Ideal xfmr(s) for connecting 480/277Y inverters (60kW total) to 12470/7200 grid

[PVfarmer]

It is very doubtful that you would get 2 _separate_ 120/240V single phase services. In addition to other issues, such a setup violates code.

Could you explain that? As in, two services attached to one building violates code?
I haven't seen anything that prohibits having more than one meter on one property.

Your inverters are limited to 480/277V wye. Your choices are:
1 have the utility supply 480/277V, and then provide your own transformer to give 120/240V,
2 have the utility supply 208/120V, and then provide your own transformer to give 480/277V
3 have the utility supply 240/120V high leg delta, and then provide your own transformer to give 480/277V

All of the 480V to 208 or 240V transformers that you are looking at have _delta_ primaries. This means that each individual coil on the secondary side actually loads _two_ of the inverters on the primary side.

1 is too pricey and too complicated.
2 208 won't be used, so why not stay with 120/240 split/single phase?
3 i don't think they'll supply high leg- they supply 200A, 400A, and over 400A, which is this- Single-Phase, 3 Wire, 120/240V for
Loads 72 to 100kVA Demand or Three-Phase up to 800A From Overhead Line

So it could turn out that they don't mind all 75kVA/60kW of PV going into just one of the overhead lines, at either 120/240 OR 208/120.
Which I suppose simplifies things somewhat.

And that last 277 to 120/240 xfmr I posted was this-

Primary configuration	One Winding
Secondary configuration	Two Winding

 

[ggunn]

Could you explain that? As in, two services attached to one building violates code?
I haven't seen anything that prohibits having more than one meter on one property.

I think that what he is saying is that you cannot connect a single inverter to multiple services. I ask again: why are you so dead set on three phase inverters? A three phase inverter must be connected to a three phase service.

 

[PVfarmer]

Why are you going through all this just so you can use three phase inverters? There are perfectly good single phase inverters out there.

They're less $$ for the same amount of inverter (6 10000TL are ~$1000 more than 3 20000TLs).
Less $$ for wire.
And! They put out more overall.
Think of it like buying a BMW, except that the fancy inverter pays back for itself and then makes you $$!
Eek, I just found this one, talk about cost effective! Unless it breaks, but they do sound easy to replace, by an electrician of course.

This is a ridiculously low amount of $$!
http://www.proinso.net/pub/doc/File/ingl/sunny-tripower-STP60-us-10-proinso.pdf

Europeans are usually ahead of the curve...
http://www.solarpowerworldonline.co...-one-rise-three-phase-solar-string-inverters/

 

[jaggedben]

Could you explain that? As in, two services attached to one building violates code?
I haven't seen anything that prohibits having more than one meter on one property.

Code prohibits multiple services to a building except under certain conditions. One of those conditions happens to allow an additional service for a PV system; i.e. one service for the loads, and one for the PV, with no intermingling of connections and functions. I stated this as an option earlier in the thread.

A meter is not a service. The service is the wires that come from the utility. Mulitple meters can be connected in parallel to each service.

 

[jaggedben]

So it could turn out that they don't mind all 75kVA/60kW of PV going into just one of the overhead lines, at either 120/240 OR 208/120.
Which I suppose simplifies things somewhat.

Indeed. This is how the vast majority of systems are interconnected. It's how you'd be doing it if your existing service weren't too small.

They're less $$ for the same amount of inverter (6 10000TL are ~$1000 more than 3 20000TLs).
Less $$ for wire.
[etc.]

You should compare the costs for the different service configurations before making a final decision on the inverters. The savings you're citing may be offset by other additional costs.

 

[ggunn]

You should compare the costs for the different service configurations before making a final decision on the inverters. The savings you're citing may be offset by other additional costs.

Beat me to it. Look at the entire cost of the project rather than just a differential in one of the components.

 

[PVfarmer]

1 Code prohibits multiple services to a building except under certain conditions. One of those conditions happens to allow an additional service for a PV system; i.e. one service for the loads, and one for the PV, with no intermingling of connections and functions. I stated this as an option earlier in the thread.

2 A meter is not a service. The service is the wires that come from the utility. Mulitple meters can be connected in parallel to each service.

3 Indeed. This is how the vast majority of systems are interconnected. It's how you'd be doing it if your existing service weren't too small.
You should compare the costs for the different service configurations before making a final decision on the inverters. The savings you're citing may be offset by other additional costs.

1 I think that's where this could help: (2 25kVA xfmrs)

G2.1 (G210- ) TWO-PHASE TRANSFORMER BANK
OPEN-WYE PRIMARY
OPEN-DELTA, 4 WIRE SECONDARY
from my comment on the last page.
http://forums.mikeholt.com/showthread.php?t=170392&p=1660251#post1660251

Of course the POCO might say "um, we haven't done one of those since the 70s"
I'll have to check on that no intermingling thing in the code.

