String sizing for Grid Tie inverter

[jarred125]

Good day everyone, my first post here after doing as much lurking and reading as I can.

Recently I decided to finally make my jump into solar and I have 20 panels on the way from Santan Solar. These are Canadian Solar C3SW-445MB-AG bifacial panels. The basic specs are:

VoC: 48.9V
Vmp: 40.9V
Imp: 10.89A up to 14.16A @ 30% gain from the underside

I have been looking at the Fronius Primo 8.2-1 inverter (I would love to have battery backup now but with current usage it is a pipe dream .. I am working to fix this). Basic inverter specs:

Operating range: 80V-600V (max input is 600V)
MPP voltage: 270-480V
Max usable current: 18A / 18A (2 MPPT)
Max usable total current: 36A

My question comes with sizing the strings for this inverter (or any other inverter for that matter). Clearly I cannot just tie all 20 panels together in series and connect as that would greatly violate the voltage input of this inverter. Running them in parallel would mean a low voltage but high current and would then require larger conductors, etc (if I am thinking correctly). So creating multiple strings to keep the voltage lower is key, but what about keeping amps lower? The 18A max usable current confuses me a little on this and other devices.

I am thinking I would have two strings of 10 panels each for a voltage of 489V each with a max potential Imp of 14.16A if I achieved 30% gain from the under side of the panels. Those run into a combiner box would be an output of 489V with potentially almost 30A.

How would you go about configuring this? I greatly appreciate any ideas you are willing to share.

 

[DThames]

10 in series on each of the MPPT inputs would be my knee jerk response. If it gets really cold where you live, look at the voltage of the panels with the temp chart. 10s would be 500v at rated temp, which would give plenty of headroom for cold, unless you are in crazy cold location.

 

[jarred125]

10 in series on each of the MPPT inputs would be my knee jerk response. If it gets really cold where you live, look at the voltage of the panels with the temp chart. 10s would be 500v at rated temp, which would give plenty of headroom for cold, unless you are in crazy cold location.

I am outside of Cincinnati, OH. We have seen temps in the negatives only a few times (-10F at most was the worst I have seen) in my life. Our average temps the last 2 years in winter were around 31F with minimums 10-15F.

Edit: One last question is .. with the amps being higher than the MPPT usable, what happens when I cross that threshold?

 

[ncsolarelectric]

Can I recommend microinverters instead of Fronius?

 

[jarred125]

Can I recommend microinverters instead of Fronius?

I have thrown this around as well, in fact I was reading through your thread yesterday about Enphase and Hoymiles. I am a total noob so I am open to anything!

 

[DThames]

I have thrown this around as well, in fact I was reading through your thread yesterday about Enphase and Hoymiles. I am a total noob so I am open to anything!

I have micro inverters and they are great. However, if you want a battery system, they might not be the best approach. They "can" work with a battery system but often with some pretty serious limitations.

 

[jarred125]

I have micro inverters and they are great. However, if you want a battery system, they might not be the best approach. They "can" work with a battery system but often with some pretty serious limitations.

That is the biggest catch for me and why I am really thinking string inverter. I will eventually add batteries down the line, I am just not sure when. So many things to consider at this point.

 

[DThames]

That is the biggest catch for me and why I am really thinking string inverter. I will eventually add batteries down the line, I am just not sure when. So many things to consider at this point.

I know. I am trying to help a friend figure out what he wants to do and he has zero experience. I just want to tell him, "Put some panels on a ground mount and we can go from there". That part we can largely decide on. There are just too many options.

That is why I went with micro inverters. I just wanted to get my electric bill to zero.

 

[jarred125]

I know. I am trying to help a friend figure out what he wants to do and he has zero experience. I just want to tell him, "Put some panels on a ground mount and we can go from there". That part we can largely decide on. There are just too many options.

This is basically what I decided to do. The panels were on a decent sale. Originally I wanted 10, but then figured .. why not 20? The shipping was only a little bit more and to gives me a lot to play with. I am having fun with this process, I love to learn and this is a whole new world for me. I hope your friend follows your advice, get some panels installed and go from there!

