Identifying Main vs. Sub Panel for off-grid PV system

Status
Not open for further replies.
J

jbowman88

Member
Location
Renton, WA
Hey there! I am new to the electrical trade and am helping out with the installation of an off-grid PV system with propane generator back-up. My questions are about grounding and bonding of the new system.

First a little background: this system is in a small garage building with attached office way out in the middle of nowhere. It had utility service at one point which fed a 125 Amp panel with only a few branch circuits for lighting and outlets. The utility power was wiped out by floods about 15 years ago, and since then the building has only been powered as-needed by a portable generator. The service conductors have been removed from the panel/service entrance, and there is no possibility of interaction with the grid for the foreseeable future.

We installed a 1.6 kW PV array with a 4.4 kW inverter and battery bank which connects directly to the main panel to feed the premises wiring. Now a generator (12 kW) has been installed behind the building. It feeds a small panel on the back of the building through a 50 Amp breaker. A branch circuit of this panel feeds the inverter (which has AC transfer capability built-in), which subsequently feeds the building loads. The generator is not a separately derived system as the inverter switching mechanism does not interrupt the neutral conductor.

The original panel in the building was grounded through a ground rod placed next to the service entrance, and had a system bonding jumper in place, as one would expect.

My questions are:
1. Since the original "main" panel in the building is now fed by the "generator panel", is it now to be considered and treated as a sub-panel?

2. If so, then shouldn't the bonding jumper be lifted and the neutral grounded at the generator panel instead?

3. Is the grounding electrode system required to be connected directly to the "Main" panel? In other words, can we use the existing ground rod at the sub-panel or does this need to be removed and a new (set) of rods driven at the generator panel?

There has been some considerable debate about this, so I am trying to get an answer that is actually based on code...
 
It sounds like the 50A breaker at the generator is just a disconnect so I wouldn't change anything with the bond at the main or the generator as a non SDS, and ground rods are not nearly important as lots of people make them out to be. As a matter of fact, remote ground rods can cause problems, so I would leave that as it is too.
 
I pretty much agree with Dave. To go into a bit more detail:

jbowman88 said:
...

My questions are:
1. Since the original "main" panel in the building is now fed by the "generator panel", is it now to be considered and treated as a sub-panel?
2. If so, then shouldn't the bonding jumper be lifted and the neutral grounded at the generator panel instead?
Click to expand...

Assuming that the generator does not have a built-in neutral-to-ground bond, or that it can be removed according to manufacturer's instructions, I don't see why it matters where the system bonding jumper is. The most important thing is to only have one system bonding jumper in one location and one grounding electrode system connected to that location. You have two non-separately derived sources, neither of which is a service. So I think it's basically arbitrary which source is treated as 'main'. What you've got to consider is what is the fault-path for both sources if there's a fault to ground, and will the system bonding jumper carry that fault current back to the source to trip the breaker for both sources. You haven't described anything that brings that into question.

Are there any loads that can be fed by the generator without going through the inverter? If not, I see no reason to treat the generator panel as the main panel. It's just a disconnect, as Dave said. But even if yes, I'm not sure why the generator should take precedence over the inverter as being the 'main' source.

3. Is the grounding electrode system required to be connected directly to the "Main" panel? In other words, can we use the existing ground rod at the sub-panel or does this need to be removed and a new (set) of rods driven at the generator panel?
Click to expand...

You haven't mentioned anything that would justify uprooting an existing electrode. At most, if you decided there was some important reason to relocate the grounding to the generator panel, just remove the existing grounding electrode conductor and route a new grounding electrode conductor from the existing rod over to that location.
 
That is all very helpful, thank you. I should probably clarify a few things:

1. There is an additional branch circuit on the generator panel. It goes to an RV/Trailer hook-up on the front of the building, bypassing the inverter to avoid overloading it.

2. The AC input on the inverter panel is rated to 30 Amps. My concern with having the neutral grounded at the original panel was that fault currents (up to 50 amps) on the generator side would have to travel all the way to the main panel and then back through the inverter's neutral wiring, which seems to only be designed for 30 amps. It's a UL listed device otherwise I would just up-size the wiring. So, I figured it would be better to have the bonding jumper at the generator panel instead.

