings.

So that's the crux of the debate. Is it worth the expense of module level optimization to recover lost generation from partially shaded areas?

[ButchDeal]

the study NREL uses 3 parallel strings. Which as I've state IS one area where shading WILL influence the entire array.

...

So that's the crux of the debate. Is it worth the expense of module level optimization to recover lost generation from partially shaded areas?

Module level optimization is not more expensive.

Further That is NOT the crux of the debate at all.

OP asked about micros or optimizers for shade. with NEC 2017 one or the other are in the mix.

Further by your own statements optimizers and micros will produce more than strings in shaded area:

There may be some additional generation you can squeeze out of partially shaded panels since string level MPP will 'delete' partially shaded sections..

as far as OPs actual question goes. Micros and SolarEdge will likely produce about the same assuming a well designed system without clipping issues.

There may be some advantages with SolarEdge beyond the already mentioned features in that with a heavily shadowed roof, because you have pooled inverting with optimizers, you can install a smaller inverter without clipping which might be a simpler install (load side tap instead of line side tap).

[nwdiver]

Module level optimization is not more expensive.

??? How did you arrive at that conclusion? If I want a 5kW Solar Edge system it's gonna cost (~$1480 inverter) + 16($60 Optimizers) = $2440 Compared to a SMA String level system for (~$1120 inverter) + 16($30 TS4-F) = $1600. That's a $840 premium. $1320 if it's a ground mount.

I totally agree that the $2360 SE system will probably produce more kWh than the $1680 SMA system. But you could also add another kW of solar with the $840 you're saving on not getting module level optimization and 6kW of string level will produce more kWh/yr than 5kW module level...

[JSchnee21]

The reliability of electronics on the roof, and KISS aside, I've always felt (SWAG) that PV arrays with panel level optimizers/micros would put less stress on the panels themselves, busbars, and bypass diodes.

For example, in the classic string design, all of the current and voltage of the entire string (e/g 600V or higher, 10 amps or higher) is passing through each and every panel, busbar, bypass diode. If you have a chronically shaded region (or as region with cracks/fissures/corrosion), that either does or does not rise to the level of tripping the bypass diode, I would expect to see some localized heating as the power is forced through the remaining elements (I'm speaking very generally here -- hand wave).

In contrast, for the panel level electronics case, each individual panel only ever it's own small voltage and current and doesn't have to deal with conducting the power of the entire string. Likewise the potential for PID should also be reduced as the voltage is much, much lower (within the panel).

Of course some of these stresses are just handed off to the micro/optimizer. Aside from the capacitor issue, I would think micros would see the least such stress as there is no "string" per se, whereas optimizers still see the full voltage and current of the string.

Of course, I've never seen any data on this, not even from the Fraunhofer.

[RShackleford]

I'm getting really confused. If I get the "optimization" version of the Tigo TS4 (TS4-O ?), does that do module-level optimization ? IOW, is the debate between SolarEdge versus Tigo optimizers, or is it between the non-optimizing Tigo (TS4-F ?) and either the optimizing TS4 or SolarEdge ?

[ButchDeal]

??? How did you arrive at that conclusion? If I want a 5kW Solar Edge system it's gonna cost (~$1480 inverter) + 16($60 Optimizers) = $2440 Compared to a SMA String level system for (~$1120 inverter) + 16($30 TS4-F) = $1600. That's a $840 premium. $1320 if it's a ground mount.

I totally agree that the $2360 SE system will probably produce more kWh than the $1680 SMA system. But you could also add another kW of solar with the $840 you're saving on module level optimization and 6kW of sting level will produce more kWh/yr than 5kW module level...

Simple, you are Not showing retail pricing ( high for solaredge and low for SMA or based in smaller inverter) the Sunny boy 5.0-US actually runs over $1400 and the TS4-Fs are not carried by most distributors ( they were just recalled earlier this year) but go for more ~$40 retail. So more like $2,040 for SMA, And still only 97% efficient of an inverter,

and as for installing more pv, on mosts roofs you are going to have much more flexibility to do that with optimizers, allowing use of shadowed areas, multiple azimuths, and areas with room for too few for a string without optimizers.