2 I meant service meter, whoops.
I haven't seen anything that prohibits having more than one (service) meter on one property.

As in, if the PV field is far away from the existing meter, and closer to a possible new connection point.

3 That's the thing- it's hard to compare costs when we have to tell them what the system is BEFORE they tell us what the service will be.
The "work" division (where you get a # for the PV application and who puts the xfmrs in) won't do or say anything about the xfmr set up until the PV application is submitted- on the PV app, they want to know "if the customer is providing a xfmr"...
So the answer is sort of...yes, but it depends on what the other dept. is going to do...it would be counter-productive to tell Dept. B one thing, then Dept. A says, no you can't do that.
For instance, the xfmr on the pole now is inside of a tree- you'd think they'd want to put it on the ground instead just for safety reasons.
OR they might want to leave it the way it is, then putting 50 kVA of that 2 phase 120/240 on the ground would be the simplest and least $$.
So now I have to check out that intermingling stuff.

Thanks!

 

[PVfarmer]

ben-
You should compare the costs for the different service configurations before making a final decision on the inverters. The savings you're citing may be offset by other additional costs.
gg-
Beat me to it. Look at the entire cost of the project rather than just a differential in one of the components.

Well, one three phase or 3 one phase xfmrs, by the prices I can find, are pushing 20 grand (which is what my neighbor the xfmr guy implied).
No idea what the POCO charges, but three high efficiency 277 to 120/240 models are under 1/2 that, and if the POCO is providing "the usual" 50 or 25 kVA 120/240 xfmrs, that's still less $$, Guessing of course
I could be missing something, but if the higher output of three phase inverters (over 240V models) can be combined with less $$ for xfmrs...why not?

 

[ggunn]

Well, one three phase or 3 one phase xfmrs, by the prices I can find, are pushing 20 grand (which is what my neighbor the xfmr guy implied).
No idea what the POCO charges, but three high efficiency 277 to 120/240 models are under 1/2 that, and if the POCO is providing "the usual" 50 or 25 kVA 120/240 xfmrs, that's still less $$, Guessing of course
I could be missing something, but if the higher output of three phase inverters (over 240V models) can be combined with less $$ for xfmrs...why not?

Because no matter what you do with transformers you cannot connect a three phase inverter to a single phase service. A three phase high leg service, maybe; it depends on the size of the B phase because sometimes the B phase is of much lower capacity than the A and C phases. I keep saying that and you keep talking about transforming 277 to 240/120. What am I missing? Sure, you could get three transformers to step 277 down to 240 or 120, but you cannot combine their output onto a single phase service because the timing of their output voltage waveforms is all wrong.

Single phase inverters can be connected to either single phase or three phase services, but three phase inverters can only be connected to three phase services.

 

[PVfarmer]

Sure, you could get three transformers to step 277 down to 240 or 120, but you cannot combine their output onto a single phase service because the timing of their output voltage waveforms is all wrong.
Single phase inverters can be connected to either single phase or three phase services, but three phase inverters can only be connected to three phase services.

Ok. But when the 3 phases of the inverters are going to 3 one phase step down xfmrs, which are going to 3 one phase step up xfmrs...they're going to *3* single phase services, not one.
Still not sure if 600A of 120/240 service would be three 25 kva xfmrs or a 25 and a 50.
But they do offer it.

 

[jaggedben]

Ok. But when the 3 phases of the inverters are going to 3 one phase step down xfmrs, which are going to 3 one phase step up xfmrs...they're going to *3* single phase services, not one.

No no no. We keep telling you this in various ways and you keep not listening. You cannot and will not get multiple services to which you connect the pv. You will also not connect the outputs of multiple transformers to each other or parallel them in any way at any location. You can, at most, get one service for the regular loads and one for the pv, which would not be connected in any meaningful way.

How many transformers are used to derive the service conductors from the grid is completely irrelevant to how you connect the inverters to those conductors.

Still not sure if 600A of 120/240 service would be three 25 kva xfmrs or a 25 and a 50.
But they do offer it.

You can be assured that any single phase service they provide will come off a single transformer. As far as you're concerned this is essentially for a three phase service too; whether the windings used happen to come in three cans or one really makes no difference in how anything is connected at your end of the service.

 

[PVfarmer]

1 No no no. We keep telling you this in various ways and you keep not listening.

2 You cannot and will not get multiple services to which you connect the pv.

3 You will also not connect the outputs of multiple transformers to each other or parallel them in any way at any location.
You can, at most, get one service for the regular loads and one for the pv, which would not be connected in any meaningful way.

