zig zag transformer

[electrofelon]

David & electrofelon: Would you mind giving your comments on the paper in post#45. That may clarify the matter a bit.

Sahib, that paper really isn't applicable. These inverters take a neutral and it is grounded. That's the way we do things here in the us. There are 2 options: 1. The zig zag is for harmonics. 2. The zig zag is there to satisfy a utility requirement (and possible misapplied).

I guess a 3rd option is that the designed put the zig zag in for a redundant system groundING means in case the MBJ was lifted, neutral failed, etc.... IMO that would be ridiculous.

 

[Ingenieur]

Sketch the 0 seq network diagram
the Yg:Yg xfmr makes it interesting (seems like a bad option imo)
phase shift?
both sources util & pv will supply sec gnd fault i
Looks like they are trying to give each source a rtn path?

 

[Sahib]

both sources util & pv will supply sec gnd fault
Looks like they are trying to give each source a rtn path?

The issue for the POCO is a solidly grounded PV system could make the protection coordination ineffective by its contribution to fault current. So the POCO (in US, of course)usually demands an effective grounding. In some cases, however, POCO may allow solid grounding (as in OP case with solidly grounded zigzag transformer) if the PV system trips fast in case of POCO distribution line fault.

 

[mbrooke]

Was the MV-LV transformer delta at any point in time?

 

[mbrooke]

1200A at 13.2KV ??? That can't be right.

(Is this when an RFI becomes a WTF?)

1,200 is a very common standard for 15-35kv breakers.

 

[GoldDigger]

It is still a lot of watts (or maybe just VA).

Sent from my XT1585 using Tapatalk

 

[Greghollis]

The drawing says the “purpose” is for grounding .


Sent from my iPhone using Tapatalk

 

[mbrooke]

It is still a lot of watts (or maybe just VA).

Sent from my XT1585 using Tapatalk

Yup, though this post may be on to something:

http://forums.mikeholt.com/showthread.php?t=186921&page=3&p=1861899#post1861899


I was not aware of the customer generation- but that could be the reason considering it radically changes what typical fuses, reclosers and feeder breakers see during fault conditions (usually reduced fault current from the in-feed). The zig-zag might be there just for that reason, to increase short circuit currents as seen by protective devices, but if that was the case it would make much more sense on the MV side.

Heres my issue with this: there is nothing that indicates selective coordination or what currents will pass through the zig-zag during fault conditions. Short circuits out on the utility feeder could trip the 1,200amp zig-zag, 3000 amp fused main, or devices on the 13.2 kv side. Unless every detail is known that zig-zag could be major headaches.


This is one of those things that could be either a very brilliant solution to a unique problem, or very bad idea for a none issue depending on what the engineer(s) are trying to accomplish.

Electrofellon: Please keep us updated.

 

[Ingenieur]

looks live the pv sources are delta?
the legend/note shows 4w + G, dwg shows 3 plus egc

remove the zig-zag
3000 fs opens
a single phase from the pv side of the switch (not the xfmr/util side) goes to gnd
no fault rtn path, frame is elevated
if the zig-zig is in the ckt a single gnd fault will induce a ph-ph fault thru the zig-zag

if the fs is closed (or fuses good) the ph gnd fault has 2 paths
the zig-zag ph-ph
the xfmr X0 thru a winding to a ph

 

[mbrooke]

FWIW, I have worked on a number of ther 1-2 meg systems in other utilities' territory, and the interconnect was solely throught a Wye:Wye transformer.

Thanks for the solectria reference, I will look into it.

Wye-Wye is the utilities choice 99% of the time. In fact some POCOs forbid delta on any distribution transformer. Wye-Wye completes the circuit, has no phase shift, far less susceptible to ferroresonance, and inductive tank heating is solved simply with a "give me a 5 limb instead of a 3 limb"

National Grid generally requires any PV plant >500kW, and sometimes smaller ones, to have what they call "effective grounding". This can take the form of either a grounding transformer/Zig-Zag transformer or a Neutral Ground Reactor. I believe both would fall under the general category of impedance grounding. I generally see Zig-Zag transformers used with Wye-G:Wye-G and NGRs used with Wye-Gelta, although I'm not sure of the exact reasons behind that.

I'm confused. Reactors and NGRs limit zero sequence current, while wye grounded-delta and zig-zag provide a zero sequence source. Yg-delta and zig-zag would not be effective (imedpance) grounding in my book unless the N terminal to ground of the zig-zag or the secondary delta (brocken delta, acorss the terminals) had a reactor or resistor across it.

