Enphase vs. Solar Edge. What would you go with + interesting input from engineer

[J.P.M.]

the only point i was making was that max string length with traditional sting inverters is dependent on location due to the lowest expected temperature at that location... That's it. Digging deeper would just cause confusion or result in missing the point.

r.i.p.

[nwdiver]

What/which cell temp. are you referring to, the maximum design temp., or the minimum design temp. ?

The ONLY point I was making was that max string length with traditional sting inverters is dependent on location due to the lowest expected temperature at that location... that's it. Digging deeper would just cause confusion or result in missing the point.

Not quite; One of the unique features of Solar Edge inverters is that their string sizes are based on power not voltage. IIRC your SE inverter is limited to 6kW per string. 6kW/340w = 17 panels.

Most other inverters like SMA are limited to 600v open circuit. The OC voltage of the panels you're considering is ~71v but this is at 20C. Voltage varies based on ambient temperature so you need to take into consideration the voltage on the coldest day you expect.

[J.P.M.]

....... sure....... without being uselessly pedantic maximum string length will vary by geographic location due to how that location influences cell temperature because some places get much colder than others.... consult a design tool.

Pedantry has nothing to do with it - for me anyway.

What/which cell temp. are you referring to, the maximum design temp., or the minimum design temp. ?

The way I saw and read the OP's original post, the temps. the OP was writing about were max. temps. as they related to micro inverters and probably optimizers and also cell temps. As I wrote, those max. temps. will be influenced by many of the same things that influence cell operating temps., as will any minimum cell operating temps. of micros/optimizers and cells with the addition of at least one other factor I mention below.

You made a statement that "Voltage varies based on ambient temp....". I see that as an incorrect statement due to it's incompleteness. It's a partially correct statement, but more is needed. Without more explanation, your statement might lead neophytes to thing cell temp. == amb. air temp. As I've already stated, several other things come into play, with clear sky P.O.A. under normal production times usually having at least as much and often more influence on clear sky cell temp. than does ambient air temp. as can easily be confirmed by looking at the NOCT on any decent panel's spec sheet where amb. air temp == 20 C and NOCT is usually ~ 45 C or so.

On the low cell temp. side of things, if that's what you're concerned about for string sizing or other reasons, design low ambient air temp. is often used because irradiance is usually low when such low amb. temps. are likely - often around sunup - making amb. air temp. at such times the biggest factor influencing cell temps. I get that as well how higher voltages under those conditions can influence design. At such times your statement is a bit more correct. But, and FWIW, radiant sky temps. which can cause cell temps to fall below amb. air temps. by 5 - 10 C in places that experience cold but dry winters with very low coincident dew points can easily get cell temp. below amb. air temp. by about the same amount. Sharp designers know that and may adjust their design and low design temps. accordingly as they see best.

If you believe adding information for the sake of correctness, accuracy and completeness is pedantry or unnecessary, so be it. Ignore what I write.

I say, what's the worst thing that can happen by adding the information ? Someone may learn something ? They can always ignore it too.

As usual, take what you want of the above. Scrap the rest.

[bird95134]

Thanks everyone again for assistance with questions and comments! I have decided and moved forward tonight after finishing up a few reference calls. I selected the following configuration:


Production Design, 14,960 kW. Estimated production: 27,963
qty 44 - Panasonic VMHN340SA17 PV Panels, qty 44 - SE P400 Power Optimizers, qty 2 - SE7600H-US Inverters, added Solar Ready 100 AM power panel upgrade
Price installed per kW before tax credit, $2.50 (not including power panel upgrade)

Other:
Installing new roof (lift and replace -- new 50 yr synthetic underlayment, adding o'hagen vents)
All PV panels on south facing roof mount.

Will post on this thread as I progress with project and come on line.

Thanks again.

[nwdiver]

For STC it's 25 C. Serious and thorough design considerations use cell temp. which is a function of more than ambient temp.

....... sure....... without being uselessly pedantic maximum string length will vary by geographic location due to how that location influences cell temperature because some places get much colder than others.... consult a design tool.

[J.P.M.]

