Transformer sizing for step up and step down transformers

Status
Not open for further replies.

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
but what would be happening with H2 from the 480V side and X2 on the 120/240V side?
Don't yell at me!
But I can sort of see how, if current has to go from X2 to X1 and X3, (if that's physically possible) it would cause some serious issues.

Nothing would be happening at X2 because there's nothing at the service to hook it up to. Or rather, anything happening at X2 is just flowing through the transformer coils between X1 and X3 as if there's no terminal at X2.

Ergo on the 480V side there will not be a proper voltage reference, and the inverter will not turn on. And thus nothing would be happening on either side, as far as current flow is concerned. Golddigger or someone else will have to confirm for you what the voltage measurements would be from H2 to other terminals. I could take a guess but I may get it wrong. (Hey, that's something you should try PVfarmer, i.e. keeping your mouth shut when you're not sure of a statement. ;))

Now supposing a 480V 3-phase generator that is not an interactive inverter was hooked up to the 480 side, and somehow properly synced to the grid, then I gather that any loads hooked to the 240/120 phase will draw current. But that's where the 5% problem with the transformer would come in. Again, an engineer could explain why that problem comes up. In any case, iit would be illegal without a transfer switch, and we're just trying to understand theory here because actually doing it would be stupid.
 

kwired

Electron manager
Location
NE Nebraska
Could you elaborate on that some?
"Supply three single phase loads"... but not with a GTI alone, right?

If you had a 15kW 480V 3ph grid tied inverter.
That's 18A at 480V and 62.5A at 240V.

But you can't just attach a grid tied inverter to three 480V to 120/240V 1 ph xfmrs and then 1 motor each to those that draw 5000w / 20.83A each, without the grid being involved, correct?
The motors wouldn't give a "reference voltage" to the inverters...at all?
There has to be a MV grid and step up transformer attached to any brand of GTI for reference AND protection I believe.



Right! Like I said before, over 50kVA of 1 phase PV to grid around here would get a "why?" from the POCO, and it is 100% their call on allowing it.
My comment was just general purpose transformer information, which I think the unit mentioned is nothing more then a 480 x 240/120 with high leg general use transformer. Not impossible to supply single phase loads from a three phase transformer, but say you have a 75 kVA transformer, you are not going to supply a single 75 kVA single phase load but you can supply three 25 kVA loads if they are balanced across the phases.

That said forget any 120% rules for the moment that relate to PV applications or the fact the inverter needs to see utility voltage to "turn on" and try to reverse feed said three phase transformer - you would be able to get single phase out - but it would only be at one third of the full kVA rating before overloading occurs as you are only connecting to one third of the coils in the transformer.
 
1 Nothing would be happening at X2 because there's nothing at the service to hook it up to. Or rather, anything happening at X2 is just flowing through the transformer coils between X1 and X3 as if there's no terminal at X2.

2 Ergo on the 480V side there will not be a proper voltage reference,

3 we're just trying to understand theory here because actually doing it would be stupid.

1 Thanks, that's what I figured- there's an air gap between H side and X side, but it means nothing because only X1 and X3 are connected.

2 Therefore the inverter has the *wrong* voltage ref, outside of any of it's parameters or settings.

3 I'll try to make my theories sound less like statements then! :huh:
Yes, it is a bad idea overall- why create a "bottleneck" on the customer side AND an imbalance on the POCO side, even if there is still a magical 3ph to 1ph setup that would "work".
You can't even combine Sunny Island 1ph with any Tripower 3ph (I think in Europe that flies, so maybe here...next year?)

try to reverse feed said three phase transformer - you would be able to get single phase out -

Aha. The three phase inverter is only "seeing" a single phase grid, in other words. I'm looking at the "reverse feed" as the same circuit as the 120/240V xfmr output.
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
As for that 5% limitation, it does not apply to 240V line to line loads from A to C or even balanced 120V loads. AFAIK it applies to loads corresponding to current in the neutral (i. e. unbalanced loads).
The existence of the center tap wire has no effect on the ability or inability of the transformer to handle unbalanced 240V loads.
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
Nope. Because of grounded conductor and voltage to ground limitations I do not see any point (short of some weird industrial process maybe) for having a center tap on more than one of the three (or two for open delta) windings.
All 120 V loads will have either A or C as the hot conductor and will use the same grounded neutral.
So I am referring to balanced 120V loads which are electrically indistinguishable from a 240V load between A and C.
A set of 120V loads all wired from A to neutral only would be using only 1/6 of the secondary winding length and so from that alone would be limited to no more than 16%.
But there are other heat flow and core flux issues that lead the manufacturer to impose an even lower 5% limit.
 