 

[Hedges]

Voc limits are to be calculated using historical record cold temperature for your location, not "only a few times" or "in my lifetime". Get it wrong once and you kill the SCC or inverter, maybe the battery.

My first conservative estimate is Voc increases 16% in freezing weather. Sharpen your pencil by using your record temperatures and Temperature Coefficient of Voc from panel data sheet.

VoC: 48.9V x 1.16 = 56.7 Voc in extreme cold
600V limit / 57.6 Voc = 10.6 panels; do not exceed 10 panels in series unless sharper pencil says it is OK.

You can put 10s into each MPPT. 10A or 14A is well under 18A.
You can put 10s2p into one MPPT. Available current exceeds 18A, simply won't be used. Is there a "max short circuit current" spec on the inverter?
If you orient 10s toward AM sun, 10s toward PM, and connect in parallel, peak current will be reduced to about what presented area is equal to. This could let you use four, 10s strings on the inverter.

String inverter for now - is this grid tie? Can you backfeed for net metering? use current transformers for zero export?
High voltage string inverter is most efficient and most cost effective for producing AC.
If you are require to have RSD box per panel, that will make cost closer to microinverter. Either one is annoyingly expensive compared to used panels.
Used and new-old-stock string inverters, not implementing latest grid-support functions in many locations, can be had for about $0.10/W.

Either string or micro inverter can have AC coupled battery inverter added later, if the PV inverter supports frequency-watts. But it is a fairly expensive battery inverter.

Some batteries-optional inverters from MPP and others are much lower purchase price. If you can find something that is also grid-interactive, that could support all usage models.

 

[jarred125]

Voc limits are to be calculated using historical record cold temperature for your location, not "only a few times" or "in my lifetime". Get it wrong once and you kill the SCC or inverter, maybe the battery.

My first conservative estimate is Voc increases 16% in freezing weather. Sharpen your pencil by using your record temperatures and Temperature Coefficient of Voc from panel data sheet.

VoC: 48.9V x 1.16 = 56.7 Voc in extreme cold
600V limit / 57.6 Voc = 10.6 panels; do not exceed 10 panels in series unless sharper pencil says it is OK.

You can put 10s into each MPPT. 10A or 14A is well under 18A.
You can put 10s2p into one MPPT. Available current exceeds 18A, simply won't be used. Is there a "max short circuit current" spec on the inverter?
If you orient 10s toward AM sun, 10s toward PM, and connect in parallel, peak current will be reduced to about what presented area is equal to. This could let you use four, 10s strings on the inverter.

String inverter for now - is this grid tie? Can you backfeed for net metering? use current transformers for zero export?
High voltage string inverter is most efficient and most cost effective for producing AC.
If you are require to have RSD box per panel, that will make cost closer to microinverter. Either one is annoyingly expensive compared to used panels.
Used and new-old-stock string inverters, not implementing latest grid-support functions in many locations, can be had for about $0.10/W.

Either string or micro inverter can have AC coupled battery inverter added later, if the PV inverter supports frequency-watts. But it is a fairly expensive battery inverter.

Some batteries-optional inverters from MPP and others are much lower purchase price. If you can find something that is also grid-interactive, that could support all usage models.

Whew! Thank you for commenting here and helping me out!

I will review the record lows for my area and calculate based off of that.
The data sheet from Fronius for this inverter says : Max. array short circuit current (1.5 * Imax) 27A / 27A

The inverter will be grid tie, we do have net metering (Duke Energy in Cincinnati, OH). I will check on the requirement for the RSD per panel but my initial thoughts based on what I have seen installed locally is they are not required per panel.

I will reach out to my provider today and see what the requirements are. Thank you again!

 

[DThames]

Voc limits are to be calculated using historical record cold temperature for your location, not "only a few times" or "in my lifetime". Get it wrong once and you kill the SCC or inverter, maybe the battery.

