Inverter N=G connection

Status
Not open for further replies.
augie47

augie47

Moderator
Staff member
Location
Tennessee
Occupation
State Electrical Inspector (Retired)
(You can read any response I have given in this forum and see that PV is one of my weak areas:), hence my question)
I recently looked at a PV install where the inverter had a nametag noting the output was "240". In the accompanying instructions it shows a neutral and "ground" connection. The points identified were obviously common to each other (verified by a meter). Inspector's dilemma: This is an interactive system. If a neutral and equipment ground are both brought from the premises wiring to the inverter we have a violation of 250.24(A)(5) by connecting them together.
How is this addressed in installations ?
 
augie47 said:
(You can read any response I have given in this forum and see that PV is one of my weak areas:), hence my question)
I recently looked at a PV install where the inverter had a nametag noting the output was "240". In the accompanying instructions it shows a neutral and "ground" connection. The points identified were obviously common to each other (verified by a meter). Inspector's dilemma: This is an interactive system. If a neutral and equipment ground are both brought from the premises wiring to the inverter we have a violation of 250.24(A)(5) by connecting them together.
How is this addressed in installations ?
Click to expand...
It seems to me that if neutral and ground are bonded inside the inverter, you cannot run the neutral from the service to the inverter. It's interesting, though; I have never tested for continuity between neutral and ground at the terminals of an inverter right out of the box. Is there maybe a removable jumper between them in your case?
 
Last edited:
No jumper. You can actually see the connection between the two etched in the circuit board.
With a nametag showing the output at "240" I was curious as to why it even showed a neutral connection.
The E/C connected it committing the neutral from the inverter to the service (actually to the combiner box).
 
augie47 said:
No jumper. You can actually see the connection between the two etched in the circuit board.
With a nametag showing the output at "240" I was curious as to why it even showed a neutral connection.
The E/C connected it committing the neutral from the inverter to the service (actually to the combiner box).
Click to expand...
Would you care to share the make and model of that inverter?
 
There's no reason to run the neutral from the inverter to the building's main service. The PV power is generated in DC, converted to AC, and can be connected to the service in single or three phase without bringing the neutral. Why they even have a lug for it at the inverter, I have no idea. It's not needed at all. I've heard the reason is because some inspectors require it, but that doesn't make any sense. When you do tie it in, it often causes nuisance tripping of the inverter offline.
 
augie47 said:
(You can read any response I have given in this forum and see that PV is one of my weak areas:), hence my question)
I recently looked at a PV install where the inverter had a nametag noting the output was "240". In the accompanying instructions it shows a neutral and "ground" connection. The points identified were obviously common to each other (verified by a meter). Inspector's dilemma: This is an interactive system. If a neutral and equipment ground are both brought from the premises wiring to the inverter we have a violation of 250.24(A)(5) by connecting them together.
How is this addressed in installations ?
Click to expand...


I don't think I've ever heard of an inverter that had an N=G bond at the inverter. I'm pretty sure that would violate a UL standard, at least for utility interactive inverters. (There might be some off-grid inverters with a bond.)

If you metered it while it was wired up, then you were measuring N and G as common through the main-bonding-jumper at the service. :happyyes:

Most inverters that require a neutral do so only because they use it to sense the voltage and phase relationship of the utility source. If so it can be sized the same as the equipment ground. In some cases an inverter is actually hooked up to the neutral as a power output, but I've only heard of that with 120V and 277V.
 
Pharon said:
There's no reason to run the neutral from the inverter to the building's main service. The PV power is generated in DC, converted to AC, and can be connected to the service in single or three phase without bringing the neutral. Why they even have a lug for it at the inverter, I have no idea. It's not needed at all. I've heard the reason is because some inspectors require it, but that doesn't make any sense. When you do tie it in, it often causes nuisance tripping of the inverter offline.
Click to expand...

Transformerless inverters often (always?) require a neutral connection for voltage sensing. It doesn't have to be as large as the phase conductors but it has to be at least as large as the EGC.
 
augie47 said:
SMA
SunnyBoy SB 8000US-12
Click to expand...
I am pretty sure that you can run that inverter at 240V phase to phase without the neutral if you set the jumpers accordingly, but it really surprises me that the neutral and ground are tied internally. I would call SMA tech support with your concerns. They have always been responsive to my questions.
 
Last edited:
ggunn said:
Transformerless inverters often (always?) require a neutral connection for voltage sensing. It doesn't have to be as large as the phase conductors but it has to be at least as large as the EGC.
Click to expand...
I've been thinking about this for several days now, to the point where it's even affected my sleep (I need to get a life, I think).

