Tap Question

[csoc64]

In general, is it permissible to tap a PV AC output circuit onto one of two parallel conductors on the load side of a MCB, assuming fusing and length rules are properly applied?

 

[Carultch]

In general, is it permissible to tap a PV AC output circuit onto one of two parallel conductors on the load side of a MCB, assuming fusing and length rules are properly applied?

No. You are asking for trouble by doing this. Anytime there are parallel conductors, you have to preserve the symmetry as much as possible.

I made a graphic illustrating the problem with only tapping onto one of the parallel conductors. Problem: calculate the current in R1, R1', and R2, given the resistances and the voltages V1, V2, and the current IPV. This represents a long feeder that is made up of parallel 250 kcmils with a 2% voltage drop across the source feeder. The PV is tapped very close to the load on set #1 of that feeder.

Do you exceed the ampacity of the 250 kcmil wire, in any portion of it?


The proper way to do this, is with a 5-terminal insulated tap connector, that has an input and output port for both wires in the existing parallel set, and then a 5th port for the PV. Or any other method that allows you to connect onto both existing conductors. A nice way to do this, is when you have an unused lug position on the existing breaker.

 

[ggunn]

No. You are asking for trouble by doing this. Anytime there are parallel conductors, you have to preserve the symmetry as much as possible.

I made a graphic illustrating the problem with only tapping onto one of the parallel conductors. Problem: calculate the current in R1, R1', and R2, given the resistances and the voltages V1, V2, and the current IPV. This represents a long feeder that is made up of parallel 250 kcmils with a 2% voltage drop across the source feeder. The PV is tapped very close to the load on set #1 of that feeder.

Do you exceed the ampacity of the 250 kcmil wire, in any portion of it?
View attachment 15420

The proper way to do this, is with a 5-terminal insulated tap connector, that has an input and output port for both wires in the existing parallel set, and then a 5th port for the PV. Or any other method that allows you to connect onto both existing conductors. A nice way to do this, is when you have an unused lug position on the existing breaker.

FWIW, I see it done and approved by the AHJ pretty often.

 

[jaggedben]

The 5 terminal connector suggestion is wise. If you use this type (i.e. wires can enter from both sides), you need a minimum of slack to do it right and make any worries about this go away. Worth the extra money if you ask me.

 

[iwire]

The proper way to do this, is with a 5-terminal insulated tap connector, that has an input and output port for both wires in the existing parallel set, and then a 5th port for the PV. Or any other method that allows you to connect onto both existing conductors.

I agree 100% about that.

I had to do the same with soem feeder taps, If I remember correctly I needed seven port 600 kcmil connectors for it. The boss was not happy about it at $160 each for the four I needed.

 

[electrofelon]

I agree that it is generally good practice to connect to every conductor in a set, however is this explicitly required by code for feeder taps and/or pv interconnections? I guess I don't really see an issue if done properly....

 

[ggunn]

I agree that it is generally good practice to connect to every conductor in a set, however is this explicitly required by code for feeder taps and/or pv interconnections? I guess I don't really see an issue if done properly....

I know of several systems in a couple of different jurisdictions where a single conductor of a parallel set was tapped on each phase. They were inspected and approved by their respective AHJs. I have looked for code language that either expressly allows or denies this and I haven't found anything.

 

[jaggedben]

I think it's a violation of 310.10(H)(2)(1) and (5).

If I were an AHJ, that is what I'd hang my hat on, along with Carultch's diagram. It's an issue to be taken seriously. You can thank yourselves I'm not your AHJ.

 

[Smart $]

I think it's a violation of 310.10(H)(2)(1)...

I agree.

 

[electrofelon]

I think it's a violation of 310.10(H)(2)(1) and (5).

I dont agree with 1, but I see 5 as plausible. Perhaps 240.21(B) could be used also. We could use a proof by contradiction:

Let there exist a set of 2 parallel conductors sized 1/0 or larger.
1) suppose we tap one set of the parallel set.
2) Now by the description in 240.21(B) we have tapped a feeder so that one set is a feeder
3) By the definition of feeder, the untapped conductors are a feeder.
4) thus we have two feeders
5) this is a contradiction, therefore we cannot tap one set.

 

[Smart $]

I dont agree with 1...

How can you not agree? The conductors are not the same length.

Look at Carultch's diagram. You have parallel conductors from the V1 connection point to the IPV tap point. The length of conductor for R1 path is definitely shorter than the R1' + R2 path.

 

[electrofelon]

How can you not agree? The conductors are not the same length.

Look at Carultch's diagram. You have parallel conductors from the V1 connection point to the IPV tap point. The length of conductor for R1 path is definitely shorter than the R1' + R2 path.

Yeah i guess. The "original" parallel set becomes a parallel set for the tap also, whether you "wanted it to" or not.

 

[ggunn]

I think it's a violation of 310.10(H)(2)(1) and (5).

If I were an AHJ, that is what I'd hang my hat on, along with Carultch's diagram. It's an issue to be taken seriously. You can thank yourselves I'm not your AHJ.

