Grounding a PV system to a Sub Panel in a detached structure

Status
Not open for further replies.
me-
There's a GEC coming out of the inverter, the neutral..

The GEC is NOT the neutral. It's it's own thing.

Ok. The neutral is not the GEC. Got it.
There's still a neutral coming from the inverter which has to be bonded somewhere.
If the inverter neutral in this diagram doesn't count as the ground conductor, the EGC of the inverter must be the GEC then?

When the GEC, as GoldD said:
"connects a Ground Electrode to the bond point. And it cannot contain connectors or reversible splices"

and:
"The grounded connector is bonded to the EGC and GES at exactly one point in the customer side of the service point.

while the GES:
"ends up connected to the main bond location (often at the main disconnect or main panel)."

It's that "one point" on the customer side that's bugging me. I can't quite figure out where it is, it doesn't seem to be where GoldD is saying it should be.



From post #33, GoldD-
A grounded supply of electricity has one grounded conductor (often it is also a neutral, but that is not always true)
In some cases (often in a PV installation) the same conductor may serve as both a GEC and an EGC.
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
The GEC is NOT the neutral. It's it's own thing.
Although it (the GEC) may share that function with the EGC, definitely not the neutral.

PV,
That is just the way panels and devices are labelled. A 200A 120/240 panel will have two 200A buses. A 208Y/120 three phase panel will have three 200A buses. A two phase five wire panel (if they existed) would have four 200A buses.
You do not get to change the industry standard naming convention based or your sense of how it should be.
A 200A breaker, regardless of how many poles it has, will limit each pole to 200A.
 

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
There's still a neutral coming from the inverter which has to be bonded somewhere.
It is grounded by being connected to the feeder neutral, which in turn is connected to the service neutral, which in turn is grounded.

If the inverter neutral in this diagram doesn't count as the ground conductor, the EGC of the inverter must be the GEC then?
In the second diagram yes. 690.47(C)(3) in the 2011 code allows one to combine the functions of the EGC and GEC into one conductor for a solar inverter. This is the only place in the code that allows such a combination, AFAIK. A new section was added to the code in 2011 explicitly prohibiting it in general.

It's that "one point" on the customer side that's bugging me. I can't quite figure out where it is, it doesn't seem to be where GoldD is saying it should be.

It's at the main service. There is a wire shown connecting the neutral and ground bars. That's the main bonding jumper.
The wire shown going to the existing UFER and water is the GEC at the service.
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
The neutral coming from the inverter goes to the panel or subpanel where the AC breaker is located. If the system bond is at the main panel, there will not be an electrical separation between neutral and ground even if two bars are present.
An inverter output circuit going to the main panel will have its neutral connected to the neutral bar along with all the other neutrals.
If the inverter output goes to a subpanel, then the neutral goes to the neutral bar there and the EGC goes to the EGC bar. From those two bars separate feeder wires go back to the main panel.
If you have a line side connection for PV the neutral and the EGC go to the PV disconnect (with OCPD) and most inspectors will require a ground to neutral bond there, on the PV side of the disconnect. Even though that may seem redundant or unnecessary.
 
Although it (the GEC) may share that function with the EGC, definitely not the neutral.

PV,
That is just the way panels and devices are labelled. A 200A 120/240 panel will have two 200A buses. A 208Y/120 three phase panel will have three 200A buses. A two phase five wire panel (if they existed) would have four 200A buses.
You do not get to change the industry standard naming convention based or your sense of how it should be.
A 200A breaker, regardless of how many poles it has, will limit each pole to 200A.

I get why there are two 200A busses.

The 100A secondary breaker in diagram has 2 poles of 100A each.
So let me rewrite this- I'm wondering more about what's going on with the grounding.


Say it's noon and the PV is at max output (32 amps).
The loads drawn from the secondary panel are cooking along at 50 amps of 240V (which is the same as 100 amps of 120V loads. The load is 12kVA)

That means there's 32A from PV and 18A from the grid supplying the sec. panel loads. Fine.

What if:
Someone decides to to bit of welding or something in addition.
A 60 amp welder.

That puts the secondary panel loads at 110A > than the 100A main breaker, less than the bus rating.
What happens then?

Another 50A coming from the grid (for a total of 68A coming from the grid) isn't going to trip the 100A breaker. Won't that cause bus issues?

Or...if it is going to trip the 100A breaker, what about all the other breakers (PV and loads) remaining closed on the load side of the tripped breaker? Won't 36A of PV output still be going to 110A of loads?
Seems like of course the PV breaker would also trip.

I'm wondering about when GoldD said-
"There is no cross section of the bus carrying more than 100A."