Because if a neutral is required for voltage sensing at the meter, how is the neutral-ground bond made? Is it to be considered a non-separately derived system at this point, where the neutral and ground run separately from the inverter and get tied to the main service entrance, like a feeder?

It's not just PV I'm wondering about - I also have applications with reciprocating cogeneration units that would run across the same issue.

It's so cut and dried with open transition transfer switches - if it's 4-pole, the generator can get a neutral-ground bond as a separately derived system, and if the ATS's are 3-pole, it's non-separately derived and the neutral and ground run separate from the generator back to the main service.

I'm guessing that if you have a source (i.e. inverter, cogen, etc.) running in parallel with the main service, it must be non-separately derived. Does this sound correct?
 
Last edited:
In most cases the POCO meter does not require a neutral for voltage sensing or any other purpose.
The meter assumes that any current carrying neutral is midway between the ungrounded conductors. To the extent that this is not true (e.g. voltage drop from an unbalanced load) the small metering error is in POCOs favor.
 
Pharon said:
...Because if a neutral is required for voltage sensing at the meter [inverter?], how is the neutral-ground bond made? Is it to be considered a non-separately derived system at this point, where the neutral and ground run separately from the inverter and get tied to the main service entrance, like a feeder?

...

I'm guessing that if you have a source (i.e. inverter, cogen, etc.) running in parallel with the main service, it must be non-separately derived. Does this sound correct?
Click to expand...

Voltage sensing on the neutral is required by most inverters. 3 phase meters measure the neutral current, single phase meters do not, but I think that's irrelevant to the subject here.

A transformerless inverter is not separately derived system; it does not meet the NEC definition in any way. An inverter with a transformer should probably not be considered a separately derived system either. A strict reading of the NEC definition might make it seem so, but to my knowledge none of the rules for separately-derived systems were written with inverters in mind and it does not make sense to apply them. Articles 690 or 705 or some other article will cover that ground and supersede article 250 etc..

An AC system only has N-G bonding at the service. No utility-interactive inverter should have an AC N-G bond at the inverter.




... It's not just PV I'm wondering about - I also have applications with reciprocating cogeneration units that would run across the same issue.

It's so cut and dried with open transition transfer switches - if it's 4-pole, the generator can get a neutral-ground bond as a separately derived system, and if the ATS's are 3-pole, it's non-separately derived and the neutral and ground run separate from the generator back to the main service.
...
Click to expand...

If there are transfer switches, then there is really no commonality of issues with grid-tie only inverters, since the latter doesn't have them. There might be some commonality with off-grid systems that have battery backup and a critical loads panel.

I'm not entirely sure what's keeping you up at night.:huh:
 
Thanks for the response. What's keeping me up at night is the fact that I have stamped multiple drawing sets with no neutral between the inverter and the main service, and if one is required for voltage sensing at the [revenue] meter, I'm worried that could come back and bite me.

Aside from that, I also want to make sure that I'm showing the correct neutral to ground bonds for all parallel source installs. But that's more of a curiosity than an insomnia issue.
 
It is not the meter that needs the neutral. When a GTI is feeding balanced current into 120/240 single phase, UL requires that the inverter disconnect if the line to neutral voltage becomes unbalanced beyond a set limit.
Not quite sure why, but the requirement is there.
 
GoldDigger said:
In most cases the POCO meter does not require a neutral for voltage sensing or any other purpose.
The meter assumes that any current carrying neutral is midway between the ungrounded conductors. To the extent that this is not true (e.g. voltage drop from an unbalanced load) the small metering error is in POCOs favor.
Click to expand...
In Austin Energy's jurisdiction they want a neutral the same size as the phase conductors run from the service through the PV meter (required by AE) to the next component (whatever it may be) irrespective of whether the inverter uses the neutral or not.
 
Pharon said:
Thanks for the response. What's keeping me up at night is the fact that I have stamped multiple drawing sets with no neutral between the inverter and the main service, and if one is required for voltage sensing at the [revenue] meter, I'm worried that could come back and bite me.
Click to expand...

I have seen many central inverters that do not have any sort of neutral connection lug at all. I have seen string inverters that have both a neutral lug and a ground lug whose manuals give the option of tying the neutral and ground together and not running the neutral back to the service. It's situational; I RTFM and do what it says to do.
 
Last edited:
Status
Not open for further replies.
Top