The best way I understand it is the utility is concerned with the impact of the PV system during a nearby ground fault on the transmission system. With a solidly grounded Wye-Wye system the fault current has a return path through the PV system secondary, which could desensitize the utility ground fault detection and cause high fault currents to flow. It makes sense to me that adding impedance between the PV system neutral and ground could reduce these fault currents.

Transmission to distribution, or sub transmission to distribution transformers are almost always delta-wye (wye ungrounded-delta-wye grounded or delta zig-zag when 0* phase shift is needed). Faults on those systems would have no effect on the PV system or the ground step distance protection of the tranmission system. In fact one major reason behind the delta wye selction is so ground faults on the distrbution system do not pass to the transmission system.

As for increased fault current on the 13.8kv system, a grounding transfomer would make it far worse, a grounding transfomer with an impedance would have no effect in reducing fault current.

Exactly correct here. For even more interesting information on the subject you can read IEEE 142 the Green Book on grounding and IEEE 1547 interconnection requirements for PV systems. Effective grounding is a method used for decades on rotating generators, it's questionable that it does anything for static inverters used in PV systems but utilities are used to requiring them on generators so they thought, "why not just keep going with that."

Effective grounding usually calls for the neutral to be grounded through an impedance and not solidly, so that's a little unusual here. I would not be surprised if this effective grounding system was not improperly designed for the application.

Effective grounding is supposed to limit ground fault overvoltage and load rejection overvoltage after the local substation isolates a distribution line with PV attached and the PV is slow to disconnect. The alternative is direct transfer trip where the local substation directly commands the PV systems to disconnect at the same time it isolates the line. That is usually much more expensive so people were happy to have the effective grounding option.

The remote trip on the zig-zag breaker is to make sure the PV system is disconnected from the distribution system if it has a fault or if it's intentionally disconnected, but usually that only trips the PV system interconnection and not the whole site.

Can you elborate more on how this is typically done? And how the equipment is sized? The way I see it is that zig zag will either prvide a zero sequnce source or not- unless as you mentiuon hjas the ability to limit certain transients.

Ok this is starting to make a littler more sense. It sounds like the transformer was supposed to be delta on the PV side and the system grounding comes from the zig zag. The designer misunderstood the intent of the utility requirement. Do others agree with this hypothesis?

That could be it- and thats what I would have done if I wanted a major zero sequence source out on the feeder. But the question lays as to why if that was the case.

Sketch the 0 seq network diagram
the Yg:Yg xfmr makes it interesting (seems like a bad option imo)
phase shift?
both sources util & pv will supply sec gnd fault i
Looks like they are trying to give each source a rtn path?

Honestly, unless a see A LOT of detailed numbers, and some proof based reasoning, bad option would be correct.

 

[mbrooke]

looks live the pv sources are delta?
the legend/note shows 4w + G, dwg shows 3 plus egc

remove the zig-zag
3000 fs opens
a single phase from the pv side of the switch (not the xfmr/util side) goes to gnd
no fault rtn path, frame is elevated
if the zig-zig is in the ckt a single gnd fault will induce a ph-ph fault thru the zig-zag

if the fs is closed (or fuses good) the ph gnd fault has 2 paths
the zig-zag ph-ph
the xfmr X0 thru a winding to a ph

All true- and good analysis- however in that case the POCO must change to a delta-wye transformer.

 

[jaggedben]

looks live the pv sources are delta?
the legend/note shows 4w + G, dwg shows 3 plus egc

I believe the PV sources are wye, and the drawing shows 4W, with the neutrals not detailed inside the switchboard, and EGCs not shown.

The drawing does not seem to show a neutral sent to the zig-zag.

 

[Ingenieur]

I believe the PV sources are wye, and the drawing shows 4W, with the neutrals not detailed inside the switchboard, and EGCs not shown.

The drawing does not seem to show a neutral sent to the zig-zag.

dwg shows 3 ph plus egc tied to the frame (maybe the neut?)
the note shows 3 ph, n and g (same for the 3000 A feeder, choice of Cu or Al)
if it had a neut you would think it would be connected to the xfmr X0?
maybe they just did not show it (or the egc's)

the zig-zag derives a neut, actually a ground

still the same y or d, if the switch is open no gnd fault path

 

[mbrooke]

dwg shows 3 ph plus egc tied to the frame (maybe the neut?)
the note shows 3 ph, n and g (same for the 3000 A feeder, choice of Cu or Al)
if it had a neut you would think it would be connected to the xfmr X0?
maybe they just did not show it (or the egc's)

the zig-zag derives a neut, actually a ground

still the same y or d, if the switch is open no gnd fault path

Ok- correct me if I am wrong.

