Feeder vs Tap Feeder

[charlie b]

I am getting into this one late, and I need to digest the discussion before offering an opinion. But I will start with this:

Anywhere NEC text = "Feeder", can you replace with "Tap Feeder" as an equivalent statement ?

Nowhere in the discussion did I see evidence that there are any "taps" involved in the installation at all. Specifically, I have not seen mention of wire sizes or ampacities. To directly answer this question, no they are not the same. You don't have a "tap" unless, (1) You connect wire 2 to wire 1, (2) The OCPD protecting wire 1 is rated higher than would normally be needed to protect wire 2, and (3) You don't install an OCPD at the point of connection of wire 2 to wire 1, so that that OCPD protects wire 2 at its ampacity. I think we have item (1) in this installation, but we don't yet know about items (2) or (3).

 

[wwhitney]

Your (A₁, A₂, B₁, B₂, C₁, C₂, N) scenario describes exactly a 3 apartment riser with taps connected to the 4W riser and ALL sharing a common N.
Visualize the N simply looping its way up the riser.

300.3(B) requires all conductors of the same circuit to be within the same raceway or cable, with certain exceptions. Let's ignore for the moment the possibility of parallel conductors. [Or treat any group of conductors in parallel as a single conductor.]

So a common neutral in the sense of 215.4 occurs when a single raceway or cable contains two different circuits (meaning there will be two different ungrounded conductors on the same phase) but contains only one neutral conductor shared by both circuits.

If I understand the wiring diagram you're describing, at any point there is only one circuit in the raceway or cable. So there are no common neutrals.

Cheers, Wayne

 

[don_resqcapt19]

Can we take this back to the your first post regarding feeder vs tap feeder

There are feeder conductors and tap conductors....I have no idea what a "tap feeder" is.

In the text of 220.82(A)...single set of 3W...feeder & 220.84(A)(1)...no dwelling supplied by more than one feeder
In both of these statements are you saying a feeder and tap feeder are equivalent ?

There is only one feeder to each dwelling unit. That feeder is a combination of a feeder circuit and a tap circuit.

Does not the entire scope of art 220 demand that the single or one feeder be capable of delivery the designed ampacity whether it is from the standard calc or optional calc ?

The conductors in the feeder circuit are sized based on the total connected load that the feeder serves. The conductors in the tap circuit are sized based on the connected load for the dwelling unit that the tap circuit feeds.

I content, if you allow the 4W riser to be derated, it would be impossible to deliver design amps without tripping out riser OCD.
Again, Delta 3W loads tapped from a 4W riser, reflect back onto the riser at a factor of 1.73.

I have no idea what that means.

Also Charlie brought up a good point. Without knowing conductor sizes and the overcurrent protective device ratings, we don't really know that you even have a tap circuit, and if you do there are some very specific rules that apply to the length of tap circuits.

 

[charlie b]

This may duplicate what others have said. But you asked my opinion via PM, so here it is.

Supply conductors to apartment, "Tap Feeder" is then the same as "Feeder"

The phrase “Tap Feeder” is not defined, and makes no sense. I suggest you forgo using it in the rest of this discussion.

Does not a Tap Feeder (3W), connected to a 4W riser result in apartment served via TWO feeders

It does not, and I think this is where you are misunderstanding the code language. Let me put it this way: Stand inside the apartment and look for wires coming into the apartment from the outside world. You are only going to see one set, one feeder. You don’t have to follow that feeder and trace it back to the point at which it gets its power. There is only one feeder coming into the apartment, and that is all that matters.

Same for 220.84A)((1), apartment is serviced via two feeders, and therefore access to derating table 220.84 is not allowed

Here again, not true, for the reason given above.

All 3W Tap Feeders are connected to the 4W riser, therefore you end up with 3-6 3W Tap Feeders SHARING a common neutral....215.4(A) states this is not allowed.

When this article talks about feeders sharing neutrals, it means two circuits, powered by two different circuit breakers, each comprising its own A, B, and C wires, but having a common neutral. So leaving the panel through a common conduit you have A1, B1, C1, A2, B2, C2, N, and G, for a total of 8 wires. Somewhere downstream, in a junction box, you connect the incoming N to a pair of outgoing N1 and N2, and you connect the incoming G to a pair of outgoing G1 and G2. From the left side of the junction box, you have A1, B1, C1, N1, and G1 in one conduit going somewhere. From the right side of the junction box, you have A2, B2, C2, N2, and G2 in a different conduit going somewhere else. At that point, circuits 1 and 2 are no longer sharing neutrals.

 

[MyCleveland]

300.3(B) requires all conductors of the same circuit to be within the same raceway or cable, with certain exceptions. Let's ignore for the moment the possibility of parallel conductors. [Or treat any group of conductors in parallel as a single conductor.]

So a common neutral in the sense of 215.4 occurs when a single raceway or cable contains two different circuits (meaning there will be two different ungrounded conductors on the same phase) but contains only one neutral conductor shared by both circuits.