<snip>
You can put 10s2p into one MPPT. Available current exceeds 18A, simply won't be used. Is there a "max short circuit current" spec on the inverter?
<snip>

Inverter short circuit current spec......never heard of that. Even in the context of MPPT input.

 

[jarred125]

Found our interconnection documentation, for the RSD:

A disconnect switch that meets the following criteria:

(i) The disconnect switch must be capable of isolating the DER for the purposes of safety during EDU system maintenance and during emergency conditions.

(ii) The disconnect switch should be located externally to the equipment package and must be accessible to and lockable by EDU personnel, within an acceptable distance of the meter as determined by the EDU, at either the primary voltage level, which may include load-break cutouts, switches, and elbows, or at the secondary voltage level, which may include a secondary breaker or switch.

(iii) The disconnect switch must be clearly labeled as a DER disconnect switch.

(iv) The disconnect switch will be provided, installed, and paid for by the applicant, whether it is an integrated feature of the equipment package or a compatible external device.

So it seems to be a single point of disconnect is what they require.

 

[Hedges]

Yes, utility cares about simply disconnecting, so you don't backfeed and electrocute a lineman.
My original installation, that was a visible-blade disconnect. Today, option not to install one, but utility might unplug meter to isolate.

RSD would be a code requirement, for fireman safety, so they can hack a hole through panels and roof if needed to vent smoke. Or maybe just to direct water on damaged array. (Also likely a setback requirement from ridge, and a walkway.)

Earlier RSD was just low voltage outside array. Later RSD was < 80V between any wires even within array.
Does not apply if ground mount or some outbuildings, just rooftop of dwelling.

 

[Hedges]

Inverter short circuit current spec......never heard of that. Even in the context of MPPT input.

"Max. operating input current per MPPT ... 10 A
Max. short circuit current per MPPT ... 18 A"

https://files.sma.de/downloads/SBxx-US-DS-en-41.pdf

One reason is they have a diode to short out reverse-polarity connected array. Inverters without that get killed regularly by installation mistakes.
Some products also have a pulldown FET as over-voltage protection, but I don't think these models do.
It is probably also related to U.L. tests; no matter what fails inside inverter you don't want it to catch fire.

 

[jarred125]

Yes, utility cares about simply disconnecting, so you don't backfeed and electrocute a lineman.
My original installation, that was a visible-blade disconnect. Today, option not to install one, but utility might unplug meter to isolate.

RSD would be a code requirement, for fireman safety, so they can hack a hole through panels and roof if needed to vent smoke. Or maybe just to direct water on damaged array. (Also likely a setback requirement from ridge, and a walkway.)

Earlier RSD was just low voltage outside array. Later RSD was < 80V between any wires even within array.
Does not apply if ground mount or some outbuildings, just rooftop of dwelling.

Ill still check code but this is for a ground mount system, my roof is not something I want to add a ton of weight to .. it was not built that great from the manufacturer.

Aside from that, the calculations I just did for the low temperature Voc would be 56.4 (-32C is the lowest we have seen on record) and the change in Voc is 0.27% per °C. So it appears that 10 would be the max for this set of panels.

 

[Hedges]

Ground mount no RSD. Unless maybe because DC wires run into house; still, just an external DC disconnect ought to be sufficient.

So my +16% on voltage was close. I assumed -25C and 0.4%/degree.

Ground mount, if you've got the sun exposure, two orientations could extend production for more hours.
I aimed mine at 2:00 PM sun due to afternoon time of use rates, but rate schedules have different hours now.
More important for off-grid loads and having battery charge kept high later in the day.

You probably have peak wattage limits due to 120% rule for PV breakers in panel and continuous current 80% of breaker rating.
It is common to have 7.7kW inverter on 200A panel or 3.8kW inverter on 100A panel, but some allow more and there are some ways around it.

I put in SMA Sunny Boys long ago, Sunny Island recently.
My setup does (edit: NOT) time charging/discharging to game time of use rates. I just enable loads accordingly.
(edit: Other models and brands have settings for shaving peak loads or storing power to export later. Not a feature of the battery inverter I'm using.)

