Zigzag Transformers for Islanding TOV on PV Projects

[cyriousn]

I'm trying to understand why grounding transformers are put on solar projects where the transformer is Yg-Yg. Is it due to the condition where the ground source is lost if the breaker at the local substation is tripped and the PV site is islanded with the distribution circuit until the PV site shuts down in 2 seconds or less due to islanding detection?

I'm working on a project where there is a zigzag transformer connected on the 600V side of the 23kv/600v Yg-Yg Transformer. The zigzag is fed from a 400A electronic breaker in a switchboard that shunt trips the 2500A MCB. Typically I see Yg - Delta grounding transformers on solar projects so this is a new one for me.

Wouldn't the grounding transformer, under steady state conditions, be a grounding source for the local distribution circuit for imbalances? The rating of the transformer is 867A continuous on the neutral. Is there an easy way to determine what conditions on the distribution circuit could trip the circuit breaker and dump the project?

 

[pv_n00b]

The effective grounding XFMR is not used under normal steady state operation, it is used when there is a L-G fault on the utility distribution system that causes it to be isolated at the substation. This removes the ground reference for the distribution system in the substation and now you have a PV system feeding an ungrounded distribution system with a L-G fault until the inverter realizes something is wrong and shuts down. During that period it could cause a TOV event on the distribution system that may damage other customer's electrical systems. The effective grounding transformer will limit the TOV voltage and prevent that damage. That's the short explanation.

 

[pv_n00b]

Wouldn't the grounding transformer, under steady state conditions, be a grounding source for the local distribution circuit for imbalances? The rating of the transformer is 867A continuous on the neutral. Is there an easy way to determine what conditions on the distribution circuit could trip the circuit breaker and dump the project?

This is true and why it is an effective grounding transformer (EGT) and not a grounding transformer. EGTs are not designed to provide a solid ground like grounding transformers are. They are designed with specific impedances to limit TOV and at the same time limit ground fault current through them so the utility protection system is not desensitized.
The CB protecting the EGT is linked to the main CB so that if the EGT is isolated the PV system cannot back feed the utility distribution system. The CB for the EGT should not trip unless there is a fault on the EGT circuit. Sometimes the neutral current of the EGT is monitored by a relay and if ground fault current over a set point is detected the PV system main CB is opened to take the system offline.

 

[cyriousn]

This might be a stupid question but could the 400A breaker rating be too high for protecting the grounding transformer? Is there a condition where the 867A continuous neutral rating is violated from having more than 289A on each phase? We could dial down the breaker rating but I didn't know if we would be looking at any potential nuisance tripping of the solar site.

 

[ggunn]

This might be a stupid question but could the 400A breaker rating be too high for protecting the grounding transformer? Is there a condition where the 867A continuous neutral rating is violated from having more than 289A on each phase? We could dial down the breaker rating but I didn't know if we would be looking at any potential nuisance tripping of the solar site.

Phase currents do not add to make line current.

 

[pv_n00b]

This might be a stupid question but could the 400A breaker rating be too high for protecting the grounding transformer? Is there a condition where the 867A continuous neutral rating is violated from having more than 289A on each phase? We could dial down the breaker rating but I didn't know if we would be looking at any potential nuisance tripping of the solar site.

You should size the OCPD using Art. 450.5 for grounding autotransformers. Find out the inrush current and stay above that to prevent nuisance tripping on startup.

 

[electrofelon]

The effective grounding XFMR is not used under normal steady state operation, it is used when there is a L-G fault on the utility distribution system that causes it to be isolated at the substation. This removes the ground reference for the distribution system in the substation and now you have a PV system feeding an ungrounded distribution system with a L-G fault until the inverter realizes something is wrong and shuts down. During that period it could cause a TOV event on the distribution system that may damage other customer's electrical systems. The effective grounding transformer will limit the TOV voltage and prevent that damage. That's the short explanation.

This makes no sense to me. Almost every grounded distribution system is an MGN, so its going to have tons of ground reference on the line. Furthermore, why does the substation lift the grounded conductor when the line trips out? I Wouldnt think they would regardless of if its an MGN or uni-grounded.

 

[pv_n00b]

This makes no sense to me. Almost every grounded distribution system is an MGN, so its going to have tons of ground reference on the line. Furthermore, why does the substation lift the grounded conductor when the line trips out? I Wouldnt think they would regardless of if its an MGN or uni-grounded.

It requires a bit of a deep dive. When there is a L-G fault on the distribution system there are several levels of protection that are used to clear it if it is temporary and isolate it if it is not. The end result is that with the distribution isolated from the utility generation the only thing left energizing the faulted line would be DG on the distribution system. Eventually the DG will notice that the utility is no longer energizing the line and trip off but during that period between when the utility is isolated and the DG trips off there is a danger of TOV. That is because the DG is typically an ungrounded source. The L-G potential of the unfaulted phases can rise to the L-L potential and damage anything connected to the distribution system. The purpose of the effective grounding transformer is to ground the DG during this period and limit the L-G potential on the unfaulted phases.

