Supply-side vs Load-side PV Interconnection

[Opensolar]

I spent some time to figure out that PV AC disconnect is considered to be a service disconnect and has to have bonding at the switch and a GEC from AC switch into main panel board grounding electrode. (The existing main panel board already have G-N bonding there)

Question:

The rationale for not having a neutral to ground bonding at the AC disconnect for load-side taps is to avoid parallel current paths. How does supply-side tap avoid this considering we are tapping into the same neutral bus? ( regardless it's on supply or load side.)

Any thoughts?

Thanks

 

[Opensolar]

Here,I found this answer:

EGC carries fault current back to the electrical source under fault-conditions and Neutral carries unbalanced current. NEC wants to keep these two different currents separate. "Source Grounding and Bonding study David Burtt"

In this case, for supply-side tap NEC doesn't want to keep them separate?

 

[iwire]

In this case, for supply-side tap NEC doesn't want to keep them separate?

That is correct. In most cases when you have more than one service disconnect enclosure you will have parallel paths for neutral current.

I would say that the NEC would prefer not to have the parallel paths however the need for reliable bonding outweighs the negatives of parallel paths in a confined area.

 

[ggunn]

We have had debates on this in here, but really, what difference could it make? Typically the AC disco in a supply side tap is 10' or less as the wire flies from the tap point, and the neutral and ground are tied together at the service. The neutral and ground are not broken at the disco and even if they aren't tied together there they are connected close by through a low resistance path. What disaster could be averted by connecting the EGC and neutral in the disco?

Beyond that, in the case of an unbalanced set of single phase inverters feeding a three phase tap, there is current on the neutral, and tying neutral to ground in the disco would put current on the grounding conductor between the disco and the tap; I thought the point was to keep current off the EGC except in the case of a fault.

 

[texie]

We have had debates on this in here, but really, what difference could it make? Typically the AC disco in a supply side tap is 10' or less as the wire flies from the tap point, and the neutral and ground are tied together at the service. The neutral and ground are not broken at the disco and even if they aren't tied together there they are connected close by through a low resistance path. What disaster could be averted by connecting the EGC and neutral in the disco?

Beyond that, in the case of an unbalanced set of single phase inverters feeding a three phase tap, there is current on the neutral, and tying neutral to ground in the disco would put current on the grounding conductor between the disco and the tap; I thought the point was to keep current off the EGC except in the case of a fault.

I think that are installation scenarios that could lead to catastrophic consequences if the neutral is not bonded at the PV line side tap. What if the installer does not use a metallic method to connect the disconnect or if metallic that does not comply with 250.92? The fault return path would then be a long and round about way through the undersized EGC on the load side.

 

[iwire]

We have had debates on this in here, but really, what difference could it make?

Not bonding this new supply side disconnect could result in hazardous voltages on the enclosure.

Typically the AC disco in a supply side tap is 10' or less as the wire flies from the tap point, and the neutral and ground are tied together at the service.

There are no tap distances or tap rules for service conductors.

The neutral and ground are not broken at the disco and even if they aren't tied together there they are connected close by through a low resistance path. What disaster could be averted by connecting the EGC and neutral in the disco?

The service conductor supplying the disconnect faults to the enclosure, without proper bonding the enclosure could go hot or cause arcs at fittings.

Beyond that, in the case of an unbalanced set of single phase inverters feeding a three phase tap, there is current on the neutral, and tying neutral to ground in the disco would put current on the grounding conductor between the disco and the tap; I thought the point was to keep current off the EGC except in the case of a fault.

It would be nice to keep the current off of the enclosures but the need for proper bonding exceeds that.

 

[ggunn]

I think that are installation scenarios that could lead to catastrophic consequences if the neutral is not bonded at the PV line side tap. What if the installer does not use a metallic method to connect the disconnect or if metallic that does not comply with 250.92? The fault return path would then be a long and round about way through the undersized EGC on the load side.

Maybe (probably) I am not understanding you. Neutral and EGC pass through the disco back to the tap. Neutral and EGC are bonded in the service at or near the tap. All conductors are sized correctly. The only difference I see with bonding in the disco is that fault current for a phase to ground fault out past the disco will return through both the neutral and the EGC from the disco to the tap instead of just though the EGC. Available fault current is limited by the OCPD in the disco and EGC is sized accordingly. Where are the catastrophic consequences?