This leads me to the question about the main vs sub panel. I am not necessarily concerned with the labelling of the panels. I want to know if/why the grounding electrode system needs to be connected directly to the same panel/location as the bonding jumper. In other words, can you have the bonding jumper in one panel and the GES on the other? Are there codes that pertain to this?

Thanks again
 
jbowman88 said:
That is all very helpful, thank you. I should probably clarify a few things:

1. There is an additional branch circuit on the generator panel. It goes to an RV/Trailer hook-up on the front of the building, bypassing the inverter to avoid overloading it.

2. The AC input on the inverter panel is rated to 30 Amps. My concern with having the neutral grounded at the original panel was that fault currents (up to 50 amps) on the generator side would have to travel all the way to the main panel and then back through the inverter's neutral wiring, which seems to only be designed for 30 amps. It's a UL listed device otherwise I would just up-size the wiring. So, I figured it would be better to have the bonding jumper at the generator panel instead.

This leads me to the question about the main vs sub panel. I am not necessarily concerned with the labelling of the panels. I want to know if/why the grounding electrode system needs to be connected directly to the same panel/location as the bonding jumper. In other words, can you have the bonding jumper in one panel and the GES on the other? Are there codes that pertain to this?

Thanks again
Click to expand...

A fault current from the generator should most likely be fine to travel on the EGC for the 30A inverter feed. (10awg wire size is the EGC requirement for 30-60 amps.). Yes, the fault current has to travel all the way to the other panel and back, but this just becomes the opposite case for the inverter as a source if the bonding-jumper is moved to the generator panel. Perhaps there is an argument that you need a 'supply side bonding jumper' instead of an EGC from the system bonding jumper to the ground bar for both sources, but that only moves you up one wire size (8awg), and is probably of no practical safety significance in this case.

I'm not aware of any code sections that directly address your situation. There are relatively clear rules for services, for separately-derived generators, and for solar systems, but not for a non-separately derived generator connected to the same system and a solar inverter only.

If one of the sources were much larger, then for fault current reasons that would argue for considering the larger source to be the 'main'. But if your EGC or supply side bonding jumper size requirement is the same under 250.122 and 250.102 for both sources, as it is in your case, then I think the choice of which is 'main' is really arbitrary. The inverter was your 'main' source before, so just keep it that way.

I think the grounding system should be connected to the same location as the bonding jumper, or on the 'main' supply side of the bonding jumper, whichever source you decide to be 'main'. That is more or less standard practice for grounded systems throughout the code.

Short summary: I don't think there's a clear safety reason or a clear code reason you have to do extra work on this system.
 
jbowman88 said:
This leads me to the question about the main vs sub panel. I am not necessarily concerned with the labelling of the panels. I want to know if/why the grounding electrode system needs to be connected directly to the same panel/location as the bonding jumper. In other words, can you have the bonding jumper in one panel and the GES on the other? Are there codes that pertain to this?

Thanks again
Click to expand...
250.24 tells you to bond the GEC at the service neutral/grounded conductor, but you don't have grounded service conductor. 690.47 gives you a bunch of different ways to attach a DC system to a AC GEC. I don't know anything about inverters, but I'll take jaggedben's word for it that it will be fine under a fault.

There is so much hand wringing about ground rods that is unnecessary. Check out this video by Mr. Holt himself....

 
ActionDave said:
250.24 tells you to bond the GEC at the service neutral/grounded conductor, but you don't have grounded service conductor. 690.47 gives you a bunch of different ways to attach a DC system to a AC GEC. I don't know anything about inverters, but I'll take jaggedben's word for it that it will be fine under a fault.

There is so much hand wringing about ground rods that is unnecessary. Check out this video by Mr. Holt himself....

Click to expand...

