Transformer sizing for PV system

[Master of None !]

Hello all, I think i know the answer , but looking to see if anyone has another approach to it. I have a PV installation that used SMA tripower which is 480 volt. Im connecting( Back feeding from solar ) to 3phase 208volt MDP . If i do basic sizing , i have an max output production at the 480 side of 90 amps , im looking at a 75 KVA transformer which lists FLA at 208amps on the 208volt side. Im sizing my overcurrent on the 208volt size at 250amps OCPD. It makes sense on paper , but just throwing me for a little loop when looking at the calcs. Anyway any thoughts would be appreciated . Thanks

 

[Smart $]

Is that 90/208 amps before or after applying the 125% to the inverter's rated [maximum] output current?

 

[PVfarmer]

SMA tripower which is 480 volt.

max output production at the 480 side of 90 amps ,

im looking at a 75 KVA transformer

Smart$ -Is that 90/208 amps before or after applying the 125% to the inverter's rated [maximum] output current?

Edit: Yes, right if those are 3 30A breakers and 72A of actual output then 75KVA is fine.

But if it's 90A of output current, unfortunately, 75kVA isn't large enough, according to this. You're limited to 81.2 amps with a 75kVA.
Not really sure why, but it is official.
Three 20000TL-US, yes. Three 24000TL-US, no.

The maximum apparent power of all inverter (SPV= ΣSPVi) connected to the low voltage side is:
• Less than or equal to 90% of the rated power of the transformer (SXMFR) SPV≤ 0,9 · SXMFR.

http://files.sma.de/dl/7418/STP24-US_MV_Trafo-TI-en-10.pdf

 

[Master of None !]

Transformer sizing for Solar

Transformer sizing for Solar

Edit: Yes, right if those are 3 30A breakers and 72A of actual output then 75KVA is fine.

But if it's 90A of output current, unfortunately, 75kVA isn't large enough, according to this. You're limited to 81.2 amps with a 75kVA.
Not really sure why, but it is official.
Three 20000TL-US, yes. Three 24000TL-US, no.

The maximum apparent power of all inverter (SPV= ΣSPVi) connected to the low voltage side is:
• Less than or equal to 90% of the rated power of the transformer (SXMFR) SPV≤ 0,9 · SXMFR.

http://files.sma.de/dl/7418/STP24-US_MV_Trafo-TI-en-10.pdf

Ok , So the install design and size changed . Here is more where the design will be. Its now 4 -30kTTL and 1 -20kTTL. From SMA data sheet , 36.2 Amp on the 30k and 24 A on the 20k. This is the max output current added together is 168.8 amps . If I do the 125% rule , heres what i get .

168.8 A x 125% is actually 211 amps @480v side . 211 amps would be a 225 KVA transformer . The 208volt side is 624 FLA . ( got those off ACME transformer chart ).

It seems like an awfully big sizing if i go with those calcs. I can understand if i was sizing a transforemer based on standard loads . Basing how many compresors or motors , lighting load etc. But is transformer sizing on the production side with solar sized the same way ?

The way i see it , 168amps is the largest , ideal perfect storm generation. Adding in the 125% is load based . I would think i go with 150 KVA transformer which lands 180amps on the 480 volt side .

Open for help. Thanks.

PS I really appreceiate the oppurtunity to be able to reach out to a knowlegable community.

 

[ggunn]

Ok , So the install design and size changed . Here is more where the design will be. Its now 4 -30kTTL and 1 -20kTTL. From SMA data sheet , 36.2 Amp on the 30k and 24 A on the 20k. This is the max output current added together is 168.8 amps . If I do the 125% rule , heres what i get .

168.8 A x 125% is actually 211 amps @480v side . 211 amps would be a 225 KVA transformer . The 208volt side is 624 FLA . ( got those off ACME transformer chart ).

It seems like an awfully big sizing if i go with those calcs. I can understand if i was sizing a transforemer based on standard loads . Basing how many compresors or motors , lighting load etc. But is transformer sizing on the production side with solar sized the same way ?

The way i see it , 168amps is the largest , ideal perfect storm generation. Adding in the 125% is load based . I would think i go with 150 KVA transformer which lands 180amps on the 480 volt side .

Open for help. Thanks.

PS I really appreceiate the oppurtunity to be able to reach out to a knowlegable community.

Unfortunately, you cannot rely on production estimates to size your equipment. It must be sized to 125% of your combined inverter nameplate rating.

 

[PVfarmer]

ggunn Unfortunately, you cannot rely on production estimates to size your equipment. It must be sized to 125% of your combined inverter nameplate rating.

Which code section is that for PV step-down xfmrs?
Because then that SMA pdf should say 80%, not 90%, regardless of step up or down..

This is the max output current added together is 168.8 amps . If I do the 125% rule , heres what i get .

168.8 A x 125% is actually 211 amps @480v side . 211 amps would be a 225 KVA transformer . The 208volt side is 624 FLA . ( got those off ACME transformer chart ).