 

[electrofelon]

It requires a bit of a deep dive. When there is a L-G fault on the distribution system there are several levels of protection that are used to clear it if it is temporary and isolate it if it is not. The end result is that with the distribution isolated from the utility generation the only thing left energizing the faulted line would be DG on the distribution system. Eventually the DG will notice that the utility is no longer energizing the line and trip off but during that period between when the utility is isolated and the DG trips off there is a danger of TOV. That is because the DG is typically an ungrounded source. The L-G potential of the unfaulted phases can rise to the L-L potential and damage anything connected to the distribution system. The purpose of the effective grounding transformer is to ground the DG during this period and limit the L-G potential on the unfaulted phases.

Ok sure that makes sense for delta transformers, but if the transformer is gY->gY and there is MGN distribution, then what is the zigzag adding that isn't already done?

 

[pv_n00b]

Ok sure that makes sense for delta transformers, but if the transformer is gY->gY and there is MGN distribution, then what is the zigzag adding that isn't already done?

It gets pretty deep to go farther and it's not something that I can go into in the forum. There is too much to go into. I can just say this is how it works and why effective grounding transformers are required. There is some good material on the subject on the internet but it's hard to find and you have to filter out a lot of irrelevant grounding and bonding information.
I will say that you are confusing Ygyg transformers with grounded sources. If you have a 3W ungrounded source and connect it to a Ygyg transformer that does not ground the source on either side. It's still an ungrounded source. Now if you connect that ungrounded source to the delta side of a Ygd transformer then the WYE side is a grounded source.

 

[electrofelon]

I will say that you are confusing Ygyg transformers with grounded sources. If you have a 3W ungrounded source and connect it to a Ygyg transformer that does not ground the source on either side. It's still an ungrounded source. Now if you connect that ungrounded source to the delta side of a Ygd transformer then the WYE side is a grounded source.

Right, I would say its a potential issue of a wye winding, on its own, not making a good grounding transformer. The derived neutral has too high an impedance so the neutral can shift. Add a delta winding on the other side and you now have a good grounding transformer because current circulates in the delta winding which "holds" the neutral in place (someone can feel free to refine my explanation there if its a little off). So I guess adding another wye winding doesnt really add much to your other wye winding as far as deriving a low impedance neutral relative to the line terminals. It would disagree a little and say it does indeed ground the source, but perhaps not super well - but does it need to? Im not really seeing how the neutral would shift with balanced line currents and no neutral current coming from the inverter. My thinking is it should be adequate for modern inverters, which seems to be supported by (E) below which is from the National Grid spec book:

7.3.2.1 Effectively Grounded, Four-wire Multi-grounded 3-phase Wye EPS
To avoid overvoltage on the distribution EPS and other customers, the Company’s policy requires any DER facility 500 kW and above connected to a four-wire multi-grounded distribution feeder to provide an effectively grounded system with respect to the Company’s EPS. Refer to Section 7.1.4 for specific grounding requirements. The IC’s DER facility shall be designed such that it cannot connect to the Company’s EPS without the means of effective grounding in service. In all instances, when the means of effective grounding is out of service or disconnected from the system, the DER facility shall be disconnected from the Company’s EPS. Generally, a DER or aggregate DER below 500 kW in an IC’s facility may be permitted to utilize an ungrounded interconnection. The Company reserves the right to require an effectively grounded source for generation at 250 kW and above depending on DER saturation and other conditions on individual distribution EPS feeders. Grounding transformer impedance shall be specified in ohms per phase, as stated in ANSI C57.32 Section 6.4.6. Grounding transformer nameplates shall comply with ANSI C57.32 Section 6.8.2. Effective grounding may be accomplished with the following configurations:

(A) A wye-grounded to wye-grounded transformer with a grounded generator source (a generator which can demonstrate production of a sufficient amount of zero-sequence fault current). A neutral grounding reactor or grounding resistor between the generator neutral and ground may be required in event the generator's contribution to faults on the Company’s EPS results in undesirable fault current values. See Section 7.1.5.

(B) A wye-grounded connected primary winding with a fully insulated neutral and a delta connected secondary winding. The insulated neutral is to establish provisions for the addition of a grounding reactor or grounding resistor in the event the generator's contribution to faults on the Company’s EPS results in undesirable fault current values. See Section 7.1.5.

(C) A wye-grounded to wye-grounded transformer with an associated grounding transformer.

(D) Ungrounded primary configurations where zero sequence continuity is not maintained and having an associated primary side grounding transformer (Note: primary protection scheme with three phase gang operated breaker or recloser may be required).