4 How many transformers are used to derive the service conductors from the grid is completely irrelevant to how you connect the inverters to those conductors.

5 You can be assured that any single phase service they provide will come off a single transformer.

6 As far as you're concerned this is essentially for a three phase service too; whether the windings used happen to come in three cans or one really makes no difference in how anything is connected at your end of the service.

Thanks for hanging in here, ben.
First of all, I'm almost ready to just go with the 200A 480/277. But my neighbor the xfmr guy said the POCO likes to run the xfmrs at close to 100% capacity- the 72A of PV is only 36% of that.
Hmm. That IS max efficiency for the xfmrs. Yet another angle. And they offer 277/480 single phase service at 100A also. Arrgh.

1 I am listening, but this is seriously confusing. It would be a big help if there was a single source with all the options laid out in one line diagrams.

2 However, it could be two services with separate PV systems on each, and those PV panels could all be in the same spot, just wired to different inverter(s) for each service. Does the NEC address that anyhwere, anything about physical separation of panels/inverters as opposed to electrical separation?

3 You can connect 3 1ph xfmrs together into a delta bank. Isn't that the preferred POCO/PCC set up? Their side is Y, they keep the neutral on that side, so the customer side is a delta bank?

4 Got it. Thanks.

5 There's that weird 2 phase 120/240 thing I linked to earlier- that looks like two 25kVA models running in parallel- wouldn't that be a way to double a 200A 120/240 service?

6 Right, it doesn't affect thing electrically, just affects the $$.

So... one 12.47 to 480/277 delta to wye model to put the PV straight into the grid.
Inverters going out at 277V because the N is connected (leading out) on the LV side.
Then three 277 to 120/240 xfmrs for the load- I get what you're saying- that's the only way for the grid-interface and load balancing features of the fancy inverters to work, right?

That makes the POCO/MV side 12.47 kV delta, the LV side 480 wired in Y so 277, and the three 277 to 120/240s for the load are fed from the 277Y, letting the inverters balance it?

 

[PVfarmer]

Index G on page 209- not sure what G2.1 does, but G2.1G is single *and* 3 phase from 2 xfmrs.
Note: I'm not saying that I comprehend how these two work. But they are UDSA approved! Not sure if they're Grade A however.

If you click on them on pg 209, it jumps to the diagrams.

G2.1 (G210- ) TWO-PHASE TRANSFORMER BANK
OPEN-WYE PRIMARY
OPEN-DELTA, 4 WIRE SECONDARY

G2.1G TRANSFORMER / METER CONNECTION GUIDE
THREE-PHASE, OPEN-WYE - OPEN DELTA
FOR 120/240 VOLT POWER LOADS
http://www.rd.usda.gov/files/UEP_Bulletin_1728F-804.pdf

 

[winnie]

It looks like jaggedben has been injecting useful knowledge into this thread while I have been off-line.

PVfarmer, one of the reasons that this forum is explicitly _not_ 'DIY' is that there is no way for us to know what you don't know, and specifically there is no way for us to know the things that you are so clueless about that you don't even know to ask about. I think that this thread is being allowed to continue because 1) there is a PE involved who _will_ make sure that things are done correctly, and 2) you are asking interesting questions that are making people think outside of the box in useful ways.

In any case, there are many things that are _possible_ but which are very unlikely. Some approaches to a particular problem might be clever and elegant in a specific situation, but simply non-standard, and not used because you couldn't easily get support.

It _is_ possible to put transformers in parallel, and in theory a pair of 50kva single phase transformers could be connected to a single primary phase with secondaries in parallel, but this is unlikely. There are lots of reasons for this, but I suspect that if the power company puts multiple transformers on the pole, then they will be arranged to provide 3 phase power.

It _is_ possible to set 3 single phase 7200:120/240V transformers in a 3 phase bank, and then join the neutrals of these transformers, giving 'hexaphase' (mentioned earlier), and with custom (read expensive) transformers you could connect your three phase inverters to this non-standard power system. In addition to custom transformers, you would also need custom circuit breaker assemblies to properly protect the transformers, and it might not be possible to implement the required protection in a code compliant fashion. (This is a physical possibility, but so 'out to lunch' that it isn't worth exploring the details of the code involved.)

It is possible to take 3 phase power and convert it to single phase power; but to do so a transformer is not sufficient. You need something that stores energy from the different parts of the AC cycle and delivers it to other parts. You know, something like an inverter (or a motor-generator set, or some other systems). Any such system would totally wipe out an savings that you get from using three phase inverters rather than more expensive single phase inverters.

If you get a single phase '600A' service, it is _very_ likely that the power company will simply set a single 50kVA transformer, possibly even using the existing 25kVA transformer and simply giving you fatter wires to your service drop. The power company likes to load things as much as possible, know that transformers can take substantial overloading for considerable time, and know that a '600A' service will _never_ be used at the full 600A for more than moments at a time.