Basing how many compresors or motors , lighting load etc. But 1) is transformer sizing on the production side with solar sized the same way ?

The way i see it , 168amps is the largest , ideal perfect storm generation. 2) Adding in the 125% is load based . I would think i go with 150 KVA transformer which lands 180amps on the 480 volt side .

I'd call SMA or an SMA dealer about that 90% thing, instead of going by my theory- that pdf is talking about 480V to MV, not 480V to 208V...so the MV shown in the pdf is 208/120V 3-phase in your case.

Which leads to a very important point- you can make 3-phase 208Y/120V out of 3x Sunny Boys.
(As in 3x SB 5000TL-US = one 15000TL Tripower, just @ 208/120 3-ph instead of 480/277.
3x 4000TL = 1 12000TL Tripower....6x 4000TL = a 24000 Tripower....6x 5000TL = 30000 Tripower)

Yes, it's more $$, but...225kVA xfmrs are also lots of $$!
There are many variables in your case that we don't know...

The way I would personally go is (if I had to buy a big xfmr) is oversize the xfmr for the 168.8A of PV output by 154%.
Because.... I'd be looking at a "exceeds DOE efficiency at 65% of rated kVA" xfmr.
168.8 / .65 = 260A.
260A * 480V * 1.73 =215kVA.
Boom, 225kVA is perfect!

(4 -30kTTL and 1 -20kTTL) also = you have 140kVA of PV output.
140 / .9 (SMA's 90%)= you'd want a 155kVA xfmr.
-and-
140 * 1.25 (code's 125%) = you'd want a 175kVA.
Those last two don't exist.

Therefore, 150kVA isn't enough.

My general theory is: with the previous 90kW of output from Tripowers (also = to 90KVA with 480 3-phase).

If you use a 75kVA xfmr, you're running it at 120%
With a 112.5kVA, you're running it at 80%.

Obviously, 80% is more efficient than 120%.
Therefore, 90% is more efficient than 100%.
But...65% makes more % of $$!

I'd be surprised if SMA said "if you run them at 92% of xfmr KVA they will shut off"... but you never know until you ask.

I'd personally go for 225kVA or 112.5 even if 150kVA or 75 was somehow allowed and the 208V loads were < 50 amps.

---

Re: the bolded stuff-

1- I believe you have to take both sides of the xfmer (PV output side // grid-load side) into account. (see #2)

2- Adding 250% could be load based, too. (Motor inrush)

 

[jaggedben]

Which code section is that for PV step-down xfmrs?
Because then that SMA pdf should say 80%, not 90%, regardless of step up or down..

I'm not entirely sure what the code requirement is, but if you have a manufacturer requirement and a code requirement that aren't the same you just have to keep track of which is greater. It would be helpful if SMA was crystal clear about what their documentation applies to, but it could apply to places where the NEC isn't code and you can't expect them to always be 100% on top of what the various local codes are.

I could be wrong but I believe that a transformer may not be required to be rating 125% of the inverter output rating. Unlike conductors and OCPDs.

 

[electrofelon]

I'm not entirely sure what the code requirement is, but if you have a manufacturer requirement and a code requirement that aren't the same you just have to keep track of which is greater. It would be helpful if SMA was crystal clear about what their documentation applies to, but it could apply to places where the NEC isn't code and you can't expect them to always be 100% on top of what the various local codes are.

I could be wrong but I believe that a transformer may not be required to be rating 125% of the inverter output rating. Unlike conductors and OCPDs.

The way I see it:
1. Transformers have no 125% rule/continuous load fudge factor. In fact, 110.3(B) is the only thing that indirectly specifies the sizing/loading of transformers.
2. I dislike these silly manufacturer requirements like this one saying the transformer must be sized at 90%. They should just stick to making inverters. That said, if they state that in the instructions, 110.3 compels us to follow it.

 

[Smart $]

...
I could be wrong but I believe that a transformer may not be required to be rating 125% of the inverter output rating. Unlike conductors and OCPDs.

Transformer rated 125% is not required by the NEC... but indirectly it is required to be at least at or near 80%... but sizing that small would be a total blunder on the design. Going with 100% is a marginal design IMO.

 

[ggunn]

I could be wrong but I believe that a transformer may not be required to be rating 125% of the inverter output rating. Unlike conductors and OCPDs.

Duh, you are correct; I posted without thinking. I match kW output to kVA and pick the next size up.

 

[PVfarmer]

1 if you have a manufacturer requirement and a code requirement that aren't the same you just have to keep track of which is greater.
2 It would be helpful if SMA was crystal clear about what their documentation applies to,
3 but it could apply to places where the NEC isn't code and you can't expect them to always be 100% on top of what the various local codes are.
4 I could be wrong but I believe that a transformer may not be required to be rating 125% of the inverter output rating. Unlike conductors and OCPDs.

