Wye Inverter on a Delta Service

[Rock-Noll]

So, I haven’t spec’d equipment on anything this big before, and it’s looking like I might have messed up. The building has a 480 v delta service, I wired 4 solar inverters that are looking like they will only work on a Wye service. The thing that tripped me up is that the inverter manual says “When N is present and connected to the corresponding terminal, remove the bridge installed as standard between N and the enclosure” The inverter has a jumper from their “N” terminal to the inverter’s metal frame. So I thought that’s ok, don’t have a neutral, no problem, leave this jumper in. Then after turning things on and getting a grid error message, I looked closer at the spec sheet and see that it says Wye grid only. I am thinking I may need to install two Delta to Wye transformers now, ~ 50 kVA and 125 kVA (backfeeding to two identical services, four total inverters)

So I may have some questions about spec’ing transformers, but first I am confused about the nature of the service. When I was done with my wiring and started to turn things on, I checked for AC voltage coming into the inverter and I see ~480 volts between phases and ~277 phase to ground. There’s no neutral coming from the POCO, so how am I getting a voltage to my to grounding reference?

Any suggestions on avoiding installing transformer? I called SMA (inverter tech support) and 1st tier guy I spoke to is stumped. He’s consulting with 2nd tier, while I listen to EuroTech elevator music and looking things up on the web and going back to my old text book reviewing the difference between Wye and Delta services.

Thanks in advance!

 

[zbang]

Is this an ungrounded delta, corner-grounded, or high-leg? Sounds like the first (and there should be ground detectors somewhere) so you'll probably need to go the transformer route.

 

[ggunn]

If your inverter indeed needs a wye connection and if the service is indeed delta, you don't need two transformers; you can do it with one connected to both inverters, but your voltage references imply to me that it is a center grounded wye service with no neutral run, i.e., there is a neutral tied to ground at the transformer.

What specific inverter are you using? “When N is present and connected to the corresponding terminal, remove the bridge installed as standard between N and the enclosure” sounds like it's from an SMA Tripower manual, which is neutral agnostic and is shipped with neutral strapped to ground. If it couldn't run without a neutral, that strap would not be there.

I have designed a lot of systems with SMA Tripower inverters, all without neutrals. They may not run at all on a truly delta service, but I don't think that is what you have.

 

[jaggedben]

...When I was done with my wiring and started to turn things on, I checked for AC voltage coming into the inverter and I see ~480 volts between phases and ~277 phase to ground. There’s no neutral coming from the POCO, so how am I getting a voltage to my to grounding reference?

...

Those are wye voltages. Why do you say that there is no neutral coming to the service? I'm rather skeptical of that. Perhaps you have a bare neutral from the POCO and you aren't recognizing that for what it is. It would be an NEC violation and quite dangerous, as well as probably also more unlikely that you would read those voltages, without an grounded conductor from the POCO. You need to answer that question you asked that I quoted here. Except it would probably be put better this way: "Where is the neutral conductor that I haven't identified so far?"

I would also bet a sandwich that ggunn is right and you don't need a transformer.

 

[Rock-Noll]

Ggunn, yes you are correct, SMA inverters. I also hope you are correct about it’s actually a WYE service with a neutral I’ve failed to recognize. In the switch gear where I installed my breaker, there’s no neutrals to be seen. I didn’t get into the portion with the actual main disconnect, but am pondering if I should look there. But also thinks this might be fruitless because I am not sure if I can tell from there if it’s an uninsulated groundED or groundING wire, unless I get into the utility CT cabinet, which I can’t do.

I hope the utility transformer label will tell me the exact nature of the service, If not, how can I tell?

It’s a utility owned pad mounted transformer, with high and low voltage wires run underground. Looking at pictures I took of the CT cabinet last fall when we first bid this job, I do see one bare copper conductor. But from the picture I took, I can’t tell where it’s coming from. There’s six wires in parallel for each phase, ~500 kcmil, and then the one bare conductor, which I suspect is a grounding conductor, not grounded, is a bit smaller than that? Maybe 4/0. This was all live while the utility had the CT cabinet opened, so I didn’t get too close.

SMA says that it does not need a neutral run to it, but needs a ?center grounded transformer? (not center grounded on one phase) to maintain the 480 volts between phases. I am getting out of my league here, at this scale, I have always been installing equipment spec’d by others. I’ll just quote their “Requirements for Transformers”

Sunny Tripower CORE1 (STP 50-US-40) inverters require connection to a grounded Wye transformer winding. A neutral conductor connection to the transformer's grounded neutral point is optional.

The phase displacement of transformers does not affect the operation of Sunny Tripower inverters. Transformers with compatible winding configurations can be specified with any available phase displacement.

Connecting inverters to transformers that do not comply with these guidelines can result in unstable inverter operation, excessive nuisance tripping, and disruption or damage to the inverters or other connected equipment.

