Lets say you run 12/4 w/ground and use 3 as hots, 1 as neutral and 1 as ground. The three hots all go to 20 amp breakers. Two of the three circuits will feed out to my kitchen countertop required small appliance branch circuits (GFI protected). The other circuit will be a dedicated receptacle for a plug and cord garbage disposal. At what point do I know that I am overloading the neutral? What are the rules about how many circuits you can run off of one neutral? Any help would be appreciated.

What is the system voltage in this kitchen you are refering to?

If this is not 3 legs of a 3p wye it is a violation and you will be overloading the neutral.

Art 100 definition of multiwire;
Branch Circuit, Multiwire. A branch circuit that consists of two or more ungrounded conductors that have a voltage between them, and a grounded conductor that has equal voltage between it and each ungrounded conductor of the circuit and that is connected to the neutral or grounded conductor of the system.

If this is not 3 legs of a 3p wye it is a violation and you will be overloading the neutral.
Art 100 definition of multiwire;
Branch Circuit, Multiwire. A branch circuit that consists of two or more ungrounded conductors that have a voltage between them, and a grounded conductor that has equal voltage between it and each ungrounded conductor of the circuit and that is connected to the neutral or grounded conductor of the system.

Perhaps I'm just dense, but how does that definition make it illegal if it is not 3 legs of a wye. If you used a #10 gc and 3 #14s from any 240 V source; IMO it would be legal and satisfy the NEC.

Dlhoule, in your example it would be legal. The #10 would carry the maximum amperage of the two # 14 ungrounded conductors on the same leg.

Roger

I don't agree with the example of using a #10 and 3 #14's for a few reasons:

1. Each ungrounded conductor must be a different color- which is not clearified in the example.
2. If it is a multi-wire branch circuit by virture of the fact that it is using 2 ungrounded conductiors that have a voltage between them per the definition, then I would not approve the additional ungrounded conductor tagging along on the grounded conductor of the multiwire branch circuit.
3. I am assuming that "#10 gc" means grounded conductor.
4. 210.4 (C) makes it to debateable and on top of that ir's wierd.

George than I have to ask if you think this is a case where the handbook is wrong?


Exhibit 225.3 A 120/240-volt, single-phase, 3-wire system (branch circuits rated at 20 amperes; maximum unbalanced current of 80 amperes).

Bob

[This message has been edited by iwire (edited 05-03-2006).]

The diagram posted by Iwire comes from Article 225- Outside Branch Circuits and Feeders and is permitted outdoors (225.7(B)), but would not be permitted indoors as far as I can tell since there is no similar language in Article 210 Branch Circuits.

True Tom, but where is the language that prevents it?

What code section can be cited to fail it and why should there be one.

It is as safe as a feeder.

Bob

You are correct Bob. That is a good example. When I made that post I somehow knew you'd bring that up. I still say it's a poor design and an installer who installs this way is just asking for trouble. I can't give him a violation but I can first of all try and talk him out of it due to the potential for confusinon down the road and if that don't work you can bet your beppi he will be submitting an asbuilt showing all the intimate details of the cicuiting, wire numbering, loading and not just a skimpy sketch on piece of scratch paper either.

I have never seen anyone install anything even remotely close to that in real life. I've been in the trad for 40 years and an inspector for over 20.

I take that back- I did have a guy one time install some festoon lighting where he had a common neutral.

I think that we discussed this one previously.

Article 225 _Outside_ Branch Circuits and Feeders clearly permits 'shared neutral' circuits for lighting ( 225.7(B) )

Common neutral circuits are also described for _feeders_.

Article 210 does not mention 'shared neutral' circuits. 210.4 defines and permits multiwire circuits, and as described a multiwire circuit is a specific type of shared neutral circuit. But as far as I can tell nothing in article 210 either prohibits or permits shared neutral circuits.

-Jon

George I have never seen it or considered it.

Beside the fact I can think of very few ways to make it economically attractive I just don't like it for it's 'weirdness' factor.

Bob

You might be able to call that shared neutral circuit something but it isn't a "multiwire circuit".
I posed a similar question a while ago when I asked if you could have a remote junction box with a number of branch circuits and a common neutral, sized by 250.66 to be on the safe side. (As in a service upgrade where the conductors didn't reach the new panel)
Answers were mixed.

I had a proposal on the issue of common neutrals that was rejected. My proposal was to delete the wording shown below in bold.

215.4 Feeders with Common Neutral
(A) Feeders with Common Neutral Two or three sets of 3-wire feeders or two sets of 4-wire or 5-wire feeders shall be permitted to utilize a common neutral.
(B) In Metal Raceway or Enclosure Where installed in a metal raceway or other metal enclosure, all conductors of all feeders using a common neutral shall be enclosed within the same raceway or other enclosure as required in 300.20.

4. Statement of Problem and Substantiation for Proposal:

There is no code violation in using a common neutral so the specific provision to use one is not required

I had a number of proposals rejected that tend to overturn the idea that if the code doesn't say you can't, then you can.
Don



[This message has been edited by resqcapt19 (edited 05-03-2006).]