[ButchDeal]

I'm getting really confused. If I get the "optimization" version of the Tigo TS4 (TS4-O ?), does that do module-level optimization ? IOW, is the debate between SolarEdge versus Tigo optimizers, or is it between the non-optimizing Tigo (TS4-F ?) and either the optimizing TS4 or SolarEdge ?

Newdiver is trying to make it about non-optimizer versions against solaredge.
the tigo optimized version does do module level optimization but differently than solaredge.

the main thread is about micros vs optimizers though.

[nwdiver]

Simple, you are Not showing retail pricing ( high for solaredge and low for SMA or based in smaller inverter) the Sunny boy 5.0-US actually runs over $1400 and the TS4-Fs are not carried by most distributors ( they were just recalled earlier this year) but go for more ~$40 retail. So more like $2,040 for SMA, And still only 97% efficient of an inverter,

and as for installing more pv, on mosts roofs you are going to have much more flexibility to do that with optimizers, allowing use of shadowed areas, multiple azimuths, and areas with room for too few for a string without optimizers.

I was comparing apples-apples. I can get a 5kW SMA system that's NEC 2017 compliant for ~$1600. Where can I get a 5kW Solaredge system for less than ~$2400? If you get special volume pricing from SolarEdge that's not really available to @RShackelford and a bit of a moot point... for him it's ~$1600 for SMA vs ~$2400 for SolarEdge. I don't think he's gonna get $800 more energy from the SE system... not in ~20 years...

Newdiver is trying to make it about non-optimizer versions against solaredge.

Correct. My point is that with string level MPPT the cost of optimizers is not worth the benefit. You do not need module level optimization if all the panels on each string can face the same direction. The only real benefit to module level optimization is recovering generation from shaded areas... if the shade persists long enough to make that worth while then it's probably not a good location for solar to begin with.

I'm sick of the misleading graphics from SolarEdge and Enphase showing a 3kW String of 10 panels with one panel partially shaded and claiming the string is now reduced to 1.5kW with a string inverter. That is NOT what happens. And if the shade is only over one cell group the result would be roughly identical. ~2.9kW. Only if the entire panel is partially shaded is there a difference and even then it would be ~2.7kW vs ~2.75kW. NOT 1.5kW vs 2.75...

[scrambler]

I'm getting really confused. If I get the "optimization" version of the Tigo TS4 (TS4-O ?),

Yes that is the one that does Panel/module optimization.

I think the debate has been all over the place both between Tigo and SE (for which I think is efficiency difference is minor but reliability is not) and between Optimizer or not, which heavily depends on String configuration and the actual shade situation.

I don't think there is anyway to make a blanket statement as each configuration is going to be different, but it seems that if there is any significant shade, it is probably a good idea to consider some optimization.
Then for me the debate is really about the cost and reliability of the chosen method.

[RShackleford]

My point is that with string level MPPT the cost of optimizers is not worth the benefit.

I think the debate has been all over the place both between Tigo and SE (for which I think is efficiency difference is minor

This is why I'm leaning heavily towards string inverters, and Tigo optimizers IF I have optimizers. Tigo gives one the opportunity to decide later to include optimizers, and to optimize only "problem" panels. With SE, you gotta buy into the whole system from the get-go.

Also, with the non-SE plan, I have the option of the SMA inverter with "secure power", meeting my desire for a little backup power during grid outages. Though the jury still seems to be out, here, over whether using optimizers will defeat the "secure power" feature.

P.S. I can include my EV in the backup power scheme. I'm already using an inverter connected to its 12v system for backup power, so maybe I just use the SMA to charge the Leaf's battery using the L1 cable, and don't backup from the SMA directly.