Great video, however it brings up a question... How is there not a voltage gradient during a lightning strike from say the water pipe Bond which may be at the southeast corner of a house, to the ground rods or CEE at the service which may be at the northwest corner? That could easily be a hundred foot of distance... I understand everything is bonded together, but doesn't such a setup automatically have more than one connection to Earth? I understand that an auxiliary rod at a generator or solar array could have a very large current flow during a lightning strike through the grounding electrode system to another point in Earth, and that would be bad for a house... Why is it not bad with a water pipe bond and ground rods being a different locations?
 
JFletcher said:
Great video, however it brings up a question... How is there not a voltage gradient during a lightning strike from say the water pipe Bond which may be at the southeast corner of a house, to the ground rods or CEE at the service which may be at the northwest corner? That could easily be a hundred foot of distance... I understand everything is bonded together, but doesn't such a setup automatically have more than one connection to Earth? I understand that an auxiliary rod at a generator or solar array could have a very large current flow during a lightning strike through the grounding electrode system to another point in Earth, and that would be bad for a house... Why is it not bad with a water pipe bond and ground rods being a different locations?
Click to expand...
That's not the point. The point is that so many people think that the best thing you can do to make anything electrical safer is to drive a ground rod or make some other connection to earth and the fact is it's not true.

Watch the video again. The point is not that a water pipe underground and a CCE underground and a ground rod underground cause problems. The point is that driving a ground rod and connecting something above ground to can be a bad idea. Dirt worshipers will never concede this fact.
 
Getting PV to play nice with a generator is a tricky proposition, especially in a completely off grid situation. In your system, what happens when the PV and generator are both running and the loads are consuming less than the output of the PV system? Is the PV inverter a battery inverter or a simple grid tied inverter that is accepting the generator output as the "grid" (a bad idea)? If it has batteries, does the PV inverter shut down when the batteries are fully charged and the generator is running?

If the PV backfeeds the generator, most of the time this is very bad, possibly catastrophic, for the generator. The generator can source power on demand but it cannot sink power, and, being essentailly a current source, a simple GT inverter cannot supply power on demand.
 
ggunn said:
Getting PV to play nice with a generator is a tricky proposition, especially in a completely off grid situation. In your system, what happens when the PV and generator are both running and the loads are consuming less than the output of the PV system? Is the PV inverter a battery inverter or a simple grid tied inverter that is accepting the generator output as the "grid" (a bad idea)? If it has batteries, does the PV inverter shut down when the batteries are fully charged and the generator is running?

If the PV backfeeds the generator, most of the time this is very bad, possibly catastrophic, for the generator. The generator can source power on demand but it cannot sink power, and, being essentailly a current source, a simple GT inverter cannot supply power on demand.
Click to expand...

It took me a bit to understand how this would work as well, but it's a pretty smart system. The inverter has an add-on generator control module that communicates with the inverter and charge controller and tells the generator when to start/stop. So the if the inverter senses that the batteries are getting too low (or too hot, or have been to low for too long, etc.) then it tells the generator to start up. Once the generator is running and warm, the inverter transfers the AC loads to the generator through internal relay and simultaneously stops inverter. It then switches to charging mode and uses the AC input from the generator to charge the batteries. Once the batteries are full, it shuts down the generator and switches back to inverting.

That is why the Generator MUST be connected to the premises wiring through the inverter. It will not work with a transfer switch. and is not like a typical grid-tied (utility interactive) inverter. It's designed specifically for battery based off-grid systems with optional generator or other intermittent AC sources in mind.
 
jbowman88 said:
It took me a bit to understand how this would work as well, but it's a pretty smart system. The inverter has an add-on generator control module that communicates with the inverter and charge controller and tells the generator when to start/stop. So the if the inverter senses that the batteries are getting too low (or too hot, or have been to low for too long, etc.) then it tells the generator to start up. Once the generator is running and warm, the inverter transfers the AC loads to the generator through internal relay and simultaneously stops inverter. It then switches to charging mode and uses the AC input from the generator to charge the batteries. Once the batteries are full, it shuts down the generator and switches back to inverting.

That is why the Generator MUST be connected to the premises wiring through the inverter. It will not work with a transfer switch. and is not like a typical grid-tied (utility interactive) inverter. It's designed specifically for battery based off-grid systems with optional generator or other intermittent AC sources in mind.
Click to expand...