It does seem like the PV breaker would trip too, but even if it does, the N and G from the inverter are still connected, all the way to the main panel, and possibly the LV side service ground.
With all the load equipment breakers still closed.
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
If you start with more than about 90A 240V load before the welder, when it adds in 40A neither PV backfeed breaker nor feeder breaker will not *necessarily* trip, depending on how the current divides. If the PV produces more than 30A the feeder will not trip since it is supplying 130 - 30 or less. The PV breaker cannot trip first since it cannot have more than 32 A flowing. If the feeder breaker happens to trip the PV breaker still cannot trip, but the GTI will shut down.
If you turn off the PV or a cloud goes by then the feeder will probably trip.
Remember 32A of PV plus 40A of welder load makes 8A of load, not 72. I think this may be part of your problem.
 
me-
There's still a neutral coming from the inverter which has to be bonded somewhere.
jaggedben
It is grounded by being connected to the feeder neutral, which in turn is connected to the service neutral, which in turn is grounded.

So the "terminations" don't count as "points" then, towards the "only one point"?
And if the GEC/EGC from the inverter EGC lug is suposed to...
"connect a Ground Electrode to the bond point. And it cannot contain connectors or reversible splices"

the neutral bars aren't connectors either?

1 If the inverter output goes to a subpanel, then the neutral goes to the neutral bar there and the EGC goes to the EGC bar. From those two bars separate feeder wires go back to the main panel.

2 And a 200A 120/240 single phase 3 wire service can supply 400A worth of 120V loads if they are balanced and there are no 240V loads.

1 jben said that yes, the EGC from the inverter is its' GEC- so isn't it connected to more than one point?

2 Hoo boy. This is from 2009- the discussion seems, um, unsettled. The "reply" loses me...

http://forums.mikeholt.com/showthread.php?t=109846&p=998321#post998321
I don't agree.

The (200A 120/240V) panel can provide:
1 x 200A @ 240V
or
2 x 200A @ 120V

It can not provide:
1 x 400A @ 120V

It may seem like splitting hairs but the reality is that no single 120V load can be greater than 2400VA

reply:
Yes, you can supply 400a worth of 120v loads, but no, you cannot supply 400a at 120v.
Sorry, because the discussion is about main breakers, the 200a side wins.
 
If you start with more than about 90A 240V load before the welder, when it adds in 40A neither PV backfeed breaker nor feeder breaker will not *necessarily* trip, depending on how the current divides. If the PV produces more than 30A the feeder will not trip since it is supplying 130 - 30 or less. The PV breaker cannot trip first since it cannot have more than 32 A flowing. If the feeder breaker happens to trip the PV breaker still cannot trip, but the GTI will shut down.
If you turn off the PV or a cloud goes by then the feeder will probably trip.
Remember 32A of PV plus 40A of welder load makes 8A of load, not 72. I think this may be part of your problem.

Thanks.

I'm saying there's 50A of 240V load at noon, at the same time the PV output = 32A max.
Then the 60A of welder kicks in, putting the total load at 110A.
So the 100A breaker won't trip because there's only 78A coming from grid, the other 32 is going straight to loads from PV.

Is this just a case of 110A would be ok, but 135A on the 125A bus would cause a serious problem?

What if the PV is only outputting 10A into the 110A loads, and the main 100A breaker trips?
No danger of the bus staying energized and the inverter neutral becoming a conductor? This is really helpful, thanks again.
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
You are welcome, PVfarmer. If you put the PV backfeed breaker next to the main breaker the you will have 110A flowing in the buses just below that point, ansvit will overheat! If you put the PV breaker at the other end no part of the bus can carry more than 68 A.
Hence the 120% rule requirement for opposite end feeding.

PS: I do think that you are salvageable, it is just taking a long time. :)
 

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
me-
So the "terminations" don't count as "points" then, towards the "only one point"?

They don't count. The 'one point' is where that neutral conductor is connected to ground, (i.e. the dirt of the earth) via a main bonding jumper that in turn connects it to the grounding electrode conductor and electrode.

Neutrals are not to be connected to ground or to grounded parts (not counting other neutrals) in more that one place on the premises.
One exception to that, which is somewhat relevant to this thread, is that old codes allowed a neutral to be run to a detached building and for each detached building to have it's own neutral connection to the ground. There's still a grandfather clause in the code for that provided certain other conditions are met.

You know, our host Mike Holt has published a lot of resources that might help clear up a lot of this for you. Lots of nice illustrations. Try some of his Youtube vids.
 

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
Thanks.

I'm saying there's 50A of 240V load at noon, at the same time the PV output = 32A max.
Then the 60A of welder kicks in, putting the total load at 110A.
So the 100A breaker won't trip because there's only 78A coming from grid, the other 32 is going straight to loads from PV.

Just to note...
If you have that much potential load on the feeder then you had probably better upsize the 100A feeder and breaker anyway, because you can't count on the sun shining when you want to use that power.

This is one reason the 120% is a little bit silly, as it would take someone very nimble and knowledgeable to consistently use loads that high without suffering nuisance tripping.
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
Just to note...
If you have that much potential load on the feeder then you had probably better upsize the 100A feeder and breaker anyway, because you can't count on the sun shining when you want to use that power.