If I understand the wiring diagram you're describing, at any point there is only one circuit in the raceway or cable. So there are no common neutrals.

Cheers, Wayne

Instead of tapping the N on the 4W riser at each 3W apartment, I simply loop from apartment to apartment up the riser serving 3-6 apartments, this is not sharing ?
The 4W and the 3W are in the same panel interior.

 

[don_resqcapt19]

Instead of tapping the N on the 4W riser at each 3W apartment, I simply loop from apartment to apartment up the riser serving 3-6 apartments, this is not sharing ?
The 4W and the 3W are in the same panel interior.

No...they are still two separate circuits, even when you loop the neutral.

 

[MyCleveland]

This may duplicate what others have said. But you asked my opinion via PM, so here it is. The phrase “Tap Feeder” is not defined, and makes no sense. I suggest you forgo using it in the rest of this discussion.
It does not, and I think this is where you are misunderstanding the code language. Let me put it this way: Stand inside the apartment and look for wires coming into the apartment from the outside world. You are only going to see one set, one feeder. You don’t have to follow that feeder and trace it back to the point at which it gets its power. There is only one feeder coming into the apartment, and that is all that matters.
Here again, not true, for the reason given above.
When this article talks about feeders sharing neutrals, it means two circuits, powered by two different circuit breakers, each comprising its own A, B, and C wires, but having a common neutral. So leaving the panel through a common conduit you have A1, B1, C1, A2, B2, C2, N, and G, for a total of 8 wires. Somewhere downstream, in a junction box, you connect the incoming N to a pair of outgoing N1 and N2, and you connect the incoming G to a pair of outgoing G1 and G2. From the left side of the junction box, you have A1, B1, C1, N1, and G1 in one conduit going somewhere. From the right side of the junction box, you have A2, B2, C2, N2, and G2 in a different conduit going somewhere else. At that point, circuits 1 and 2 are no longer sharing neutrals.

Thank you

 

[jaggedben]

I see your point with the math...simply change the apartments to standard calc design at 20.6kW each.
You still end up with three 99amp apartments served via a 80amp 4W (derated N relative to each tap) riser

I don't see how you end up with an 80A main feeder if each apartment requires a 100A feeder minimum under the smallest permitted code option. I refer to my previous post.

 

[MyCleveland]

I don't see how you end up with an 80A main feeder if each apartment requires a 100A feeder minimum under the smallest permitted code option. I refer to my previous post.

This is the basis of my entire argument and exactly why I believe that derating table 220.84 should not be allowed in this 4W common riser design.
Your rebuttal on my 100amp-3W apartment example and 80A-4W riser assumed I arrived at the apartment load via the optional calc and if true your 4W riser size would be much larger than 80amps.

I came back with a 99amp apartment design, which would not be allowed to use 220.82, but the apartment 3W "Feeder Tap" remains 100amp.
The math to size the 4W-riser still results in an 80A-4W common riser, thereby smaller in phase and N than the TAPS it serves.

Staying on when one is allowed access to the derating table...
The text of 220.84(A)(1) states .... no dwelling served by more than one feeder.
The connection to the 4W-riser is a feeder-tap

Staying on when one is allowed access to the optional calc for the individual dwelling unit (Please don't confuse this with the 99amp referenced above)
The text of 220.82(A) states single set 3W feeder
The connection to the 4W-riser is a feeder-tap

Forgive me all for starting with Tap-Feeder....at the start.

Can one replace the text AS written in both of these locations and infer that feeder can be replaced with feeder-tap ?
Knowing these loads are delta on a Y system and increase in size by a factor of 1.73 once reflected onto the 4W riser.

 

[Smart $]

...
I came back with a 99amp apartment design, which would not be allowed to use 220.82, but the apartment 3W "Feeder Tap" remains 100amp.
The math to size the 4W-riser still results in an 80A-4W common riser, thereby smaller in phase and N than the TAPS it serves.
....

You have to sum the before-demand load values of the three apartments to calculate the 45% demand under 220.84. Using the sum of apartment after-demand values (i.e. 100% first 10kVA plus 40% of remainder) is producing the error you are seeing.

 

[Smart $]

...Knowing these loads are delta on a Y system and increase in size by a factor of 1.73 once reflected onto the 4W riser.

The apartment loads are open wye... not delta.

Yes combined line loads increase approximately at a factor of 1.732.... but that's better than by a factor of 2.000, is it not? Another way to look at it is the combined currents are 86.6% of what they would be when combined on a single phase system.