So, I failed to establish communication with the inverters yesterday, but while talking to tech support I got a few clues on how to proceed with that (FYI, the ser. #, password etc is under the inverter, ~3” off the ground, great place to hide it if you ever deal with these!) So, I am going back to hopefully get that done, adjust the grid profiles and see if I can get these running, but the last part of that manual quote there has me concerned “result in unstable inverter operation, excessive nuisance tripping, and disruption or damage to the inverters or other connected equipment.”

Also, two other items,

Another post on the forum mentioned that VFDs, MOVs, innate capacitance etc can provide the illusion of a center grounded system. I can’t find that again, so not sure what that’s about.

Also, on one side of the building, the voltage readings all looked like what I’d expect, a bit high, but that’s probably by design, I am reading voltages ~50’ from the main disconnect & feeder breakers. All the branch circuit panels, other transformers etc are 200’+ from this main gear. On the other side of the building, my readings phase to phase were all about 489 volts, but Brown to EGC, 279v, orange to EGC 308v, yellow to EGC 265 v. I first thought it was something odd on the utility side, but then realized might be an inherent issue with the Delta system, and draw on any one leg at that particular time.

So I am going back there this morning to see what the utility labels say (low hopes there, ~40+ years old, everything is faded) and see if I can adjust the grid profile to accept what I have. If I’m unsuccessful, then I’ll be emailing the utility engineer to see if he can shed some light, but that’s usually a long, torturous conversation.

Thanks!

 

[winnie]

I'm going make a guess:
1) The POCO transformer is a wye type transformer with its neutral point grounded. There is a neutral brought to the service, and the neutral is connected to a grounding electrode.

2) Only 3 circuit conductors are brought to the other side of the building, the neutral service conductor was left at the service. The EGC is not intact on the feeder, possibly there is no EGC, or possibly there is an EGC that was damaged by prior faults.

Jonathan

 

[ggunn]

I'll just add that there is no difference between a delta service and a wye service where the neutral is grounded at the transformer secondary but terminated there except that the phase voltages are referenced to ground, but that is enough difference for the SMA inverters to run with no neutral conductor and their neutral terminals strapped to the EGC. The fact that your phase voltages all appear to be within normal tolerance to 277V implies the latter; if all else fails can you get the POCO to open the transformer secondary so that you can see if there is a (grounded) neutral terminal?

 

[zbang]

Called the PoCo yet? They ought to be able to tell you exactly what's there, or send a lineman to check.

 

[ggunn]

Called the PoCo yet? They ought to be able to tell you exactly what's there, or send a lineman to check.

If it truly is a delta service, you may have to do as you theorized, i.e., connect the inverters back to the service through a grounded neutral delta-wye isolation transformer.

 

[electrofelon]

Just to throw one more possibility into the mix: measuring 277 line to ground on a truly ungrounded service would not be unusual. I have seen it. Definitely an ungrounded service but L-G voltages were all nice and "centered". I am not sure how the inverter is able to see through that, perhaps it uses a lower impedance measurement to see how much the voltage shifts as the capacitance is shorted out.

 

[Rock-Noll]

At the moment, it looks like Jaggedben could be correct. I managed to communicate with the inverters (part of this job I dread) and got them up up and running. With the inverter reporting on the line voltages, they all looked just fine! A little high, about 283 volts line to g, 490 volts line to line, on both services. So I am cautiously optimistic. They ran for about 2 hours with out a hiccup. I had to turn them off, but will be back next week for some loose ends and will turn them on again and see what happens.
I read some other posts about measuring 277 to ground on a truly ungrounded service, stuff about innate capacitance, VFDs, MOVs, that i don't really get. So, I'll be keeping a close eye on it and see what happens.
THANKS FOR YOUR HELP!

 

[ggunn]

At the moment, it looks like Jaggedben could be correct. I managed to communicate with the inverters (part of this job I dread) and got them up up and running. With the inverter reporting on the line voltages, they all looked just fine! A little high, about 283 volts line to g, 490 volts line to line, on both services. So I am cautiously optimistic. They ran for about 2 hours with out a hiccup. I had to turn them off, but will be back next week for some loose ends and will turn them on again and see what happens.
I read some other posts about measuring 277 to ground on a truly ungrounded service, stuff about innate capacitance, VFDs, MOVs, that i don't really get. So, I'll be keeping a close eye on it and see what happens.
THANKS FOR YOUR HELP!

IMO you should still investigate the service and verify that it is a grounded wye with no neutral. The data sheet clearly specifies that the SMA Tripower requires a wye connection; even if you can get yours to work on an ungrounded delta service there is no guarantee that they will continue to do so, and if any service on them were to be required the warranty would almost certainly not be honored.

 

[Rock-Noll]

THANKS FOR THAT!!! That's been on the back of my mind all day. I think I will. The more I do this solar stuff, the more I appreciate CYA!!

Thanks again for your lessons! I reread alot about 3 phase transformers this morning, and kind of get it, but not enough to be really confident.

WAIT A MINUTE!!! GROUNDED WYE WITH NO NEUTRAL!?! i HOPE YOU MEAN GROUNDED delta WITH NO NEUTRAL, or I am still really more confused.

