240 delta with high leg solar installation

[PVfarmer]

B is 208V to the grounded neutral. It is also 240V to A and 240V to C. That's all the "PV" transformer needs is A, B, and C at 240V to each other.

I get that too- but it seems as if the PV xfmr is connected to a service panel which is not wired in the way it needs.

If it was supply side connection and the 240 delta/PV output wasn't going thru the bus in the MSP...the neutral/ground SSBJ/GEC would still have to be connected in the MSP, correct?

That last sentence is exactly backwards. The neutral for the PV side SDS must end at the PV side of the xfmr. Any further connection from there is considered an SSBJ or a GEC and is not supposed to carry any current.

The fact is that the installation manual for the OP's inverter says neutral is required. It can't end on the PV side of the PV xfmr.

"Required to have a grid connection"

So not backwards really.

The PV neutral is connected along with the (incoming) POCO xfmr center-tapped neutral at the MSP ground/neutral bar, and therefore the same GES, correct?

The name "neutral" ends on the PV side, the electrical connection does not.

The important thing it that *some inverters* need a grid neutral connection to work, some don't- the OP's does.

 

[ggunn]

The fact is that the installation manual for the OP's inverter says neutral is required. It can't end on the PV side of the PV xfmr.

Not so. The isolation transformer creates a neutral on the 480/277V side, which is grounded. The neutral is NOT carried all the way back to the service from the 480/277V side of the transformer and it wouldn't even if the transformer had a neutral on both sides. It's separately derived. Look, I do this stuff all the time and I know for a fact that it works and that the inverter companies are OK with it.

 

[ggunn]

The PV neutral is connected along with the (incoming) POCO xfmr center-tapped neutral at the MSP ground/neutral bar, and therefore the same GES, correct?

The name "neutral" ends on the PV side, the electrical connection does not.

The important thing it that *some inverters* need a grid neutral connection to work, some don't- the OP's does.

No, NOT correct. The PV neutral has NO connection to the service neutral other than through the ground. The voltages on both sides of the transformers are referenced to it as 0V but that is all. No current flows. Whether the inverter needs a neutral or not to function is irrelevant; the neutral is there if it is needed.

You might want to spend some time learning how all this stuff works rather than just cutting and pasting irrelevant passages from a manual.

 

[PVfarmer]

Whether the inverter needs a neutral or not to function is irrelevant; the neutral is there if it is needed.

You might want to spend some time learning how all this stuff works rather than just cutting and pasting irrelevant passages from a manual.

Why would I call SOlarEdge to answer the OP's question?
Where do you get this idea that ignoring the actual manual is OK because you feel like it?

Fact:
Some inverters will work with a high leg grid, and some won't.

This italicized is your comment from Aug. 28 2013- you were correct then- SMA xfmrless require a neutral.
So does the OP's SE 33.3KW inverter.
So tell me simply- where is the neutral for the OP's SE 33.3kW inverter needs connected?

The standard SMA Sunny Boys can operate without a neutral reference; page 24 of their SB5000-6000-7000-8000 manual shows a three inverter interconnection to 240 delta with a stinger. One inverter has a neutral and the other two do not.
I believe the transformerless models all require a neutral but I am not sure. The new three phase SB's are 480VAC output only.
http://forums.mikeholt.com/showthread.php?t=155814&p=1507816#post1507816

 

[PVfarmer]

Look, I do this stuff all the time and I know for a fact that it works and that the inverter companies are OK with it.

You're from TX and the OP is from CA.
So your comment is kinda irrelevant, no?
Different POCOs?
It's the OP's POCO's high-leg service we are talking about- the inverter maker doesn't tell his POCO what is OK to connect to their grid.

NatGrid in RI would say no for sure, I have zero doubt about that.

 

[ggunn]

Fact:
Some inverters will work with a high leg grid, and some won't.

This italicized is your comment from Aug. 28 2013- you were correct then- SMA xfmrless require a neutral.
So does the OP's SE 33.3KW inverter.
So tell me simply- where is the neutral the SE 33.3kW inverter needs connected?

The standard SMA Sunny Boys can operate without a neutral reference; page 24 of their SB5000-6000-7000-8000 manual shows a three inverter interconnection to 240 delta with a stinger. One inverter has a neutral and the other two do not.
I believe the transformerless models all require a neutral but I am not sure. The new three phase SB's are 480VAC output only.
http://forums.mikeholt.com/showthread.php?t=155814&p=1507816#post1507816

Heavy sigh. The neutral the inverter needs is the neutral on the 480/277V wye isolation transformer. Get this through your head: it's separately derived; the asymmetry of the neutral with respect to the phase voltages does not pass through the transformer. The inverter only cares about what it is directly connected to; it does not know or care that its path to the utility passes through a high leg stage.