 

[c_picard]

"There are no tap distances or tap rules for service conductors. "

705.31 in 2014 NEC. I'm not a fan of this requirement as there is no precedent for non-PV service entrance conductors. I'm proposing this either get deleted in 2017, or apply the requirement uniformly for all service entrance conductors.

Apologies to the OP for the distraction. To the original point, the grounded conductor should be bonded at a PV supply-side connection for all the same reasons it would be at a conventional service disconnect.

 

[c_picard]

@ggunn

There is typically no EGC on the line side of a service enclosure; only a grounded conductor and GEC. If the grounded conductor is not bonded, and a line-to enclosure fault occurs on the supply side, the enclosure and the downstream EGC system is what would attempt to carry the fault back to the source, in a roundabout an high-resistance way.

 

[texie]

Maybe (probably) I am not understanding you. Neutral and EGC pass through the disco back to the tap. Neutral and EGC are bonded in the service at or near the tap. All conductors are sized correctly. The only difference I see with bonding in the disco is that fault current for a phase to ground fault out past the disco will return through both the neutral and the EGC from the disco to the tap instead of just though the EGC. Available fault current is limited by the OCPD in the disco and EGC is sized accordingly. Where are the catastrophic consequences?

I think we are misunderstanding each other. Picture a supply side tap terminated at a fusible disco. that the inverter connects to. Assume the installer chose to use a PVC nipple from the existing supply point to the disco which is commonly done on the line side of service discos). There would be no EGC in this raceway between the connection point and the disco. as it is not required (or even allowed). The EGC from the inverter stops at the supply side connection disco. Without the neutral bonded here the fault current path of one of the supply conductors shorting to the disco would now be back through the EGC of the inverter circuit which it not sized to handle, nor is it a compliant fault path. The result would likely be catastrophic as you would have the full current of whatever the service conductors could deliver flowing over the EGC back to the inverter and the EGCs upstream of that back to the service in a long and undersized path.

 

[texie]

@ggunn

There is typically no EGC on the line side of a service enclosure; only a grounded conductor and GEC. If the grounded conductor is not bonded, and a line-to enclosure fault occurs on the supply side, the enclosure and the downstream EGC system is what would attempt to carry the fault back to the source, in a roundabout an high-resistance way.

Since you and I have discussed this before, I think we are on the same page. I hope my scenario in my previous post of what could happen with no neutral bond makes sense.

 

[Smart $]

Not bonding this new supply side disconnect could result in hazardous voltages on the enclosure.

...

The service conductor supplying the disconnect faults to the enclosure, without proper bonding the enclosure could go hot or cause arcs at fittings.

It would be nice to keep the current off of the enclosures but the need for proper bonding exceeds that.

The debates haven't been whether or not to bond the disconnect enclosure. It's how to bond it. You have to establish an EGC for the PV side of the disconnect.

To generalize, is the disconnect is a service disconnecting means? It doesn't supply power to a GTI PV system (unless it's a hybrid). From there is the debate wheter or not a main bonding jumper is required if a GEC is run to the disconnect. If we install a main bonding jumper, is an EBJ and/or a GEC required to be run to it? ...and so on.

 

[ggunn]

@ggunn

There is typically no EGC on the line side of a service enclosure; only a grounded conductor and GEC. If the grounded conductor is not bonded, and a line-to enclosure fault occurs on the supply side, the enclosure and the downstream EGC system is what would attempt to carry the fault back to the source, in a roundabout an high-resistance way.

OK, but in PV systems which are interconnected with a supply side tap, there is typically (always?) an EGC from the PV system through the disconnect back to the existing service sized to the OCPD in the disco. Maybe I'm just dense but I still don't see the danger from not tying them together in the disco when they are already tied together in the existing service.

There doesn't even necessarily have to be a neutral; with most three phase inverters the neutral, if it's there at all, is only a voltage reference and can be sized the same as the EGC.

 

[Opensolar]

That is correct. In most cases when you have more than one service disconnect enclosure you will have parallel paths for neutral current.