Either way, I'll be following up with the POCO engineer on Monday

 

[ggunn]

THANKS FOR THAT!!! That's been on the back of my mind all day. I think I will. The more I do this solar stuff, the more I appreciate CYA!!

Thanks again for your lessons! I reread alot about 3 phase transformers this morning, and kind of get it, but not enough to be really confident.

WAIT A MINUTE!!! GROUNDED WYE WITH NO NEUTRAL!?! i HOPE YOU MEAN GROUNDED delta WITH NO NEUTRAL, or I am still really more confused.

Either way, I'll be following up with the POCO engineer on Monday

No, I mean a grounded wye with no neutral. What I mean is this: the utility transformer may have a wye secondary with a grounded neutral (X0) terminal but with no neutral conductor run out to the building, and your voltage measurements are consistent with this. If this is the case your inverters are fine with their neutral terminals strapped to the EGC. If it truly is an ungrounded delta service then there is an issue you will need to address.

 

[Hv&Lv]

Can you post the pic of the CT cabinet with the wires?
For a 480 service we put a N with it and connect it to the transformers. It's a 277/480. Use what you want.
My guess is there is the utility N in the CT cabinet that's squeezed to bonding copper, unless the contractor pulled the wire in. Take a pic of the padmount data plate. I have only dealt with two delta pad mounts in 35 years. They are aggravating behind a set of breakers(OCRs).
They have to be on three phase tripping protection devices. No triple singles.

 

[pv_n00b]

So I may have some questions about spec’ing transformers, but first I am confused about the nature of the service. When I was done with my wiring and started to turn things on, I checked for AC voltage coming into the inverter and I see ~480 volts between phases and ~277 phase to ground. There’s no neutral coming from the POCO, so how am I getting a voltage to my to grounding reference?

Everything is grounded through capacitive coupling of the conductors to ground. If you have a high impedance DMM and measure an ungrounded delta line to ground you will measure this capacitively coupled voltage. In a 480V system that will be 277V.
This is the basis for the really old school 3 lamp ground detectors. Hook up 3 lightbulbs line to ground in a WYE in an ungrounded delta system, if it's ungrounded they will all dimly glow the same amount. If one phase gets grounded the lamp on that phase goes out and the other two glow bright.

SMA says that it does not need a neutral run to it, but needs a ?center grounded transformer? (not center grounded on one phase) to maintain the 480 volts between phases. I am getting out of my league here, at this scale, I have always been installing equipment spec’d by others. I’ll just quote their “Requirements for Transformers”

This trips up a lot of people. They see in the manual that they don't need to run a neutral to the inverter and assume that means they don't need a grounded service. What the inverter is doing if it does not have the neutral is using the EGC as a neutral reference to measure phase to ground voltage since the EGC and neutral are bonded at the service. It's just a hack to not have to run a neutral to the inverter, but a grounded WYE service is still required or the inverter will fault on phase voltage. The manual has to specifically say that the inverter can connect to an ungrounded service.

 

[pv_n00b]

Those are wye voltages. Why do you say that there is no neutral coming to the service? I'm rather skeptical of that. Perhaps you have a bare neutral from the POCO and you aren't recognizing that for what it is. It would be an NEC violation and quite dangerous, as well as probably also more unlikely that you would read those voltages, without an grounded conductor from the POCO. You need to answer that question you asked that I quoted here. Except it would probably be put better this way: "Where is the neutral conductor that I haven't identified so far?"

I would also bet a sandwich that ggunn is right and you don't need a transformer.

I have seen on one project in decades of work where a utility upgraded an ungrounded delta service and replaced the delta transformer with a WYE and just did not run the neutral to the existing service. It was crazy and took quite a while to find out what was going on. No one expects the ungrounded WYE service.

 

[ggunn]

I have seen on one project in decades of work where a utility upgraded an ungrounded delta service and replaced the delta transformer with a WYE and just did not run the neutral to the existing service. It was crazy and took quite a while to find out what was going on. No one expects the ungrounded WYE service.

If the X0 of the transformer is grounded, it's a grounded wye service even if there is no neutral conductor from it to the service disconnect.

 

[ggunn]

This trips up a lot of people. They see in the manual that they don't need to run a neutral to the inverter and assume that means they don't need a grounded service. What the inverter is doing if it does not have the neutral is using the EGC as a neutral reference to measure phase to ground voltage since the EGC and neutral are bonded at the service. It's just a hack to not have to run a neutral to the inverter, but a grounded WYE service is still required or the inverter will fault on phase voltage. The manual has to specifically say that the inverter can connect to an ungrounded service.

The SMA inverter data sheet plainly states that the service must be wye, but the output is balanced so there is no neutral current. Why is it a "hack" to strap the neutral to the EGC instead of running a conductor back to the service? Either way, it's only a voltage reference; there are lots of three phase inverters that allow this, and for SMA Tripower inverters it is the default.

 

[electrofelon]

Everything is grounded through capacitive coupling of the conductors to ground..

Perhaps a nitpick , but not sure I like that phraseology, it's an oxymoron. I would leave out the"everything is grounded" part.