5 Transformers have no 125% rule/continuous load fudge factor. In fact, 110.3(B) is the only thing that indirectly specifies the sizing/loading of transformers.
6 I dislike these silly manufacturer requirements like this one saying the transformer must be sized at 90%. They should just stick to making inverters. That said, if they state that in the instructions, 110.3 compels us to follow it.

1- Agreed. However, I find this discussion at this link a bit... alarming.
( https://www.solarpaneltalk.com/foru...n/18207-sma-tripower-connects-with-lost-phase )
The installation consists of 4-SMA Tripower 20000TL inverters with a load side connection via a 480/208 Y-Y 100kW transformer to the MDP. The inverters are 3 phase 277/480 and the grid is 3 phase 120/208.

2- Right, the pdf is about connecting to MV, but it doesn't explain the why re: 90% at all, and I can't find a pdf about stepping down.

3- Delta and wye are the same in Germany....but the grid isn't...so...who knows?

4- Pretty sure there's a post on this forum about that somewhere.

5- Do you think the loads or the inverters are more important, if picking one of A,B,C?
Inverters aren't technically computers....or?
The grid doesn't like harmonics, so get a harmonic mitigating model makes sense to me.

A Energy efficient k-factor transformer designed to tolerate heating due to harmonics associated with nonlinear loads.
-
B Harmonic mitigating transformers are superior to K-Rated and conventional transformers in reducing voltage distortion (flat-topping) and power losses due to current harmonics created by single-phase, nonlinear loads such as computer equipment.
-
C These energy efficient general purpose transformers are designed for linear loads and are most frequently used for applications such as commercial buildings which will supply a variety of general loads.
http://www.temcoindustrialpower.com...8Y/120&sort=price - lowest&c=0005160&n=ACAg.p

6- I really don't know- maybe the impedance increases with a xfmr running at 98%, and it bothers the inverters somehow?

 

[GoldDigger]

Re point #6:
The lower the rating of the transformer, the higher the voltage drop (voltage rise) across the transformer. And the harder it is to properly detect the low impedance of the grid for anti-islanding purposes.
The difference between 80% and 125% is not enormous, but you have to draw the line somewhere.

Sent from my XT1585 using Tapatalk

 

[PVfarmer]

Re point #6:
The lower the rating of the transformer, the higher the voltage drop (voltage rise) across the transformer. And the harder it is to properly detect the low impedance of the grid for anti-islanding purposes.

This is intriguing, pardon what may be clueless questions and/or layman's terms...
(point #6 might have something to do with that link I posted in point #1 re:islanding then? Hmm.)

Example equipment is:
1. 3x 24.1kW (29A) inverters, 480Y/277V output, so the 3 inverters = 87A / 72.3kVA.
2. 75kVA xfmr = 480Y/277V to MV (or 208V)
3. (Side note: That'd be a 200A 208Y120V or 100A 480Y/277V "POCO service", but you'd be outputting 87A through a 90.2A/75kVA xfmr, sounds kinda-a-bit-too-close for me. (And 72.3KVA = 200.9A or 200.6A when converted to 208Y/120V...)


With the 72.3kVA/87A of PV the xfmr is running at 96.4% of its 75kVA/100% rating.

So when you said "lower the rating" you meant the 100% xfmr as compared to the PV output?

This lower rating could be said as "higher the (PV:xfmr rating) ratio", or that 96.4% (ratio is 0.964:1).

The xfrmr runs less efficiently at 125% (1.25:1) than it does at 80% (0.8:1), and less efficiently at 96.4% than at 90% or 80% or 65%.
This is because of higher voltage drop (voltage rise).

The inverter(s) have to detect the low impedance of the grid.

This low impedance is actually "approaching infinite" on grid side.
When you are going ---> inverter 480Y277V - connected to LV side / air gap/ then grid side /MV side of POCO's xfmr, the impedance of POCO xfmr is a factor.
(Side note: Now, if there is a step-down from inverter-to 208Y/120V xfmr, then another step-up-to-POCO MV xfmr involved, the impedance is MUCH different than straight 480-MV?)

But what I want to know is-
With this voltage drop/rise "across" the xfmr...which is caused by the xfmr being closer to its 100% rating and less efficient/running hotter...
do the inverters "have to"...or...do they "end up"... decreasing their voltage?

The inverters could have more DC power than they can handle and therefore be clipping, ok.
When so, they still want to put their max (72.3kVA for instance) output into a grid with as little impedance as possible, at as close to 480Y/277V as possible.

When say a 75kVA step-down xfmr is "holding them back" so to speak, it is because...

the xfmr is putting out <208V, and therefore "asking for" <480V from the inverters?

Could one say the 208V impedance is "clogging up" the xfmr on the inverters' side?

Are amps lost to heat because of lower voltage?
-or-
Do I have it backwards and the inverters "have to" or "end up" INCREASING their AC voltage to compensate for lower 208V voltage?

Or!
Does the xfmr just heat up due to physics because 72.3KVA of PV is 96.4% of its 100%, and the windings are more efficient at 90%, and voltage drops on both sides?

Thanks again, if this makes sense...