So I am referring to balanced 120V loads which are electrically indistinguishable from a 240V load between A and C.
A set of 120V loads all wired from A to neutral only would be using only 1/6 of the secondary winding length and so from that alone would be limited to no more than 16%.
But there are other heat flow and core flux issues that lead the manufacturer to impose an even lower 5% limit.

Thanks a lot- core flux issues as in trapped harmonics? I think I've got it.

The balanced 120V loads are still limited to 33% without heat and flux issues, so 10% with the issues by the maker then?

So with a 75kVA 240D CT xfmr, that would be 7.5kVA available total split between A and C, which is only 62.5 amps at 120V. A lot less than the 550 or so amps of 120V being proposed!

zman- Really! Check with the POCO about 480/277V service.
 
Thanks a lot- core flux issues as in trapped harmonics? I think I've got it.

The balanced 120V loads are still limited to 33% without heat and flux issues, so 10% with the issues by the maker then?

So with a 75kVA 240D CT xfmr, that would be 7.5kVA available total split between A and C, which is only 62.5 amps at 120V. A lot less than the 550 or so amps of 120V being proposed!

zman- Really! Check with the POCO about 480/277V service.

Just remember that the "center tap limit" would be typical of a three phase, "single unit" transformer. You could build whatever you wanted out of three single phase units, and a utility could of course do the same with pole mounts. I have seen both setups with a utility bank: An open delta with relatively large center tapped pot implying limited three phase loads, and a closed delta with all the pots the same size implying limited 120V loads.
 

zman990

Member
Location
US
Nothing would be happening at X2 because there's nothing at the service to hook it up to. Or rather, anything happening at X2 is just flowing through the transformer coils between X1 and X3 as if there's no terminal at X2.

Ergo on the 480V side there will not be a proper voltage reference, and the inverter will not turn on. And thus nothing would be happening on either side, as far as current flow is concerned. Golddigger or someone else will have to confirm for you what the voltage measurements would be from H2 to other terminals. I could take a guess but I may get it wrong. (Hey, that's something you should try PVfarmer, i.e. keeping your mouth shut when you're not sure of a statement. ;))

Now supposing a 480V 3-phase generator that is not an interactive inverter was hooked up to the 480 side, and somehow properly synced to the grid, then I gather that any loads hooked to the 240/120 phase will draw current. But that's where the 5% problem with the transformer would come in. Again, an engineer could explain why that problem comes up. In any case, iit would be illegal without a transfer switch, and we're just trying to understand theory here because actually doing it would be stupid.
I asked simple question it got complicated obviously so wow have some respect
5. Dude- seriously. From your link. It doesn't say it supports 120/240V grids anywhere.
Data sheet-
SE33.3K
(REQUIRES MEDIUM VOLTAGE TRANSFORMER) // Grids Supported - Three Phase 3 / N / PE (WYE with Neutral)

Plus, 480V to 240/120V is *not* a MV xfmr!

1. I am doing it another way in one case- getting 480/277V service for the 480/277V inverters. Not SE inverters, SMA.
So there are no xfmr losses on customer side going from PV to grid. The farm's loads are 120/240 split phase, but almost all the motors say 208V on their plates, so 45kVA of xfmr with 208/120V 3ph secondary for the 25kVA max of loads.

2. Check your local zoning- a real farm can be in a residential zone, but there are agricultural "variances". I mean as in, if the farm is incorporated as an agricultural business, the farm buildings aren't dwellings.
Aren't you inverters going outside, so all of the 600+V wires will be outside?
Also, look at page 2 here, about Section 101.31.
http://solarenergy.advanced-energy.com/upload/File/Application Notes/ENG-600vor1000V-260-02.pdf

3. A 112.5kVA xfmr for TWO of the 33kW inverters? No!
As iwire pointed out-
the single-phase load should not exceed 5 percent... Additional loading beyond 5 percent may cause the transformer to overheat and fail.

Your 112.5kVA xfmr can only handle 5.625kVA of 120/240V!
You'd want a 1500kVA xfmr- 5% of that is 75kVA, enough for your 66kVA of PV.
The largest 480V to 240/120V xfmr I can find is 750kVA.
BUT- you have three or four engineers (and myself) telling you that what you want to do isn't possible. Even if it was, that would be WAY too much $$!

4. I have to say that doing the long run at 480/277V and then stepping *down* the voltage sort of defeats the purpose of doing the run at higher voltage- you'd want to step UP to the grid at that point (especially with a line/supply side connection!) and step down for load only.

