30kw PV Service Sizing

BackCountry

Electrician
Location
Southern California
Occupation
Licensed Electrician and General Contractor
I’m trying to plan for a job that we just bid, and I’m at a decision making point on how we’re going to tie in with our local utility.

Existing large ranch with two existing single phase 200A 240v 3w services. Our calc shows a 30kw sized PV array will handle the existing usage, we have net metering aggregation here where one meter’s backfeed can offset another meter’s consumption on the same parcel.

We’re looking at a 125A backfeed, which means we need to either use two 15kw 1ph Fronius inverters (only ones made that high in single phase to my knowledge), the tie in is 400ft from the array, so I’m thinking we take the voltage as high as we can on the DC side to overcome distance. We then either drop a third meter in with a 225A bus and derate the main breaker to 150A, thereby giving us a 120A backfeed capacity, or...

What I would much rather do is use a 35kw SMA tripower 480v 3ph inverter, and then either evaluate the cost of 3ph 480v service (the existing utility transformer is a 50kv single phase) which may be cost prohibitive, or not place a step down transformer in.

I could also upgrade the existing 200A panel to a 400a Siemens solar backfeed panel with two 200A disconnects and a third PV disconnect so no line side tap is required.

I appreciate any insight and wisdom. This will be ground mount, and is in an agricultural area. If I could have my way, I’d get a third 480v 3ph 4w service installed on a new meter and use a larger 3 phase inverter and call it a day. I’m not sure that’ll be possible or cost effective from our utility (PG&E).
 
Some thoughts/suggestions:

the 480 3 phase service would certainly be nice, but I am pretty sure that would be prohibitively expensive. I guess it wouldn't hurt to get a quote from the power company if it's not too much of a hassle.

One thing I frequently do for ground mounts is replace the meter socket with a class 320 socket. Then I use 230.40 exception 3 to run a set of service conductors to the ground Mount array. One last summer I did I ordered a custom triplexed USE with 2x 500 AL and the smallest allowed neutral (1/0). This was 16kw at about 600 feet. I did investigate sending the DC strings over (fronius Does 1kv strings if you didn't know) but found it to not be cost-effective and worth the hassle.
 

BackCountry

Electrician
Location
Southern California
Occupation
Licensed Electrician and General Contractor
That's very helpful, I didn't know Fronius did 1kw.

There is an existing 1-1/2" conduit running from the panel to the PV array that's being replaced (installed in 2003). By my calc at a 125A backfeed, I'll need 4/0 copper or 350 AL -- both are too big for that conduit.

I think you're right regarding utility fed three phase, they'll need another transformer at the utility side, likely expensive and going to take forever to get approved.

I hear you on the 320 socket -- you may have seen Siemens new solar ready 400 (320 continuous) panels that have a seperate meter connection that bypasses the bus, and can take up to 200A of backfeed. I use those often, but at 400ft, and an existing small conduit, that's my challenge.

I now see my options as:

Retrench and replace 1.5" with 2" or greater, trenching through a lot of irrigation and other customer conduits supplying their property's lighting and outbuildings

Install a three phase inverter, output at 480v, which means running three #6's and a ground, and install a 480 delta/240 stepdown transformer, output at 240v/125A and tie into the PV input of that Siemens panel I referred to above, as a service upgrade on their existing 200A panel.

Thoughts?
 
If there is a big advantage to using that existing pipe, which sounds like there is, then your biggest bang for your buck is using the fronius with 1kv strings, about 20 modules, and bringing the DC back. Then you are into to game of figuring out how to split those 100 modules across 2 inverters/4 MPPT's 🤔. I didn't do any math but that may fit in 1 1/2......

I don't follow the 480 to 240 transformer idea. Thought the existing service is single phase?
 

wwhitney

Senior Member
Location
Berkeley, CA
Occupation
Retired
the fronius with 1kv strings,
So I noticed from the Fronius Primo website that the inverters that have a maximum MPP of 800V all weigh almost twice the smaller inverters that have a maximum MPP of 480V. Do you know if internally the former handle higher string voltages by internally doubling most of the components and putting their DC inputs in series, while putting their AC outputs in parallel? Or something like that, my understanding of the subcomponents of an inverter is quite limited.

Cheers, Wayne
 
So I noticed from the Fronius Primo website that the inverters that have a maximum MPP of 800V all weigh almost twice the smaller inverters that have a maximum MPP of 480V. Do you know if internally the former handle higher string voltages by internally doubling most of the components and putting their DC inputs in series, while putting their AC outputs in parallel? Or something like that, my understanding of the subcomponents of an inverter is quite limited.

Cheers, Wayne
What are the "smaller" inverters you speak of? Are these other fronius or other brands? Be advised in case the former that there has been some confusion on some of the fronius sales literature for their "residential" inverters where they say they are 600v strings even though they in fact do 1kv.