Cool. It sounds like a Schneider (formerly Xantrax) system. What batteries are you using?

I agree with others; as long as there is only one neutral to ground bond it doesn't much matter where it is.
 
ActionDave said:
That's not the point. The point is that so many people think that the best thing you can do to make anything electrical safer is to drive a ground rod or make some other connection to earth and the fact is it's not true.

Watch the video again. The point is not that a water pipe underground and a CCE underground and a ground rod underground cause problems. The point is that driving a ground rod and connecting something above ground to can be a bad idea. Dirt worshipers will never concede this fact.
Click to expand...

Thanks for the video. I saw this a few weeks ago and it totally changed my understanding on grounding. I’m glad too, because I definitely was in the “if a little grounding is good, then lots of grounding must be better” camp. I probably would have gone crazy and pounded ground rods at the generator, generator panel, Inverter, IT rack, propane tank, garage door, and heck maybe a ground rod for each piece of metal furniture too! ??

I didn’t realize how pervasive that idea is. The contractor that installed the PV array drove a ground rod at the array, which is a pole mounted array 120’ away from the building. It’s conneted to the charge controller by a 150’ run of 1” IMC above-ground. So the inverter, charge controller, and lots of other expensive electronics are in-between two ground contact points that are 150’ apart in a lightning prone area.

The other funny part is that the ground rod is being used to ground a pole mounted array that consists of a 6” diameter galvanized steel pipe embedded 10ft into the ground and encased in 2 cubic yards of steel reinforced concrete (in very wet soil). I remember they mentioned that inspectors often wanted to see a ground rod at the array...

I guess the array is going to be functionally grounded whether there is a ground rod or not, so the only way to resolve the parallel ground path issue is to install a surge protection device on the PV input at the charge controller?
 
ggunn said:
Cool. It sounds like a Schneider (formerly Xantrax) system. What batteries are you using?

I agree with others; as long as there is only one neutral to ground bond it doesn't much matter where it is.
Click to expand...

The inverter, charge controller, and generator control module are all Magnum Energy components. The battery is 48 VDC, constructed from Outback Solar AGM cells.
 
jbowman88 said:
The inverter, charge controller, and generator control module are all Magnum Energy components. The battery is 48 VDC, constructed from Outback Solar AGM cells.
Click to expand...

Thanks. We occasionally have off grid PV requests and I did not know of Magnum Energy.
 
jbowman88 said:
Thanks for the video. I saw this a few weeks ago and it totally changed my understanding on grounding. I’m glad too, because I definitely was in the “if a little grounding is good, then lots of grounding must be better” camp. I probably would have gone crazy and pounded ground rods at the generator, generator panel, Inverter, IT rack, propane tank, garage door, and heck maybe a ground rod for each piece of metal furniture too! ??

I didn’t realize how pervasive that idea is. The contractor that installed the PV array drove a ground rod at the array, which is a pole mounted array 120’ away from the building. It’s conneted to the charge controller by a 150’ run of 1” IMC above-ground. So the inverter, charge controller, and lots of other expensive electronics are in-between two ground contact points that are 150’ apart in a lightning prone area.

The other funny part is that the ground rod is being used to ground a pole mounted array that consists of a 6” diameter galvanized steel pipe embedded 10ft into the ground and encased in 2 cubic yards of steel reinforced concrete (in very wet soil). I remember they mentioned that inspectors often wanted to see a ground rod at the array...
Click to expand...
I've done a fair amount of work on cell sites which have the most insane grounding specs that can be imagined, but I've never had to drive an extra ground rod.

I guess the array is going to be functionally grounded whether there is a ground rod or not, so the only way to resolve the parallel ground path issue is to install a surge protection device on the PV input at the charge controller?
Click to expand...
You lost me here. What parallel ground path?
 
ActionDave said:
I've done a fair amount of work on cell sites which have the most insane grounding specs that can be imagined, but I've never had to drive an extra ground rod.


You lost me here. What parallel ground path?
Click to expand...

I assume he's talking about the ground path through the ground and the one through the grounding conductor and/or conduit.
 
Status
Not open for further replies.
Top