 

[charlie b]

I think it would help if you would stop putting the word "feeder" and the word "tap" anywhere close to one another. You are somehow making it sound, incorrectly so, that there is a relationship between the words. I defined "tap" in an earlier post. A "feeder" is a set of conductors that is downstream of the utility service conductors and upstream of the last overcurrent device serving a load. It does not matter whether there is an overcurrent device at the starting point of the feeder (the most common configuration) or whether the feeder originates by a connection to another feeder. I can have a feeder start at a breaker and end at a panel, and I can connect another set of wires somewhere in the middle and run it to a different panel. That gives me two feeders going separate places. But each place is getting only one feeder. If the wire sizes are all the same, then there is no "tap" involved.

 

[Smart $]

I think it would help if you would stop putting the word "feeder" and the word "tap" anywhere close to one another. You are somehow making it sound, incorrectly so, that there is a relationship between the words. I defined "tap" in an earlier post. A "feeder" is a set of conductors that is downstream of the utility service conductors and upstream of the last overcurrent device serving a load. It does not matter whether there is an overcurrent device at the starting point of the feeder (the most common configuration) or whether the feeder originates by a connection to another feeder. I can have a feeder start at a breaker and end at a panel, and I can connect another set of wires somewhere in the middle and run it to a different panel. That gives me two feeders going separate places. But each place is getting only one feeder. If the wire sizes are all the same, then there is no "tap" involved.

MyCleveland may be looking at this from a different perspective. And the tap scenario of his installation appears to obfuscate the matter...

Say you have a main service panel which supplies four subpanels and those panels in turn supply 3 apartments each. MyCleveland appears to be looking at this as each apartment is being supplied by two feeders, 1) from service panel to subpanel, and 2) from subpanel to apartment panel. This is not what is meant by the requirement the dwelling unit be supplied by not more than one feeder, but from this perspective, it is being supplied by two feeders... one after the other.

 

[jaggedben]

This is the basis of my entire argument and exactly why I believe that derating table 220.84 should not be allowed in this 4W common riser design.
Your rebuttal on my 100amp-3W apartment example and 80A-4W riser assumed I arrived at the apartment load via the optional calc and if true your 4W riser size would be much larger than 80amps.

I came back with a 99amp apartment design, which would not be allowed to use 220.82, but the apartment 3W "Feeder Tap" remains 100amp.
The math to size the 4W-riser still results in an 80A-4W common riser, thereby smaller in phase and N than the TAPS it serves.

I see what you're saying and agree that this is something of a fluky result, but bottom line is that your 80A riser feeder will be protected and made safe by an 80A overcurrent protection device, and it's unlikely that it will trip. What is really the case is not that the 80A 4W feeder is undersized, but that each dwelling feeder is oversized due to the fluky difference between the regular and option load calculations. If there is a reason that the riser OCPD shouldn't be smaller than the feeder tap OCPDs, then I think that would be due to coordination requirements or something other than load calculations.

Staying on when one is allowed access to the derating table...
The text of 220.84(A)(1) states .... no dwelling served by more than one feeder.
The connection to the 4W-riser is a feeder-tap

Staying on when one is allowed access to the optional calc for the individual dwelling unit (Please don't confuse this with the 99amp referenced above)
The text of 220.82(A) states single set 3W feeder
The connection to the 4W-riser is a feeder-tap

Each apartment is served by a single 3W feeder. The fact that the 3W feeders are connected in series with another feeder upstream is irrelevant to this, and it is irrelevant whether they are tapped from a feeder or fed from a panelboard that is fed by a feeder. There are numerous places in the code where the number of supplies is restricted to one feeder or one branch circuit or some such - for example separate buildings - and it would make no sense if those feeders couldn't be supplied by connection in series to another feeder. For example, a building with a service could have a feeder to a subpanel, which contains a breaker to a single feeder supplying a separate building. No one would ever say that this violates the restriction to serve the separate building with a single feeder.

Can one replace the text AS written in both of these locations and infer that feeder can be replaced with feeder-tap ?
Knowing these loads are delta on a Y system and increase in size by a factor of 1.73 once reflected onto the 4W riser.

Yes. Now that you put it fully in context I would say that, yes, in those two sections you could replace 'feeder' with 'feeder tap' and the load calculations therein are still permitted. I would not say that as general rule throughout the code, but for the purposes of these two sections it does not matter if the feeder serving the dwelling(s) originates at a tap or not.

Not withstanding that line-line loads might result in 1.73 times the current on one of the 4W conductors as compared to any of the 3W conductors, that is what 220.84 plainly allows. Again, the riser feeder needs to be protected by an appropriate OCPD, regardless.

 

[iwire]

Say you have a main service panel which supplies four subpanels and those panels in turn supply 3 apartments each. MyCleveland appears to be looking at this as each apartment is being supplied by two feeders, 1) from service panel to subpanel, and 2) from subpanel to apartment panel. This is not what is meant by the requirement the dwelling unit be supplied by not more than one feeder, but from this perspective, it is being supplied by two feeders... one after the other.

Per the definition of feeder I believe we can very well call that arrangement a feeder.

I don't see anything in the definition that stops a feeder at a breaker that also supplies a feeder.

Food for thought, take it for what it is.