You cannot connect a three phase inverter that requires a neutral directly to a high leg service because the voltages to neutral are not the same for all the phases - that much you have right. But if you create a wye from that high leg service with an isolation transformer, then the phase to neutral voltages ARE all the same and the inverter can and will connect. Again: I have done this many times and my systems are all up and running. It's silly to keep insisting that this can't work.

What you quoted me as saying in 2013 doesn't have anything to do with three phase inverters. I'm flattered that you would research my comments that deeply, though.

 

[wwhitney]

The important thing it that *some inverters* need a grid neutral connection to work, some don't- the OP's does.

If an inverter manual says it needs a "grid" neutral connection, it is just using slang to mean that the voltage system the inverter is connected must have a neutral which is brought to the inverter. "Grid" here means "other power source" for the grid-tie inverter , not necessarily the utility. That other power source can be a customer-owned transformer; in fact, there need not be any utility power present.

Cheers, Wayne

 

[PVfarmer]

I'm not here to argue at all.
This thread is a bit off track, sure.
I agree with ggunn- if there are 3 xfmrs for this high-leg service, and they aren't all the same size, a new service is needed- that is no big deal.
I'd bet the POCO would like to update a high-leg service, it is probably 40+ years old.

Here's a problem I see- the PV disco switch is grounded....where? Seems neutral is (maybe?) bonded in the PV switch and grounded in the MSP- that may be the problem. The 2nd bolded part- the "GEC" goes THROUGH the PV disconnect- what does that mean?

There is a neutral TO the MSP and no neutral between MSP and PV xfmr- that is the purportedly "missing" neutral.
Are there 208V loads associated with the MSP? Then it has 208V and 240V on bus B, corrrect?

Here is everything the OP has said (we lost him on page 2- hope you are still there, dude).

Need Help! I have an installation were the main service is a 240 delta with a high leg. We are using a SolarEdge 33k 480v inverter. We have installed a 240 delta to 480/277y transformer in between. Everything was stamped by engineer and passed city plan check, but now the city inspector is saying it won't pass because if there was a fault to ground on the primary side of the transformer it would not trip the breaker. Please advise, I am looking for a solution that would not cause us to have to use 3 single phase SolarEdge inverters and have to change out all of the p700 optimizers we currently have onsite. There are 90 Canadian Solar 315watt panels. Thank you in advance for your help

I guess after talking to the Inspector more I need to make a correction to my post. On the service side of the xfmr, there is no neutral connection. There is one on the inverter side, but not service side. The GEC is tied continuously from the array, through the inverter, through the disconnect,through the xfmr which is also tied to the neutral on the inverter side of xfmr, through the disconnect, and tied the the ground/neutral bar at the service. I'm including a photo of the label on the transformer.

The xfmr is 15 ft from the service OCPD. If this is still a problem, would you happen to know of a sollution to it. I am being told that a zig zag xfmr could create the neutral I am missing.

Also need to mention that there is a fused disconnect right next to the xfmr on the service side in addition to the breaker that feeds it.

Yes there is a neutral conductor from the POCOs xrmr(s) to the MSP. Also in the event of a fault the inverter will shut down within 10seconds.

 

[PVfarmer]

But if you create a wye from that high leg service with an isolation transformer, then the phase to neutral voltages ARE all the same and the inverter can and will connect.

I'm flattered

The OP's setup is creating the Y from a high-leg service and MSP of which one leg is 208V. The xfmr creating it COULD BE high-leg, but is not.
Why not?

And don't be flattered- it was the 2nd result in google for '3 phase inverter high leg service'.

That other power source can be a customer-owned transformer;in fact, there need not be any utility power present.

Howdy Wayne- SMA and/or SE inverters won't work without grid power.
I see the power source as the POCO's xfmr, it is center-tapped.
Bus B in the MSP is 208V. How can it be both 208V and 240V?

Say this OP's customer's grid goes down and the customer plugs a 120/240V generator into the MSP.
That's not a 3-phase grid or power source, and the 3-ph inverters won't run.
Neither will a 208V motor on bus B- what happens with 240V motors running off of B-C or A-B and not A-C? (A-C being the center-tapped 120/240V)

 

[wwhitney]

Howdy Wayne- SMA and/or SE inverters won't work without grid power.