Won't that zig-zag transformer try to balance out any phase angle discrepancies from the utility? Such that a fault on the 13.2kv primary could cause the breaker for the zig-zag to trip?

 

[jaggedben]

dwg shows 3 ph plus egc tied to the frame (maybe the neut?)
...

The drawing shows the fourth wire connected to each winding of the transformer opposite each phase, and also to ground. This to me pretty clearly indicates that the fourth wire in the drawing is the neutral. Transformer note also says wye for both sides.

 

[electrofelon]

So I dont really have any significant updates with this, other than it is still a real mess and no one really seems to know what is going on. First, let me disclaim that I am not the head of this project so I dont talk to the utility, engineer, or solar developer directly - But this is what I hear from the guy I am working for.

There is a MV contractor that is separate from us, and he didnt install any of the control wiring for the shunt trip to the MV side. I have no idea if he messed up, or still has to do it, or someone decided it wasnt needed. The utility is saying they dont want a breaker feeding the zigzag. Its like they want that thing on at all costs, like a fire pump LOL. Its kinda funny, I guess they dont understand that we have codes to follow and cant just skip transformer and conductor protection.

Post #25 is right on though. This is definitely a utility requirement, not something the designer or engineer came up with because of some real or perceived problem. We have done very similar systems for other utilities and there was no zig zag, no shunt trip, just a straightforward connection to a Wye -Wye padmount.

I know the drawing is a little vague on the grounded conductor and EGC details, but it definitely has a grounded conductor to the switchboard and all the feeders.

 

[mbrooke]

If you can, I would try and make a logical case to the utility.

Just to clarify, the PV inverter has its own neutral?

 

[electrofelon]

Just to clarify, the PV inverter has its own neutral?

Yes, but it is only used for voltage and phase measurement, but there is a grounded neutral circuit conductor

 

[Sahib]

The question is what arrangement exists on PV side for a phase to ground fault not to compromise the protective coordination on POCO side, if any.

 

[JackyJohnson]

The solectria reference has non-approved IEEE P1547.8 reference

The solectria reference has non-approved IEEE P1547.8 reference

The entire thing came from IEEE P1547.8 standard which is not approved and will never be approved with the specific grounding impedance requirement. I had the similar situation where utility need specific Zig-Zag grounding transformer with specific impedance and that might be coming from the non-approved IEEE standard 1547.8. See below link from the utility:

https://www.xcelenergy.com/staticfi...cing-Requirements-and-Sample-Calculations.pdf

Is it the best practice ? Based on IEEE not because it was based on a non-approved standard and the assumption made were incorrect for X0 or impedance requirement. There are new standards which are C62.92.1 to C62.92.6 which are for utility interconnection and based on that you may not need the Zig-Zag grounding transformer. Again, not sure but refer those standards. Let me know if this helps.

National Grid generally requires any PV plant >500kW, and sometimes smaller ones, to have what they call "effective grounding". This can take the form of either a grounding transformer/Zig-Zag transformer or a Neutral Ground Reactor. I believe both would fall under the general category of impedance grounding. I generally see Zig-Zag transformers used with Wye-G:Wye-G and NGRs used with Wye-Gelta, although I'm not sure of the exact reasons behind that.

The best way I understand it is the utility is concerned with the impact of the PV system during a nearby ground fault on the transmission system. With a solidly grounded Wye-Wye system the fault current has a return path through the PV system secondary, which could desensitize the utility ground fault detection and cause high fault currents to flow. It makes sense to me that adding impedance between the PV system neutral and ground could reduce these fault currents.

I believe there are also concerns with temporary overvoltage under some of these fault scenarios, which NGRs/Zig-Zags can mitigate. I'm less clear on how this works.

I admit I don't truly understand some of the fundamentals here, this is just what I've picked up in dealing with NGRID. It also often appears that the utilities themselves don't understand what they mean by "effective grounding". We just suck it up and buy a zig-zag to make them happy.

Solectria has published some good documentation on this issue that may be helpful.