If it is, it's a violation of an interpretation of those sections, not of the letter of the code. Personally, I don't care; in fact I always recommend tapping them all, but as I said, I know of several systems, installed and inspections passed, where a single conductor of each phase was tapped. The AHJ where these systems are installed is not known for being lax on points of code, either.

One point in favor of doing it that way is that with IPC's you can make the connection without having a service shutdown. I guess you could accomplish the same thing with a 4 hole Polaris block with the the inverter(s) on one terminal and IPC's on the other three.

 

[Carultch]

If it is, it's a violation of an interpretation of those sections, not of the letter of the code. Personally, I don't care; in fact I always recommend tapping them all, but as I said, I know of several systems, installed and inspections passed, where a single conductor of each phase was tapped. The AHJ where these systems are installed is not known for being lax on points of code, either.

One point in favor of doing it that way is that with IPC's you can make the connection without having a service shutdown. I guess you could accomplish the same thing with a 4 hole Polaris block with the the inverter(s) on one terminal and IPC's on the other three.

My diagram was set up to show that interconnecting onto just one set, will put disproportionate current onto that section of conductor, which does have the possibility of exceeding the ampacity of the individual conductor.

I agree that there is an advantage to using insulation piercing connectors, over their Polaris block counterparts, due to the fact that you can leave the existing conductors intact. But breaking the symmetry of the parallel feeds, is not a safe practice, unless your situation is set up to avoid overloading any individual conductor. One example would be, if the existing conductors operate at far less than their full ampacity, even with the PV system current, due to being upsized for voltage drop curtailment.

 

[ggunn]

My diagram was set up to show that interconnecting onto just one set, will put disproportionate current onto that section of conductor, which does have the possibility of exceeding the ampacity of the individual conductor.

I agree that there is an advantage to using insulation piercing connectors, over their Polaris block counterparts, due to the fact that you can leave the existing conductors intact. But breaking the symmetry of the parallel feeds, is not a safe practice, unless your situation is set up to avoid overloading any individual conductor. One example would be, if the existing conductors operate at far less than their full ampacity, even with the PV system current, due to being upsized for voltage drop curtailment.

How would you feel about the scenario I outlined with a four terminal Polaris block and three IPC's on each phase? It certainly would be a lot cheaper than using a seven terminal block on each phase with a shutdown necessary, and it would probably be a lot easier to fit into an existing gutter.

 

[Smart $]

How would you feel about the scenario I outlined with a four terminal Polaris block and three IPC's on each phase? It certainly would be a lot cheaper than using a seven terminal block on each phase with a shutdown necessary, and it would probably be a lot easier to fit into an existing gutter.

FWIW, convenience is never an excuse for safety. You should [next to] never install IPC's without shutting down.

Also, if shutting down isn't too awful inconvenient, I'd rather use gutter taps.

 

[ggunn]

FWIW, convenience is never an excuse for safety. You should [next to] never install IPC's without shutting down.

I think that IPC's are designed to establish connections without shutting down. Solar Pro (the magazine) thinks so, too.

From http://solarprofessional.com/articles/design-installation/can-we-land/page/0/11:

Insulation-piercing tap splice connectors. Usually the easiest conductor tapping device to install is a tap splice connector that pierces the conductor’s insulation to make electrical contact. These devices are available for a variety of wire sizes and types. Insulation-piercing tap splice connectors can be installed on live service conductors, making them the best choice when a shutdown is inconvenient or costly.

 

[electrofelon]

My diagram was set up to show that interconnecting onto just one set, will put disproportionate current onto that section of conductor, which does have the possibility of exceeding the ampacity of the individual conductor.
.

But the symmetry will remain for the "original" load. I would think that if someone were to tap one set, they would assume the untapped set wasn't there for load calculation purposes of the new load/source.

 

[GoldDigger]

But the symmetry will remain for the "original" load. I would think that if someone were to tap one set, they would assume the untapped set wasn't there for load calculation purposes of the new load/source.

If the original load was large enough to require parallel conductors, it will still require parallel conductors.
If the PV contribution toward satisfying the original load is significant it will severely overcurrent (in the direction of the load) the conductors which are tapped.
The voltage drop across the shortened length of parallel conductor will be higher than the voltage on the untapped conductor resulting is a current share which is proportional to the VD divided by the resistance. Net effect will be far larger than you might expect at first glance.

 

[Smart $]

I think that IPC's are designed to establish connections without shutting down. Solar Pro (the magazine) thinks so, too.

From http://solarprofessional.com/articles/design-installation/can-we-land/page/0/11:

Insulation-piercing tap splice connectors. Usually the easiest conductor tapping device to install is a tap splice connector that pierces the conductor’s insulation to make electrical contact. These devices are available for a variety of wire sizes and types. Insulation-piercing tap splice connectors can be installed on live service conductors, making them the best choice when a shutdown is inconvenient or costly.

Yes... but that's for when you put all the nuts and bolts of safety standards together (OSHA, 70E, etc.) and energized work is permitted. If the reason for not shutting down is the highlighted section above, that will get an employer in trouble with OSHA quicker than ____ (you name it). There are likely those out there that get away with doing it all the time, and nobody seems to care... until there's a tragedy that strikes somewhere close to home (or OSHA catches you without a tragedy involved and fines you big time).