If by "grid" you mean "utility" that is not true, which was the point of my comment, "grid" doesn't have to be utility. Any grid-tie inverter just needs an external stable AC source to sync to. Usually that is a utility, but it could be a good generator or a battery inverter. [If it is not the utility, then other means have to be taken to ensure that instantaneous power production matches instantaneous demand, including the case that the availabe PV output exceeds demand.]

I see the power source as the POCO's xfmr, it is center-tapped.
Bus B in the MSP is 208V. How can it be both 208V and 240V?

You need to understand that voltage is measured between two points. If one says "point X is at 240V", there is always an implicit point that is defined as 0V, which needs to be determined from the context.

So Bus B to ground is 208V. Bus B to Bus A is 240V. Bus B to Bus C is 240 V.

Cheers, Wayne

 

[ggunn]

The OP's setup is creating the Y from a high-leg service and MSP of which one leg is 208V. The xfmr creating it COULD BE high-leg, but is not.
Why not?

It could have a high leg connection to the service but there's no reason to; all it needs are the phase conductors. There are 240/120V high leg to 480/277V wye transformers and there are 240V delta to 480/277V wye transformers. Either would work.

The delta to wye transformer takes the phase to phase voltages on the service side (forget about the neutral on the service side because it isn't even connected to the transformer, and forget about the high leg stuff because it is irrelevant) and changes them to what the inverter needs, and it creates the center point between the phase voltages for the newly created neutral. Until you ground the neutral, all four of these voltages relative to ground are undefined.

You can bond neutral to ground either in the disco on the inverter side of the transformer or at the transformer itself, but not both. You'll need a new ground rod and you'll need to bond it to the service ground.

 

[Smart $]

I get that too- but it seems as if the PV xfmr is connected to a service panel which is not wired in the way it needs.

If it was supply side connection and the 240 delta/PV output wasn't going thru the bus in the MSP...the neutral/ground SSBJ/GEC would still have to be connected in the MSP, correct?
...

I do not know how the service panel is wired, and I do not believe the OPer elaborated on that. For the sake of this discussion, all that matters is that the service is a grounded system and assume all other pertinent aspects are compliant.

Nothing changes as to how the MSP is wired regardless of whether the PV system is connected supply side or load side. The service neutral must be run to the MSP. The GEC(s) for the building or structure where the service disconnect is located must connect to the service entrance conductor neutral. And a Main Bonding Jumper is required in the service disconnecting means. That's it for the MSP.

As to a supply-side PV system disconnecting means, I'd rather not get into that because no one can agree on exactly how it is supposed to be done. Suffice it to say you either have to run a service-entrance neutral to it and add a neutral to ground bond, or run a supply-side bonding jumper from any of several permitted connection points.

 

[Smart $]

...
Howdy Wayne- SMA and/or SE inverters won't work without grid power.
I see the power source as the POCO's xfmr, it is center-tapped.
Bus B in the MSP is 208V. How can it be both 208V and 240V?

Say this OP's customer's grid goes down and the customer plugs a 120/240V generator into the MSP.
That's not a 3-phase grid or power source, and the 3-ph inverters won't run.
Neither will a 208V motor on bus B- what happens with 240V motors running off of B-C or A-B and not A-C? (A-C being the center-tapped 120/240V)

Bus B is 208V to neutral. This voltage is only stated for technical reference. It is never used in this capacity, i.e. for powering 208V single phase loads.

Bus B to Bus A is 240V.
Bus B to Bus C is 240V.
Bus A to Bus C is 240V.

Bus B is 208V to neutral at the same time because the line to line voltages are 120° out of phase with each other. Bus B to neutral is actually 90° out of phase with Bus A to Bus C voltage.

The customer should not connect a 120/240V 1Ø 3W generator to a 240/120V 3Ø 4W (delta high leg) system. A subsystem of A-N-C loads can be set up, and a 120/240V 1Ø 3W generator can be connected to this subsystem.

If the customers grid goes down, the local system must be setup as a microgrid, i.e intentional islanding. This requires a controller that makes certain the local system is disconnected from the grid among other issues with a PV system and generator working together.

 

[jaggedben]

I'm not here to argue at all.

Are you sure? :lol::lol::lol::lol::lol:

If by "grid" you mean "utility" that is not true, which was the point of my comment, "grid" doesn't have to be utility. ...

Right. I said the same thing in a slightly different way many posts ago. ('Grid' in the manual means 'AC system'.)

PVFarmer, try this thought on for size...
Functionally, what is the difference between these two hookups?