Sure... but for design considerations it's ambient temperature. You can have slightly longer strings in Panama City than you can in Fargo ND, with a traditional string inverter anyway.

For STC it's 25 C. Serious and thorough design considerations use cell temp. which is a function of more than ambient temp.

[nwdiver]

Actually, voltage of a PV device varies pretty much inversely as the device temp.

The device temp. is f(energy balance on the device) and is mainly influenced by POA irradiance, ambient temp., wind vector and to a lesser degree by some other stuff.

Sure... but for design considerations it's ambient temperature. You can have slightly longer strings in Panama City than you can in Fargo ND, with a traditional string inverter anyway.

[J.P.M.]

Not quite; One of the unique features of Solar Edge inverters is that their string sizes are based on power not voltage. IIRC your SE inverter is limited to 6kW per string. 6kW/340w = 17 panels.

Most other inverters like SMA are limited to 600v open circuit. The OC voltage of the panels you're considering is ~71v but this is at 20C. Voltage varies based on ambient temperature so you need to take into consideration the voltage on the coldest day you expect. This could be ~75v; you'd need to do the calc for your area or error conservatively. 8 panel strings might be safe. 7 would definitely be ok. So to get the desired 7.4kW with a SMA inverter you would need to use all 3 MPPT channels. If you used a lower voltage panel 2 channels would be sufficient leaving a 3rd for adding more panels in the future.

Actually, voltage of a PV device varies pretty much inversely as the device temp.

The device temp. is f(energy balance on the device) and is mainly influenced by POA irradiance, ambient temp., wind vector and to a lesser degree by some other stuff.

[nwdiver]

Do I have this correct?

Not quite; One of the unique features of Solar Edge inverters is that their string sizes are based on power not voltage. IIRC your SE inverter is limited to 6kW per string. 6kW/340w = 17 panels.

Most other inverters like SMA are limited to 600v open circuit. The OC voltage of the panels you're considering is ~71v but this is at 20C. Voltage varies based on ambient temperature so you need to take into consideration the voltage on the coldest day you expect. This could be ~75v; you'd need to do the calc for your area or error conservatively. 8 panel strings might be safe. 7 would definitely be ok. So to get the desired 7.4kW with a SMA inverter you would need to use all 3 MPPT channels. If you used a lower voltage panel 2 channels would be sufficient leaving a 3rd for adding more panels in the future.

[J.P.M.]

I once split my array exactly in half and installed a Solaredge inverter in addition to the original Xantrex. Impressed by the SolarEdge claims of "up to 2%" gains in various ways, I expected the new inverter to easily win the contest despite no shading on my PV panels. The result? The old Xantrex produced slightly more kWh every day....

Thank you.

I always like getting real experiential information and data from folks who own systems and also install them for a living.

I'd respectfully note and keeping in mind that people see what they want to see and hear what they want to hear, saying "up to 2% gains..." makes it easier for potential users to infer a lock on improvement, when that's not close to what was said and really doesn't say anything. It might be just as accurately stated "up to 2% fewer gains...".

Q: Anything about the half arrays you can think of that might be different, one to the other ? Reporting accuracy, wiring, slightly differential and very local wind patterns around the array, other stuff ?

[bird95134]

Thanks for all of the feedback and information!!

This particular post is interesting....

2. Yes; Each inverter is completely independent. Each MPPT channel is also independent which is why I think 1 String per Channel is ideal. With the panels you've selected you're somewhat limited since you can't have strings longer than ~8? This would accommodate ~16kW but you would use all 6 channels and have limited ability to expand in the future.

.

Not sure I follow the comment on # of panels per string and my limitations based on PV panel I have selected, unless you are referring to # of panels by string with or without the power optimizer configuration. My proposed configuration would include power optimizers.

Based on my calculations (without installer input, just based on my research, sizing guides and product specs), I see the capacity per string as follows, for the proposed SolarEdge solution:

A = Inverter Nominal DC Input Voltage = 480V
B = Power Optimizer Maximum Operating Output Voltage = 60V
C = Power Optimizer Maximum Output Current = 15A
D = PV Panel output = 340 W

Minimum String Length = A / B + 2 = 10
Maximum String Length = A * C / D = 21

So based on my calculations, string density (# of PV panels) is between 10 to 21 based on the SE inverter model, power optimizer, and PV panel.