 

[jarred125]

Ground mount no RSD. Unless maybe because DC wires run into house; still, just an external DC disconnect ought to be sufficient.

So my +16% on voltage was close. I assumed -25C and 0.4%/degree.

Ground mount, if you've got the sun exposure, two orientations could extend production for more hours.
I aimed mine at 2:00 PM sun due to afternoon time of use rates, but rate schedules have different hours now.
More important for off-grid loads and having battery charge kept high later in the day.

You probably have peak wattage limits due to 120% rule for PV breakers in panel and continuous current 80% of breaker rating.
It is common to have 7.7kW inverter on 200A panel or 3.8kW inverter on 100A panel, but some allow more and there are some ways around it.

I put in SMA Sunny Boys long ago, Sunny Island recently.
My setup does time charging/discharging to game time of use rates. I just enable loads accordingly.

Fantastic, I feel like I am not as out of touch with reality as I thought I was. I will check into the peak wattage limits. Thank you, thank you thank you for sharing your knowledge with me. I greatly appreciate it!

 

[ncsolarelectric]

Whew! Thank you for commenting here and helping me out!

I will review the record lows for my area and calculate based off of that.
The data sheet from Fronius for this inverter says : Max. array short circuit current (1.5 * Imax) 27A / 27A

The inverter will be grid tie, we do have net metering (Duke Energy in Cincinnati, OH). I will check on the requirement for the RSD per panel but my initial thoughts based on what I have seen installed locally is they are not required per panel.

I will reach out to my provider today and see what the requirements are. Thank you again!

OH uses NEC 2017 for now, so Rapid Shutdown is a requirement on a building. Hence, why microinverters are a much easier choice for compliance.
OH Codes

 

[ncsolarelectric]

Found our interconnection documentation, for the RSD:

A disconnect switch that meets the following criteria:

(i) The disconnect switch must be capable of isolating the DER for the purposes of safety during EDU system maintenance and during emergency conditions.

(ii) The disconnect switch should be located externally to the equipment package and must be accessible to and lockable by EDU personnel, within an acceptable distance of the meter as determined by the EDU, at either the primary voltage level, which may include load-break cutouts, switches, and elbows, or at the secondary voltage level, which may include a secondary breaker or switch.

(iii) The disconnect switch must be clearly labeled as a DER disconnect switch.

(iv) The disconnect switch will be provided, installed, and paid for by the applicant, whether it is an integrated feature of the equipment package or a compatible external device.

So it seems to be a single point of disconnect is what they require.

That's not the same as Rapid Shutdown. The switch needs to initiate de-energization of the wiring from the panels to the home. Microinverters do this automatically because the wires carry AC, when the switch is off, the AC is off. But with a string inverter, some don't discharge the DC within 30 seconds, so a contactor is required at the inverter, and TIGO or HMRSD rapid shutdown devices are required at the panels. The wires must be disconnected at both ends. It's a pain to setup, so I can understand why most off-grid systems are ground mounts to avoid RSD.

 

[ncsolarelectric]

I just compared the Fronius 8.2-1 with the Hoymiles HMS-2000-4T-NA.

Fronius Max Continuous Current = 34.2A @240V, 8.208 kW

HMS-2000 Max Continuous Current, 7.99 x 5 = 39.95A @240V, 9.588kW, or nearly 17% more power from the same 20 panels.

That's why I suggested them. These microinverters will give you the full available power without clipping. I doubt that you can get better performance than that from those panels with a string inverter.

 

[jarred125]

OH uses NEC 2017 for now, so Rapid Shutdown is a requirement on a building. Hence, why microinverters are a much easier choice for compliance.
OH Codes

You mentioned here for a building, I am guessing this wouldn't apply to standard ground mount installation? I need to read the code.

 

[ncsolarelectric]

You mentioned here for a building, I am guessing this wouldn't apply to standard ground mount installation? I need to read the code.

Correct, if you go with a ground mount, rapid shutdown is not required. Only on a building.