1) ---[service point] ----3 phase system of unknown grounding and voltage ---[XFMR approved for both sides] --- 480/277V Wye --- [PV inverter designed for 480/277V Wye]

2) ----3 phase system of unknown grounding and voltage ---[XFMR approved for both sides] --- 480/277V Wye ---service point --- [PV inverter designed for 480/277V Wye]

Remember that the service point is just a legal boundary, not an electrical one. Also note that 'Grid' isn't an NEC term, and could be on either side of the service point.

 

[PVfarmer]

Any grid-tie inverter just needs an external stable AC source to sync to.

The step-down PV xfmr isn't a source- if there was no service of any kind, the inverters wouldn't run.
I don't know if I'd call the grid a source- isn't it more of a...repository?
The inverters need to know the impedance of the grid.
This is sorta over my head, but I do know the grid people are the only ones who know what their codes and standards are.

The LCL filters employed in grid connected inverters are
used to pass the fundamental frequency and attenuate the rest
of undesired high order harmonics, which are appeared in the
current and voltage profiles. Although, this goal is more or
less achieved as an advantage of LCL filters, they interact with
the grid elements as a part of a resonance circuit at a point
close to the source of harmonics, i.e. inverter [13]. Harmonic
emission of inverter or the harmonics, which are coming
form the grid side may trigger existing parallel and/or series
resonance circuits. This phenomenon may magnify harmonic
level exceeding the limits recommended by grid codes and
standards leading to distorted and hence undesired waveforms
or in severe cases instability of inverter
https://arxiv.org/pdf/1612.02045.pdf

Until you ground the neutral, all four of these voltages relative to ground are undefined.

You can bond neutral to ground either in the disco on the inverter side of the transformer or at the transformer itself, but not both.

Seems like the OP described bonding it in the MSP, which would not be one or both- it'd be neither.

Nothing changes as to how the MSP is wired regardless of whether the PV system is connected supply side or load side. The service neutral must be run to the MSP. The GEC(s) for the building or structure where the service disconnect is located must connect to the service entrance conductor neutral. And a Main Bonding Jumper is required in the service disconnecting means. That's it for the MSP.

The OP described the PV neutral as ending at the PV disco, but also continuing on as a GEC thru the xfmr to the MSP.

There should be a bonding jumper in the PV disco *and* the MSP, and both ground bars (MSP/PV disco) should be connected to the GES by GECs.
That isn't how it was described.
I don't even understand how the "GEC goes thru the PV xfmr" as described...

Bus B to Bus A is 240V.
Bus B to Bus C is 240V.
Bus A to Bus C is 240V.

Bus A-N is 120V, and bus C-N is 120V.
I'm assuming that they are being used. Won't that throw off the inverter's reference somehow?

Are you sure? :lol::lol::lol::lol::lol:

PVFarmer, try this thought on for size...
Functionally, what is the difference between these two hookups?

1) ---[service point] ----3 phase system of unknown grounding and voltage ---[XFMR approved for both sides] --- 480/277V Wye --- [PV inverter designed for 480/277V Wye]

2) ----3 phase system of unknown grounding and voltage ---[XFMR approved for both sides] --- 480/277V Wye ---service point --- [PV inverter designed for 480/277V Wye]

Remember that the service point is just a legal boundary, not an electrical one. Also note that 'Grid' isn't an NEC term, and could be on either side of the service point.

Yes!
However....
The service point is what the inverter is supposed to use for grid reference, no?
One difference- in #1 the xfmr is customer owned, in #2, the xfmr is POCO owned.

In #1, the POCO's interconnection calcs are a factor which is influenced by the customer owned xfmr. Period.

1- You don't want to use a wye connection which is grounded on PV side and wye/not grounded on "grid" side if using an SMA Tripower, so a "3 phase system of unknown grounding and voltage" might work, or not.
SMA only restrictively recommends "Yyn_" transformers with a star connection on the medium-voltage side and a star
connection on the low-voltage side with a neutral point that leads outward (see figure below e.g. Yyn0)
http://files.sma.de/dl/7418/STP24-US_MV_Trafo-TI-en-10.pdf

Also, 1 doesn't make sense- how can you have a mystery service? The POCO doesn't say "here's some electricity, figure it out yourself. Good luck."

2- You don't need to know what the grid voltage is if the service and inverter are both = to 480Y/277V.

Service Point. The point of connection between the facilities
of the serving utility and the premises wiring.
Informational Note: The service point can be described as
the point of demarcation between where the serving utility
ends and the premises wiring begins. The serving utility
generally specifies the location of the service point based
on the conditions of service.