Do I have this correct?

[solarix]

I once split my array exactly in half and installed a Solaredge inverter in addition to the original Xantrex. Impressed by the SolarEdge claims of "up to 2%" gains in various ways, I expected the new inverter to easily win the contest despite no shading on my PV panels. The result? The old Xantrex produced slightly more kWh every day....

[nwdiver]

And white papers from Solaredge and Enphase show how theirs is better. So that's not very convincing.
Solaredge claims 98% efficiency at their optimizer and 99% at the inverter - which means the two together would be 97% efficient.

The comparisons I've seen from Solaredge and Enphase are module level vs multiple parallel strings. In those cases there can be some gains. String level MPPT is a different animal. If you shade a panel in a string that is parallel to another string the lost production is greater than the area that's shaded. It CAN effect unshaded sections. This is generally not true for string level MPPT. With string level MPPT production of unshaded areas is not reduced by shaded sections.

I've seen graphics from Solaredge and Enphase that are categorically untrue and have done a lot of damage with misinformation.

Like this; This is 100% not true.

Screen Shot 2020-05-21 at 7.48.37 PM.png

[foo1bar]

The cost and reliability components of the new SMA inverters is the biggest selling point IMO. Instead of ~$60/panel for optimizers it's <$30/panel for a rapid shut down device.

personally - I have 32 panels and a 7.6kW inverter.
Looking at the current pricing I find at renvu:
Solaredge: $1781 for inverter + 73.86 * 32 = $4145
SMA: $1646 for inverter + $40.25 * 32 == $2934

So $1200 more for Solaredge for a 32-panel system. (I thought it'd come out closer than that - but it didn't... Maybe renvu doesn't have as good of prices at the moment.)

If the white paper published by SMA is correct your annual production is also higher without the losses from module level electronics.

And white papers from Solaredge and Enphase show how theirs is better. So that's not very convincing.
Solaredge claims 98% efficiency at their optimizer and 99% at the inverter - which means the two together would be 97% efficient.

SMA - I think I've seen graph for them that showed 96-98% efficiency, depending on voltage and power level.

So I think they're close enough in efficiency that the other items are going to be much more of a deciding factor.


For me, solaredge is what I installed because
* similar efficiency to string inverters when not shaded
* better performance when some are shaded
* better performance with multiple orientations
* less electronics on roof than Enphase (And the electronics SE has on the roof are in my perception less susceptible to heat degradation / heat-induced failure)
* cheaper than enphase (for my configuration)
* benefits of per-panel tracking.

[nwdiver]

Questions:

1. If my current proposed Solar Edge design (qty 44) Panasonic VMHN340SA17 panels + ( qty 2) SE7600H-US Inverters + (qty 44) SE P400 Power Optimizers. This equals a 14,960 kW capacity. I would need to upgrade one of inverters (or both) in this design if I wanted to add panels for additional load, correct?

2. If one Inverter of the 2 died/broke, would I still have 1/2 capacity?

1. (2) SMA 7.6kW inverters can support up to ~24kWdc of solar*; Check out the data sheet for the inverters. Each has 3 independent MPPT channels that can support 10A of power and 18A short circuit. You would probably need different panels with lower voltage and higher amperage to realize ~24kW or run parallel strings on a single MPP channel which I would advise against. One of my projects is 22.77kW on (2) 7.6kW SMA inverters. It's been doing great for over a year.

2. Yes; Each inverter is completely independent. Each MPPT channel is also independent which is why I think 1 String per Channel is ideal. With the panels you've selected you're somewhat limited since you can't have strings longer than ~8? This would accommodate ~16kW but you would use all 6 channels and have limited ability to expand in the future.

The cost and reliability components of the new SMA inverters is the biggest selling point IMO. Instead of ~$60/panel for optimizers it's <$30/panel for a rapid shut down device. If the white paper published by SMA is correct your annual production is also higher without the losses from module level electronics. Cheaper, more reliable AND more efficient.