This is one reason the 120% is a little bit silly, as it would take someone very nimble and knowledgeable to consistently use loads that high without suffering nuisance tripping.
But the 120% rule is there whether there are any loads in the panel at all.
 

ggunn

PE (Electrical), NABCEP certified
Location
Austin, TX, USA
Occupation
Electrical Engineer - Photovoltaic Systems
1 It's gotta be either my outside of the box thinking...
Sometimes the box is there for a reason, like because of physics, NEC regulations, equipment capabilities, etc. The only time "outside the box" thinking is valid is when you understand what the box is and why it's there, otherwise it's just speculation from a position of ignorance.
 

ggunn

PE (Electrical), NABCEP certified
Location
Austin, TX, USA
Occupation
Electrical Engineer - Photovoltaic Systems
JAZ's main breaker for the sec. panel is 100A...that's 100A at 120V, correct?
His main breaker in the main panel is 200A- that means 200A of 120V and 100A of 240V.
No. A 200A breaker is 200A, period. 200A at 240V, 200A at 120V... just 200A. You do not add phase amps to get line amps.
 
They don't count. The 'one point' is where that neutral conductor is connected to ground, (i.e. the dirt of the earth) via a main bonding jumper that in turn connects it to the grounding electrode conductor and electrode.

You know, our host Mike Holt has published a lot of resources that might help clear up a lot of this for you. Lots of nice illustrations. Try some of his Youtube vids.
quote_icon.png
Originally Posted by GoldDigger

1 If the inverter output goes to a subpanel, then the neutral goes to the neutral bar there and the EGC goes to the EGC bar. From those two bars separate feeder wires go back to the main panel.
"The grounded connector is bonded to the EGC and GES at exactly one point in the customer side of the service point.


PV inverter |L1------>}sub panel
------------|L2------>}"
------------|N ------>} sub N bar
________
|
|
| ,<<< #8 EGC/GEC wire from inverter lug
|
>------->main panel ground bar

Isn't that all JAZ's POCO is asking for?
If the inverter neutral is going to the sub N bar, then the main N bar, then main ground bar and earth, they want the #8 EGC/GEC wire coming from the inverter lug to go to the main panel ground bar?
Because that's the one point?
I don't see how there's any other point to connect it (witho no splices or connections), unless as GoldD said, at the PV side of the disconnect switch for a line side connection.

Ok- jben- you said I should start a new thread.
How's this for a topic-
Line (supply) side versus load side PV connections- Pros/Cons/Experiences

But it doesn't seem like the greatest idea for a thread however- I can't really think of any cons, other than "it isn't allowed"...what say you?

PS: I do think that you are salvageable, it is just taking a long time. :)

Um...thanks!

otherwise it's just speculation from a position of ignorance.

Dude...I was trying to make a joke at my own expense referencing zman's "out of the box" thinking- and saying it may not always be a good thing.
 
This is one reason the 120% is a little bit silly, as it would take someone very nimble and knowledgeable to consistently use loads that high without suffering nuisance tripping.

So does that mean it will be even "sillier" if they raise 120% to 150%, as I've seen talk of doing?

This diagram seems a lot simpler today- JAZ's POCO doesn't want the green/G line coming from the inverter attached to the ground bar in the sub panel.
They want it to run from the inverter to the main panel ground bar unspliced/no connections.
So it's bonded at the one point, same point as the inverter N.
Correct?
 

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
So does that mean it will be even "sillier" if they raise 120% to 150%, as I've seen talk of doing?

No, it would be less silly. Basically, what people have said is that since the solar is there intermittently, any load that was over the busbar rating would probably trip the utility feed breaker at some moment when the sun was not shining. In other words, people can't count on using that much load, so really a 200% rule might not be silly, and the 120% rule is very, very conservative, at least when the source is solar. 150% seems like a good compromise n my opinion, for solar and wind, but I don't work for UL.

Don't get me wrong, the 120% rule makes technical sense in terms of preventing overcurrents. It also makes a lot of sense with sources other than solar, which might not be intermittent.

This diagram seems a lot simpler today- JAZ's POCO doesn't want the green/G line coming from the inverter attached to the ground bar in the sub panel.
They want it to run from the inverter to the main panel ground bar unspliced/no connections.
So it's bonded at the one point, same point as the inverter N.
Correct?

Who knows want they want exactly and why. Probably the guy asking for it isn't familiar with the code requirements for a GEC or else he wouldn't be asking for it. It's a good guess that giving them an unspliced/irreversibly spliced green wire back to the electrode (UFER) at the service will satisfy their request. Such a wire would still be connected to the subpanel ground bar and all the other grounded parts, but the idea is that would not be possible to remove the connection between inverter and service at any point in between without cutting the wire.
 
Last edited:

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
A lot depends on what function POCO is seeing for that wire. Technically a GEC runs uninterrupted from a ground electrode to the system bond point. If the inverter were the system bond point a GEC would run from there directly to one or more ground electrodes, not to the main panel.
An EGC, on the other hand, runs possibly indirectly to the system bond location at the main panel and may be spliced and joined along the way.
If it were not for the difficult concept of separate DC electrodes and grounds distinct from the AC grounds, there would not be any possibility of a GEC in those location.
A GEC and an EGC can share the same wire, but what runs from the inverter to the main panel seems to me to be a bonding jumper instead. And there is no specific rule that bonding jumpers cannot be spliced.
 
Status
Not open for further replies.
Top