-Jon

 

[ggunn]

Ok. But when the 3 phases of the inverters are going to 3 one phase step down xfmrs, which are going to 3 one phase step up xfmrs...they're going to *3* single phase services, not one.
Still not sure if 600A of 120/240 service would be three 25 kva xfmrs or a 25 and a 50.
But they do offer it.

No. Seriously, no. You CANNOT connect a three phase inverter to three different services.

 

[winnie]

Index G on page 209- not sure what G2.1 does, but G2.1G is single *and* 3 phase from 2 xfmrs.
Note: I'm not saying that I comprehend how these two work. But they are UDSA approved! Not sure if they're Grade A however.

This is basically a single phase transformer, supplied by a single primary phase, which gives standard 120/240V, with a _second_ single phase transformer used to derive the 'third leg' giving three phase power.

The very next system G3.1 is an arrangement of 3 single phase transformers to give a wye primary and high leg delta secondary.

G3.3 is the common approach that we have been describing to you, with 3 single phase pole mounted transformers to give a wye primary and a wye secondary, either 208/120V or 480/277V three phase.

-Jon

 

[ggunn]

Thanks for hanging in here, ben.
First of all, I'm almost ready to just go with the 200A 480/277. But my neighbor the xfmr guy said the POCO likes to run the xfmrs at close to 100% capacity- the 72A of PV is only 36% of that.
Hmm. That IS max efficiency for the xfmrs. Yet another angle. And they offer 277/480 single phase service at 100A also. Arrgh.

No, they don't. 480/277 is always three phase. That's what the two numbers mean - 480V phase to phase and 277V phase to neutral. 277V X sqrt3 = 480V; that's because the THREE phases are 120 degrees apart. 240/120 is 240V phase to phase and 120V phase to neutral. 120V X 2 = 240V; that's because the TWO phases are 180 degrees apart. *

1 I am listening, but this is seriously confusing. It would be a big help if there was a single source with all the options laid out in one line diagrams.

You say you are listening but you are not hearing. This is really very simple; I said it already. Single phase inverters can be connected to either single phase or three phase services, but three phase inverters can only be connected to three phase services.**

*Yeah, I know I am saying TWO phases on a single phase service, but the reasons for that are beyond the scope of this thread.

**With the exception of what winnie posted about time shifting two of the phases so that all the voltage waveforms match a single phase service, which is theoretically possible but which would be prohibitively expensive. No one does this and there are very good reasons why.

I do not mean to sound unkind, but the reason that all this is confusing to you is that you do not understand the fundamentals of single phase and three phase power.

 

[winnie]

So... one 12.47 to 480/277 delta to wye model to put the PV straight into the grid.
Inverters going out at 277V because the N is connected (leading out) on the LV side.
Then three 277 to 120/240 xfmrs for the load- I get what you're saying- that's the only way for the grid-interface and load balancing features of the fancy inverters to work, right?

As said before, the POCO will most likely use 3 separate single phase transformers in a 3 phase bank, but how the POCO provides the service is up to them, not you. (In fact, the wikipedia article on three phase power starts with a picture of a pole mounted 3 phase transformer https://en.wikipedia.org/wiki/Three-phase_electric_power )

But the key issue that you seem to not get: on your own property you could in theory have 3 separate single phase transformers, each converting 277V (or 480V) to 120/240V... but these would be 3 separate systems, with _different_ phase angles, meaning that you could not connect them in parallel. Each of these 120/240V supplies could feed separate loads, but they could not be _combined_ to supply a large single phase load.

-Jon

 

[winnie]

Yet another angle. And they offer 277/480 single phase service at 100A also. Arrgh.

As has already been mentioned, 277/480V is three phase.

It may be that your power company offers single 277V L-N service, but I bet you find that this is for some specific use, such as street lighting or some such.

-Jon

 

[ggunn]

As said before, the POCO will most likely use 3 separate single phase transformers in a 3 phase bank, but how the POCO provides the service is up to them, not you. (In fact, the wikipedia article on three phase power starts with a picture of a pole mounted 3 phase transformer https://en.wikipedia.org/wiki/Three-phase_electric_power )

I think the larger issue is that anything on the other side of the meter/disconnect is the POCO's issue to deal with. He can get a three phase 480/277V or 208/120V service from them, to which he can connect a three phase inverter (size permitting), a high leg 240/208/120V service to which he may be able to connect a three phase inverter (again, size permitting, especially of the B phase), or a 240/120V split/single phase service to which he cannot connect a three phase inverter.

The way that the POCO builds the service is not his problem to deal with.