I mean, really- you can't put in an xfmr of which you don't know the voltage/grounding! I don't get this one.
??
1) ---[service point] ----3 phase system of unknown grounding and voltage ---[XFMR approved for both sides] --- 480/277V Wye --- [PV inverter designed for 480/277V Wye]

 

[wwhitney]

The step-down PV xfmr isn't a source- if there was no service of any kind, the inverters wouldn't run.

For purposes of electrical theory it is a source. It relies on electrical power being present on "the other side", but any electrical source requires a source of power--fossil fuel for a generator, batteries for a battery-based inverter, AC electrical power for a transformer, DC electrical power for an inverter, etc.

The service point is what the inverter is supposed to use for grid reference, no?

No, the inverter just references the voltage system present on the wires connected to it. That will be determined by the closest upstream transformer. Everything on "the other side" of that transformer is immaterial.

Please listen to what multiple people are telling you.

Cheers, Wayne

 

[PVfarmer]

Remember that the service point is just a legal boundary,

Exactly!
That's why you can't go and call the 480Y/277V PV side of a step down xfmr a service point.

I'd say *some* inverters, including the OP's need the *supply point* for reference, as in where the service neutral originates.

You certainly can't connect a three phase inverter to 120/240V single phase service.
High leg from A-C legs is the same as 120/240V single phase.

Ok, forget high leg- if it was 240 delta service with no neutral, there would be no grid neutral connection for the inverter...how is that not a problem?

Also, I honestly don't see how "grid" doesn't mean the utility owned grid.
The words are in the same sentence.

Page 36-
When connecting multiple single phase inverters in
an installation connected to a three-phase grid, phase balancing may be
required by the utility or grid operator.
https://www.wholesalesolar.com/cms/solaredge-se14400us-208-installation-manual-3800765590.pdf

 

[PVfarmer]

No, the inverter just references the voltage system present on the wires connected to it.

This is the description we have in italics.
L1,L2 and L3 from inverter are connected to the step-down xfmr terminals X1, X2, X3.

N from the inverter is also connected to the XO of the xfmr, and the inverter N is connected to the MSP by a...SBJ, not a GEC. (I think)

That's 2 voltage systems present on the wires connected to the inverter.

If the service was 208Y/120V, the PV disco and MSP would have GECs to the GES.
But with 240 delta, the neutral is run between the disco and MSP?
Why?


The GEC is tied continuously from the array, through the inverter, through the disconnect,through the xfmr which is also tied to the neutral on the inverter side of xfmr, through the disconnect, and tied the the ground/neutral bar at the service.

The GEC is connected to a bonded PV disco *and* a bonded MSP? Doesn't seem right to me.

 

[wwhitney]

L1,L2 and L3 from inverter are connected to the step-down xfmr terminals X1, X2, X3.

N from the inverter is also connected to the XO of the xfmr, and the inverter N is connected to the MSP by a...SBJ, not a GEC. (I think)

That's 2 voltage systems present on the wires connected to the inverter.

No. X1/X2/X3/X0 on the inverter side of the transformer is one voltage system. If the earthing conductor (the GEC or SBJ or whatever the proper name is) run to the GES (grounding electrode system) were not present, then that voltage system would be floating relative to ground. To make that voltage system a grounded system, so that each conductor in it has a fixed voltage to ground, an earthing conductor is run from X0 to the GES. The fact that other voltage systems are also grounded and connected to the GES doesn't make them present at the inverter.

If the service was 208Y/120V, the PV disco and MSP would have GECs to the GES.
But with 240 delta, the neutral is run between the disco and MSP?
Why?

No, the conductor that is run is still an earthing conductor, it is not a neutral circuit conductor. The fact the service is 240V hi leg delta, rather than a wye system, is irrelevant. The service could be an ungrounded 240V delta service (if such a thing exists), the exact same set up would work, there just wouldn't be a neutral bar in the main service panel. There would still be a GES, a ground bar in the main service panel, and an earthing conductor running to the inverter side of the transformer, to make that voltage system a grounded system.

Cheers, Wayne

 

[wwhitney]

I'd say *some* inverters, including the OP's need the *supply point* for reference, as in where the service neutral originates.

Nope, a grid tied inverter doesn't need the *service* neutral. It may or may not need a neutral conductor for the voltage system that it connects to, the ones under discussion do need such a neutral conductor.

As a practical matter, most voltage systems are grounded, and if a grounded system has a neutral conductor, it is the neutral conductor which is grounded, so there will be an indirect connection to the service neutral.

Cheers, Wayne