5. 66kVA of 480 3ph PV output = 550 amps at 120V. You can't land that with 400A service, which is a 50kVA xfmr.
You have to get a 75kVA 480/277V service minimum. My engineer recommended getting 150kVA of 480/277V service for about the same amount of PV as you are installing, so the PV will be easily "expandable".
The SE inverter specs say "MV transformer required" - I read that as you can't even use a 480/277V to 208/120V step down xfmr for the 480/277V inverters and a 208/120V service..


Sent from my HTC One_M8 using Tapatalk
 
Last edited by a moderator:

zman990

Member
Location
US
Regardless I'm a solar integrator and anything is possible I love what I do it saves my life everyday.
I'm a problem solving soldier in the war against solar energy and big business aka utilities.
I do believe still this is possible and if the transformer or transformer's are cheaper that having to mount six 10kw inverters and or 55amp *6 and not being able to use my high wattage panel.
Then I will be pissed because when it comes to using 300mcm or #1 copper and turning 6 inverters into two inverters then I am all about speed and efficiency and optimization of every panel sold.
I know nothing about transformers just solar this is my next step in the evolution of my career I want to know everything about this business and transformer both up and down is pretty much my last hurdle.
I have a very good master electrician I sub and he will figure this out.
I truly feel getting oflver 1000volt dc system is the biggest hurdle.
Also if I was going to install the 33kw into a 480/277 service at commercial property I would not need a transformer to backfeed just a 50amp 3pole breakers simple.
I'm not here to argue with people I just feel I am right in saying this is possible and if transformers are not outrageous it will be most cost effective method of install on this job site.
Contract isn't signed anyway I hope it does.
So I can attach my approved permit and inspection



Sent from my HTC One_M8 using Tapatalk
 
Last edited by a moderator:

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
Keep up the good fight soldier, but when you question a general's judgement (or even more a top sergeant's) the odds are not in your favor.
 

kwired

Electron manager
Location
NE Nebraska
Regardless I'm a solar integrator and anything is possible I love what I do it saves my life everyday.
I'm a problem solving soldier in the war against solar energy and big business aka utilities.
I do believe still this is possible and if the transformer or transformer's are cheaper that having to mount six 10kw inverters and or 55amp *6 and not being able to use my high wattage panel.
Then I will be pissed because when it comes to using 300mcm or #1 copper and turning 6 inverters into two inverters then I am all about speed and efficiency and optimization of every panel sold.
I know nothing about transformers just solar this is my next step in the evolution of my career I want to know everything about this business and transformer both up and down is pretty much my last hurdle.
I have a very good master electrician I sub and he will figure this out.
I truly feel getting oflver 1000volt dc system is the biggest hurdle.
Also if I was going to install the 33kw into a 480/277 service at commercial property I would not need a transformer to backfeed just a 50amp 3pole breakers simple.
I'm not here to argue with people I just feel I am right in saying this is possible and if transformers are not outrageous it will be most cost effective method of install on this job site.
Contract isn't signed anyway I hope it does.
So I can attach my approved permit and inspection.
Oh yah love this shit


Sent from my HTC One_M8 using Tapatalk

You just said you know nothing about transformers, yet you continue to ignore the fact that at least 3 or 4 of us have told you there is no transformer that is capable of converting three phase to single phase (at the same kVA level in as out). It has also been mentioned that in general the inverters need to see all three phases coming from the utility side or they will not produce any output. That probably can be designed differently but likely presents other issues - AFAIK they are designed that way so they won't feed the utility if it is down for some reason.

Can they build a unit that converts the way you wish here - maybe, but it won't be a simple off the shelf transformer - it will have additional phase conversion equipment of some sort and likely some controls to make it all work. When all is said and done will raise the overall cost and possibly make you reconsider using more commonly used/available components.

The unit you provided a link to is three phase in three phase out. You can use single phase power off of the secondary side of it, but it is still producing three phase out. A single 240 volt single phase load would be limited to a third of the full kVA rating of the unit, and would be connected to either A-B, B-C or C-A phases. So assuming you got around other hurdles your 75 kVA unit you think you need needs to be three times larger because you are only loading one third of the unit, but that is only sized so the portion of secondary can handle the current, you still have an unbalance between primary phases and secondary phases to give you some issues.
 
1. I do believe still this is possible and if the transformer or transformer's are cheaper that having to mount six 10kw inverters and or 55amp *6 and not being able to use my high wattage panel.

2. I know nothing about transformers just solar this is my next step in the evolution of my career I want to know everything about this business and transformer both up and down is pretty much my last hurdle.

3. I truly feel getting oflver 1000volt dc system is the biggest hurdle.
Also if I was going to install the 33kw into a 480/277 service at commercial property I would not need a transformer to backfeed just a 50amp 3pole breakers simple.