I have used both fronius and sunny boy 7.7's and I don't recall the fronius being any heavier.
 

wwhitney

Senior Member
Location
Berkeley, CA
Occupation
Retired
Selected info from the spec sheet at https://www.fronius.com/~/downloads/Solar Energy/Datasheets/SE_DS_Fronius_Primo_EN_US.pdf
Where different models within each grouping have different specs, I noted the range with (varies)

FRONIUS PRIMO 3.8, 5.0, 6.0, 7.6, 8.2
47.29 lbs.
Nominal input voltage (varies) 410V-420V
Operating voltage range 80 V - 600 V
MPP Voltage Range low end (varies) 200-270V
MPP Voltage Range high end (varies) 400-480V
Max. input voltage 600 V (1000 V optional)

FRONIUS PRIMO 10.0, 11.4, 12.5, 15.0
82.5 lbs.
Nominal input voltage (varies) 655V - 680V
Operating voltage range 80 V - 1,000 V
MPP Voltage Range low end (varies) 220V-320V
MPP Voltage Range high end 800V
Max. input voltage 1000 V

I guess my main question is what the bigger models are doing inside that's different and makes the nominal input voltage around 650V vs 420V, and causes the weight to be much higher. Other basic questions I have:

- What is the difference between Operating voltage range and MPP Voltage range? I.e. what does the inverter do if the string voltage exceeds the high end of the MPP range but is still less than the Operating voltage?

- The maximum input voltage, is it enough to check the temperature corrected string STC OCV?

Cheers, Wayne
 
Selected info from the spec sheet at https://www.fronius.com/~/downloads/Solar Energy/Datasheets/SE_DS_Fronius_Primo_EN_US.pdf
Where different models within each grouping have different specs, I noted the range with (varies)

FRONIUS PRIMO 3.8, 5.0, 6.0, 7.6, 8.2
47.29 lbs.
Nominal input voltage (varies) 410V-420V
Operating voltage range 80 V - 600 V
MPP Voltage Range low end (varies) 200-270V
MPP Voltage Range high end (varies) 400-480V
Max. input voltage 600 V (1000 V optional)

FRONIUS PRIMO 10.0, 11.4, 12.5, 15.0
82.5 lbs.
Nominal input voltage (varies) 655V - 680V
Operating voltage range 80 V - 1,000 V
MPP Voltage Range low end (varies) 220V-320V
MPP Voltage Range high end 800V
Max. input voltage 1000 V

I guess my main question is what the bigger models are doing inside that's different and makes the nominal input voltage around 650V vs 420V, and causes the weight to be much higher. Other basic questions I have:

- What is the difference between Operating voltage range and MPP Voltage range? I.e. what does the inverter do if the string voltage exceeds the high end of the MPP range but is still less than the Operating voltage?

- The maximum input voltage, is it enough to check the temperature corrected string STC OCV?

Cheers, Wayne
I believe those specs are wrong. Like I said there has been some confusion on this. I know for a fact I have a primo 7.6 with 18 modules per string. I believe all primos do 1kv
 

Vines41

Member
Location
Silicon Valley CA
Occupation
PV Designer
Selected info from the spec sheet at https://www.fronius.com/~/downloads/Solar Energy/Datasheets/SE_DS_Fronius_Primo_EN_US.pdf
Where different models within each grouping have different specs, I noted the range with (varies)

FRONIUS PRIMO 3.8, 5.0, 6.0, 7.6, 8.2
47.29 lbs.
Nominal input voltage (varies) 410V-420V
Operating voltage range 80 V - 600 V
MPP Voltage Range low end (varies) 200-270V
MPP Voltage Range high end (varies) 400-480V
Max. input voltage 600 V (1000 V optional)

FRONIUS PRIMO 10.0, 11.4, 12.5, 15.0
82.5 lbs.
Nominal input voltage (varies) 655V - 680V
Operating voltage range 80 V - 1,000 V
MPP Voltage Range low end (varies) 220V-320V
MPP Voltage Range high end 800V
Max. input voltage 1000 V

I guess my main question is what the bigger models are doing inside that's different and makes the nominal input voltage around 650V vs 420V, and causes the weight to be much higher. Other basic questions I have:

- What is the difference between Operating voltage range and MPP Voltage range? I.e. what does the inverter do if the string voltage exceeds the high end of the MPP range but is still less than the Operating voltage?

- The maximum input voltage, is it enough to check the temperature corrected string STC OCV?

Cheers, Wayne
Without digging into the rest of the question too deeply, I think the 1000vdc versions are sharing a larger form factor. I do not really know much about the Fronius line unfortunately. Looking at dimensions alone, the larger has a volume of 5073 in^3 and the smaller has 3381 in^3, with weights of 47.3 and 82.5 lbs, seems reasonable.

Operating Voltage defines whether the inverter is capable of firing up and inverting. The MPPT range is where the inverter can be said to be within its efficiency window defined by CEC. If it exceeds MPPT voltage, but is still under or over its operating voltage, it will invert with less efficiency. I cannot easily quantify how much less efficient. Generally the best inverters have a very wide MPPT voltage range, as well as a high efficiency across that range.