I would say that the NEC would prefer not to have the parallel paths however the need for reliable bonding outweighs the negatives of parallel paths in a confined area.

That's what I need to know. Why NEC prefers parallel paths on supply side and not on the load side?

 

[c_picard]

That's what I need to know. Why NEC prefers parallel paths on supply side and not on the load side?

On the load side we do not want current flowing on exposed metal parts under normal operating conditions.

On the supply side we're trying to minimize arcing under a fault-to-case scenario by providing the most direct, low-resistance path back to the source possible, and this takes the form of a reliable and robust connection from the enclosure and GEC to the service "neutral".

I'll quote an inspector I spoke to recently on this matter, an inspector I couldn't disagree more with on this topic. When asked if he was concerned about a potential fault in the service enclosure, and understood why we bond neutrals in the first place, his response was "WHO CARES!!, LET IT BURN!!". I care, that's who.

 

[ggunn]

I think we are misunderstanding each other. Picture a supply side tap terminated at a fusible disco. that the inverter connects to. Assume the installer chose to use a PVC nipple from the existing supply point to the disco which is commonly done on the line side of service discos). There would be no EGC in this raceway between the connection point and the disco. as it is not required (or even allowed). The EGC from the inverter stops at the supply side connection disco. Without the neutral bonded here the fault current path of one of the supply conductors shorting to the disco would now be back through the EGC of the inverter circuit which it not sized to handle, nor is it a compliant fault path. The result would likely be catastrophic as you would have the full current of whatever the service conductors could deliver flowing over the EGC back to the inverter and the EGCs upstream of that back to the service in a long and undersized path.

Hmmm. I have designed and installed several commercial PV systems in four different AHJs' territories with supply side taps and I have in all cases connected the EGC (sized for the OCPD in the disco) from the inverters back through the fused disco to the grounding bar in the MDP. I have not tied the neutral (if there was one) to the ECG in the disco.

I have never been flagged by an inspector or in an AHJ's engineering department's interconnection study for it. Silly me; I thought that meant I was doing it correctly.

But on the other hand, I have in front of me a new (to me) AHJ's sample three line drawing for a supply side interconnection, and they show it the way you describe. The EGC stops in the disco and is bonded to the neutral.

Film at eleven.

 

[jaggedben]

Hmmm. I have designed and installed several commercial PV systems in four different AHJs' territories with supply side taps and I have in all cases connected the EGC (sized for the OCPD in the disco) from the inverters back through the fused disco to the grounding bar in the MDP. I have not tied the neutral (if there was one) to the ECG in the disco.

I have never been flagged by an inspector or in an AHJ's engineering department's interconnection study for it. Silly me; I thought that meant I was doing it correctly.

What is the overcurrent protection on the line side of that disco? There is none (or rather, it's whatever is on the utility transformer).

The available fault current on each side of that disco is NOT the same. At the very least, the 'EGC' on the line side of the disco should be sized for 250.66 and not 250.122.

 

[iwire]

That's what I need to know. Why NEC prefers parallel paths on supply side and not on the load side?

Because it is the lesser of two evils in this case. It has been explained a couple of different ways in this thread.

 

[iwire]

Hmmm. I have designed and installed several commercial PV systems in four different AHJs' territories with supply side taps and I have in all cases connected the EGC (sized for the OCPD in the disco) from the inverters back through the fused disco to the grounding bar in the MDP. I have not tied the neutral (if there was one) to the ECG in the disco.

Consider the fault current path in that arrangement if one of the service conductors feeding this disconnect faults to the enclosure.

Is that little EGC sized per 250.122 capable of handling full utility current?

 

[iwire]

To generalize, is the disconnect is a service disconnecting means? It doesn't supply power to a GTI PV system (unless it's a hybrid). .

I am not going to waste time on that debate, it is meaningless to me. Electricity does not care what the NFPA decides to call things.

From there is the debate wheter or not a main bonding jumper is required if a GEC is run to the disconnect. If we install a main bonding jumper, is an EBJ and/or a GEC required to be run to it? ...and so on

For me there is no debate, safety demands that an enclosure containing unfused service conductors requires reliable bonding back to the source and an EGC sized per 250.122 is not generally up to that task.