4. I'm not here to argue with people I just feel I am right in saying this is possible and if transformers are not outrageous it will be most cost effective method of install on this job site.
Contract isn't signed anyway I hope it does.

1. Dude.

*Please* stop believing before you order anything. You don't seem to have enough details to order stuff.

Have you talked to the POCO yet at all? You can't just look at your 400A main meter/breaker and know the service xfmr size from that.
I'm currently working on a farm PV project- the owner of the farm said that his service was 200A...of *240V*. There's also a 400A main switch (which was double throw for a generator for blackouts).

But! It isn't- it's 200A of 120V. I had to talk to the POCO guy who informed me of the 25kVA size of the service xfmr.
So 100% of that 25kVA is 208A @ 120V and 104A @ 240V- there's a 400A main because of the 200A farm panel and 100A house panel. (both @ 120V)
The farm's load is nearly 120A @ 240V for all of the motors at once- so it is maxed out and could really use a 50kVA 120/240V xfmr regardless of any PV.

But the point here is that you *might* have a 50kVA 1ph service transformer- BUT that's not enough. Period.
Go back to your first comment- you seem to still be saying you want to put 66kW (which is also 66kVA) of PV through a 50kVA xfmr to the grid. You can't!

The 66kVA of PV output is:
66,000VA divided by 480V divided by 1.73= 79.48 amps- that's the 80A output of the two 33kW inverters.
66,000VA divided by 120V = 550 amps.

There is no possible way to put 550A of PV through a xfmr with 400A of capacity!! Which is what your 50kVA service xfmr is! 50,000VA / 240V = 208A...or 416A @ 120V.
You didn't convert 3ph amperage to 1ph- that's VERY important.

2. You want to go DOWN with any xfmr you are buying. See #3 here.

3. You keep implying that you are using TWO of the 33kW inverters- so you'd have a 50A breaker for each and a 100A breaker for the combined output, so you'd be backfeeding thru a 100A 480/277V breaker.
If you had 480/277V service. you could get a smaller xfmr, like I said before, 3ph 480/277V to 3ph 208/120V and run single phase loads from that.
Seriously, even if you could step your PV voltage down to grid, why would you? You'll lose most of what you gained by using 480/277V for your 350 ft run.

About the 1000V DC and ground mounted panels- oh wait, never mind, I was going to say check out SolarDock racking, but their website is down. That's not good.
Anybody want to jump in here?
As far as "inaccessible" and 1000V DC.
Um...SolarPro? Anything changed since April/May 2013? :)
http://www.sma-america.com/fileadmi...gn/SolarProMagazine_1000VDC_AprilMay_2013.pdf

4. There are certainly possibilities, but that pic of the 75kVA 480V delta to 240V delta 120V CT xfmr IS NOT one of them!
First of all- you *can't* connect the SE inverters to 480V delta. The installation manual says so pretty clearly.
 
The unit you provided a link to is three phase in three phase out. You can use single phase power off of the secondary side of it, but it is still producing three phase out. A single 240 volt single phase load would be limited to a third of the full kVA rating of the unit, and would be connected to either A-B, B-C or C-A phases. So assuming you got around other hurdles your 75 kVA unit you think you need needs to be three times larger because you are only loading one third of the unit, but that is only sized so the portion of secondary can handle the current, you still have an unbalance between primary phases and secondary phases to give you some issues.

Right on.
If he had some sort of double magically perfectly balanced 25kVA 240V 1ph loads on both A-B and B-C, and then the 120/240V CT on A-C going to grid- that would still be "weird", because A-B and B-C would be drawing 50kVA, so only 16kVA left from the 66kVA of PV inverters to go to grid.
If the 50kVA of 240V loads was suddenly switched off...um...well that'sounds like a serious problem, if the inverter(s) didn't shut down, the xfmr would go kablooey in layman's terms!
 

SolarPro

Senior Member
Location
Austin, TX
This article covers transformer basics in solar applications:

Distribution and Substation Transformers

In this article, we cover the fundamentals of transformer construction and operation. We discuss the general use of distribution and substation transformers in commercial and utility solar applications. We consider important transformer ratings and features. Finally, we summarize factory and field tests. If you do not take these considerations into account, the humble transformer can trip up your whole project. Our goal is to provide a definitive and practical guide that will help you select and specify pad-mounted distribution transformers for commercial and utility solar facilities.