(E) A wye-grounded to wye-grounded transformer with UL1741-SB inverters. A letter from both the inverter(s) manufacturer and Customer stating that the Reference Point of Applicability location meets all the IEEE 1547-2018 requirements must be submitted for review. The customer shall also submit a certification document, signed and dated by a NRTL, for GFOV testing for the site-specific system grounding aspects to be installed as specified by IEEE 1547.1-2020 Section 5.17 and UL1741 Section SB 4.3.5.17.1, which complies with IEEE 1547-2018 Section 7.4. Not Effectively Grounded, Three-wire 3-phase EPS:

 

[pv_n00b]

I have not seen option (E) used before. It seems to be saying that if the inverter has been tested to show it will not cause GFOV then an effective grounding transformer is not needed. Seems like a valid option if anyone does that test. It's probably not used much if at all due to this requirement:

a certification document, signed and dated by a NRTL, for GFOV testing for the site-specific system grounding aspects to be installed as specified by IEEE 1547.1-2020 Section 5.17 and UL1741 Section SB 4.3.5.17.1, which complies with IEEE 1547-2018 Section 7.4. Not Effectively Grounded, Three-wire 3-phase EPS

I mean, who is going to go to the trouble of getting that when they can just add an effective grounding transformer? I see option (C) used most often in PV systems.

 

[electrofelon]

I have not seen option (E) used before. It seems to be saying that if the inverter has been tested to show it will not cause GFOV then an effective grounding transformer is not needed. Seems like a valid option if anyone does that test. It's probably not used much if at all due to this requirement:

I mean, who is going to go to the trouble of getting that when they can just add an effective grounding transformer? I see option (C) used most often in PV systems.

Yeah could be. At first read I was thinking that is something that is provided by the inverter manufacturer and doesn't require testing in the field. But reading it again, I don't really know what is involved with getting that. But anyway I still think the whole thing is nonsense and am highly skeptical a modern inverter would ever cause GFOV on a MGN YgYg setup.

 

[ggunn]

Yeah could be. At first read I was thinking that is something that is provided by the inverter manufacturer and doesn't require testing in the field. But reading it again, I don't really know what is involved with getting that. But anyway I still think the whole thing is nonsense and am highly skeptical a modern inverter would ever cause GFOV on a MGN YgYg setup.

I went to a presentation by Hammond Power Solutions for PV applications a couple of weeks ago. For 480/277V inverters connecting to a 208/120V service, they recommend a 208V delta primary, 480/277V wye secondary transformer with the wye neutral bonded to the existing service ground. It's the way I have been doing it for several years.

 

[electrofelon]

I went to a presentation by Hammond Power Solutions for PV applications a couple of weeks ago. For 480/277V inverters connecting to a 208/120V service, they recommend a 208V delta primary, 480/277V wye secondary transformer with the wye neutral bonded to the existing service ground. It's the way I have been doing it for several years.

Yes for customer transformers I def agree. That should fulfill any POCO effective grounding requirements also.

Just to be clear, I am not advocating wye-wye customer owner transformers be used. I am just questioning and skeptical that a grounding transformer is needed for wye-wye utility transformers. Yes I know because you will probably say it, if you want your PV system connected you have to do what the POCO wants

We did a bunch of multi- megawatt projects in Georgia power territory on MGN's and they didn't require grounding transformers.

 

[ggunn]

Yes I know because you will probably say it, if you want your PV system connected you have to do what the POCO wants

Perzactly.

I have designed a bunch of systems like this because of 208/120V services and the dearth of larger inverters that are 208V native. Only once or twice has an AHJ required a wye-wye transformer for the interconnection.

 

[pv_n00b]

Yeah could be. At first read I was thinking that is something that is provided by the inverter manufacturer and doesn't require testing in the field. But reading it again, I don't really know what is involved with getting that. But anyway I still think the whole thing is nonsense and am highly skeptical a modern inverter would ever cause GFOV on a MGN YgYg setup.

I agree, I've read a lot on the subject, gone through the software simulations, etc.. It seems like a boogie man to utilities that they can use to make PV harder and more expensive to interconnect. The whole subject is based on decades of experience with rotating synchronous generators that have high fault current contributions and high reactance, and that does not transfer directly to static inverters used in PV. There is still a lot of confusion in utilities about how to design effective grounding transformers for static inverters because all the research is based on the high reactance of rotating generators, so the calculations that have been used to design effective grounding transformers can't be applied to static inverter systems with their low reactance. They have basically been winging it for years and there is no consistency between utilities. If I had a nickel for every time I have asked a utility for guidance on how they want the effective grounding transformer specified and I have been told, just give us your design and we will let you know if we like it.
The only actual over voltage I have seen consistently proven is load rejection over voltage where a very large PV system experiences a loss of load when isolated on the distribution system. With nowhere for that power to go it can result in LROV if the inverter does not react very fast to the loss of the distribution system loads. The funny thing is that an effective grounding transformer does not help in this situation since there is no ground fault.