Maximum input voltage is checked by taking the record low temperature compared with 25C, and multiply that by the module coefficient, then adding that whole thing to the Voc of the module at STC to get the absolute maximum module voltage.
 

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
....

There is an existing 1-1/2" conduit running from the panel to the PV array that's being replaced (installed in 2003). By my calc at a 125A backfeed, I'll need 4/0 copper or 350 AL -- both are too big for that conduit.

....
Maybe revisit that calc? Do you have a massive temp derating? (Thought this was an underground pipe.)

125A is 2awg or 1awg copper with no temp derating. I realize that you may want to run DC strings, but still...
 

BackCountry

Electrician
Location
Southern California
Occupation
Licensed Electrician and General Contractor
Hey Ben,

This is at 400 feet, I’m showing 125A at 240vCU at 4/0?

I just learned that Fronius is also not rule 21 California rapid shutdown compliant...
 

BackCountry

Electrician
Location
Southern California
Occupation
Licensed Electrician and General Contractor
I shouldn’t have focused on rapid shutdown — rule 21 is aimed at smart grid requirements. Likely can be fixed via a firmware update, just no eta yet.

The 480v step up comment before: that was just an idea to overcome voltage drop.
 
Hey Ben,

This is at 400 feet, I’m showing 125A at 240vCU at 4/0?

I just learned that Fronius is also not rule 21 California rapid shutdown compliant...
I would agree with you on the 350 AL as being about right. To nitpick a bit, you didnt say whether the 30kw was AC or DC. Sounds from your wording like that might be DC, in which case, your AC kw might be around 90% of that which would lower your current a bit.

Actually you can fit (2) 3/0 and (1) #4 CU in a 1 1/2 PVC with 3.6% VD/ 9 volts rise which isnt too bad (actually it would be a bit less as that calculator assumes .9 PF. Jam ratio sucks, probably would want to get it plexed. Also I cringe at the cost of that length in copper, but I cant say what your cost of trenching hassle is. Also 250/250/2 AL fits believe it or not.

350/350/(4/0) URD AL is about $3 per foot and is what I would probably run if I was starting new. Just for your information that 500/500/(1/0 plexed AL I custom ordered last summer was $3.65/ft. 600/600/(3/0) had a pretty big price jump up to $4.73/ft (over 500 triggers having to go to 3/0 as minimum grounded conductor size).
 

BackCountry

Electrician
Location
Southern California
Occupation
Licensed Electrician and General Contractor
That’s not bad. We just ordered 4/0cu at $3/ft. It could probably be done, but running 1000v DC is starting to sound better.

My 30kw was DC ballpark, it’s more like 26 after the calc. This a common problem for us with ranches, how to make it work at high volume with 240v 1ph 200A service.
 
That’s not bad. We just ordered 4/0cu at $3/ft. It could probably be done, but running 1000v DC is starting to sound better.

My 30kw was DC ballpark, it’s more like 26 after the calc. This a common problem for us with ranches, how to make it work at high volume with 240v 1ph 200A service.
You would think the 1000VDC would be so much better, but If you run the numbers I think you will not see much advantage (maybe your case is different with the existing pipe situation thrown in). For one thing, I can get a sunny boy 7.7 for about $1300 where a fronius 7.7 is like $1700 - that buys a lot of cheap aluminum conductor, so that right there brings me (maybe your pricing structure is different) out of the 1000V realm and down to 600V. Then if you think about it, you kinda have to have the perfect setup to take advantage of the 1000V strings when you consider dividing the modules across multiple inverters and MPPT's, which again often puts you into 600V territory again anyway. So now, price out copper PV wire vs a big cheap aluminum service/feeder run and you really dont get ahead (I know PV wire is available in AL, but I have not run the numbers with that).

Try my 230.40 Ex #3 idea for ground mounts, it works real nice. You just change the socket and now you have a nice 200A worth of tie in available and only three conductors required out to the array.
 
I’m on the same page as you. Look at this spec sheet: https://www.siemens.com/download?BTLV_44421

That unit is the problem solver. Just impossible to find.
I dont typically need a disconnect or panel board built in, I just use a class 320 socket only. One set of SEC goes to the array, the other refeeds the original service disconnect. We don't need the outside disconnect here.
 

BackCountry

Electrician
Location
Southern California
Occupation
Licensed Electrician and General Contractor
Same idea here just already built — and already EUSERC compliant. Most of the time that’s easier on our end since we deal with CSED panels. I know we’re all different.
 

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
I admit I didn't think about voltage drop over 400ft. However if you run DC in that pipe it may be an option to just compromise on voltage drop. There was a paper in Solar Pro magazine a few years back that argued that upsizing for DC voltage drop was break-even on cost.
 
Top