If a transformer isn't necessary for an interconnection or a long wire run, you generally want to avoid introducing one. Adding a transformer to a PV system introduces losses. These losses are significant, in part because they incur whenever the transformer is energized. You can order a high efficiency transformer to minimize these losses, but that is an expensive device; it may also not be available off the shelf.
 

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
I asked simple question it got complicated obviously so wow have some respect

Dude, that sentence wasn't even referring to what you are doing. Respect also involves reading people's posts carefully enough to follow what they say if you're going to say something about it.

...It has also been mentioned that in general the inverters need to see all three phases coming from the utility side or they will not produce any output. That probably can be designed differently but likely presents other issues - AFAIK they are designed that way so they won't feed the utility if it is down for some reason. ...

That is correct. It's a bedrock part of UL 1741 and interactive inverter design. There are very good reasons it's not different. In addition to utility wiring not being energized in an outage (aka 'anti-islanding'), these inverters need to be configured as current sources to output all the available energy from the solar array. That generally involves using the utility generated voltage as a reference.
 
Last edited by a moderator:
This article covers transformer basics in solar applications:

Thanks for that article, SP! Great stuff.

zman-
Seriously.
One thing- is your 350 foot higher voltage run AC or DC?
Besides that- STEP UP! (See SolarPro's link)

I think I get it- you want the 66kW of PV to cover the farm's yearly usage, or so.
But you are neglecting to factor in that 66kW/kVA from the PV is 25% more than the farm's service xfmr. Really

There may be a reason for your customer to buy a grid connection xfmr. Hard to say.
You should find out what the grid voltage is AND whether you can even get 3 phase service before you do anything else.
The grid at the my farm project here is 12.47kV delta. I could get a 112.5kVA - 480/277V wye up to 12.47 delta xfmr for that, but it's REALLY expensive- instead, the POCO will bring their own and install it on the pad that the farm builds.
You might not have 12.47kV- it could be 13.2kV or 4160V delta. But whatever the grid is you want the xfmr going from your wye connected PV inverters to be wye/customer side up to grid voltage/delta/POCO side.

Side note: Like SolarPro said....*losses*!
It doesn't matter if your project is net metered or credit or someone is getting a check- stepping the 480/277V PV down to 120V (if at all possible) would lose whoever is supposed to save or make money *literally* thousands of $$ over 20 years!

Just let the POCO worry about the inverter to grid xfmr(s) (they have to approve it/them anyway...)- you should worry about supplying the 120/240V loads from your 480/277V wye inverters!
Which are outputting L1, L2 and L3 *and* N (and PE), all Ls being 277V.

A 480/277V service is the same thing- L1, L2 and L3 all 277V, and you have to combine two of them to get 480V, if that helps. Your farm doesn't have any 480V equipment, so you won't be using 480V anywhere really.

From SolarPro's link:

Commercial, industrial and utility solar applications use distribution and even substation transformers to step up the inverter output voltage for interconnection to the utility grid. This practice, in contrast with typical net-metered applications, results from the interconnection of large-scale PV projects to the utility grid at significantly higher voltage levels. Whereas residential systems typically interconnect at 240 Vac single-phase, and many small commercial systems interconnect at 480 Vac 3-phase, most solar farms interconnect to 3-phase utility distribution or transmission systems at voltages in the 12–115 kV range.

The most common 3-phase distribution configuration for transformers in solar applications is the delta-to-wye configuration, shown in Figure 3, with the wye grounded or ungrounded. This configuration allows for an independently derived neutral on the wye-connected secondary winding, which not only is essential for safety purposes but can also provide multiple voltages at the inverter pad without requiring additional transformers. The delta-connected primary winding provides a reliable configuration for the utility, because it allows harmonic currents to circulate within the transformer. This prevents harmonics from flowing into the electrical distribution system.
 
I'm not here to argue with people I just feel I am right in saying this is possible and if transformers are not outrageous it will be most cost effective method of install on this job site.

To put my last comment more shortly-
If you wanted to get a 112.5kVA xfmr, and your grid was 4160V delta, you might ask your engineer (and the POCO) about something along these lines:
http://www.temcoindustrialpower.com/products/Transformers_-_General/HT5438.html

Above 4160V, I can't find a link to something both higher efficiency and delta primary.
Call the transformer company, any one of them, and tell them you have 80A/66kVA of 480/277V PV inverter power that you want to feed into the grid at X thousand volts delta.
Make sure you are sitting down before you get the price for anything 7200V or 12.47kV delta primary...
For instance- this below isn't even high-efficiency- the previous 4160V model is only medium-high efficiency.
This one seems a bit outrageous.

http://www.temcoindustrialpower.com/products/Medium_Voltage_Transformers/T38594.html
 
Status
Not open for further replies.
Top