In a single phase panel you can share 2 circuits, and 3 phase 3 circuits as long as the power is on separate phases. What ever the difference is in amps between the 2 phases on a single phase, that difference will be what is on the neutral

My question refers to a 120/240 VAC Single phase panel. The kitchen circuits would be the 2 small appliance circuits that require GFCI protection above the countertops and then a feed with the 3rd ungrounded conductor to the cord connected garbage disposal. I'm just wondering where you draw the line on sharing the neutral since the receptacles would be getting use by small appliances (with motors) and the garbage disposal alone can pull up to 7 to 10 amps. What is the general rule of thumb to make sure if you run multiwire branch ciruits to different parts of a house that you do not risk a fire by overloading the neutral. Lets say one receptacle has a toaster plugged in and it is running at 10 amps and then a blender is running at 2 amps and then the garbage disposal gets turned on (another 10 amps). Well the circuit breakers are not going to trip because you have not exceeded the maximum amperage on each ungrounded conductor. But is the neutral then carrying a total of 22 amps?

Kyelectric, the problem is not what you know will be plugged in, it is what might be plugged in.

Take the two ungrounded conductors on LEG #1 and LEG #2 and load them to 20 amps each, the neutral will carry 0 amps.

Now take and add another conductor to LEG #1 and load it to 20 amps.

All three conductors are now loaded to 20 amps.

One conductor on LEG #1 will cancel with the conductor on LEG #2 which will leave the neutral carrying 20 amps. (the current on the other conductor connected to LEG #1)

Now, if the cicuit feeding the conductor on LEG #2 is turned off you no longer have a MULTI-WIRE BRANCH CIRCUIT, and the grounded conductor will no longer be a neutral, it will simply be a grounded conductor and will be carrying the total of the two ungrounded conductors on LEG #1, which is 40 amps and your # 12 would be overloaded.

Roger

Your grounded conductor will only carry the difference in current off of your ungrounded conductors as Roger just explained. A true neutral will carry no current...I find definite overheating potential in this situation and the only code reference I could give you for my reason for thinking this is a violation is once again article 100 definition of multiwire branch circuit. This definition clearly states that each ungrounded conductor will have the same voltage to the grounded conductor, and that there will be a voltage between each of the ungrounded conductors. If you run two conductors off of the same phase (which is what you are talking of doing in a single phase system) they will not have a voltage between them and that is a violation.

kyelectric,

The _general rule_ for neutrals is that the ampacity of the neutral shall not be less than the _maximum net computed load current between the neutral and _all_ ungrounded conductors connected to any one phase of the circuit.

As applied to 'multi-wire branch circuits', you simply have one, and _only_ one hot conductor per supply leg, and one neutral of the same ampacity as the hots, and call it good. In a multi-wire branch circuit, the maximum current on the neutral is equal to the maximum current on any _single_ hot conductor. This means that in a single phase system, you can run 12/3, put black on one 20A circuit, red on another 20A circuit _on the other supply bus_, giving a _maximum_ of 20A on the neutral.

Note that you are _required_ to supply the hots from different supply legs. Use the exact same cables, and the exact same circuit breakers, but put both breakers on the same supply leg, and now you can see as much as 40A on the neutral.

Further discussion, sort of a summary of the above:

A 'common neutral' circuit is one in which you have any number of ungrounded conductors sharing a single grounded conductor.

A multi-wire branch circuit is a specific case of a common neutral circuit. In a multi-wire branch circuit, you may only have _one_ ungrounded conductor per supply leg; in a single phase system a multi-wire branch circuit would have two ungrounded conductors from opposite supply legs; in a three phase system you could have either two or three ungrounded conductors from two or three of the phases.

Common neutral circuits are mentioned explicitly in article 225 for outside circuits. Multi-wire branch circuits are mentioned explicitly in article 210 for branch circuits. Multi-wire branch circuits are in common use; the more general 'common neutral' circuit is virtually unheard of, and arguably are _not_ code compliant for inside wiring. (There is also a strong argument that they are permitted since they are not prohibited.)

iwire: I think that I see a situation where using a common neutral circuit might make sense: getting around 240.4(D) and using the full thermal ampacity of small conductors. Consider a 120/240V single phase system. Say I have two fixed loads totalling 24A (say a dishwasher and a microwave). I claim that I could supply these loads and a 20A receptacle circuit using 12/4. The two fixed loads would each be placed on 20A breakers on supply leg A, using the red and blue conductors. The receptacle would be on a 20A breaker on supply leg B using the black conductor. The neutral would be the white conductor. The maximum unbalanced current on the neutral would be 24A, all conductors would be protected by 20A breakers. If the fixed loads were 12A each, then you could do this with a bundled 14/2 and 12/2 and 15A breakers...if you want to get _reallllllllly_ ugly yet remain code compliant

-Jon

Winnie,
Thank you so much for your article. It was very helpful. I do have a better understanding now than I did before. So, basically, when I use multiwire branch circuits on a 120/240v system, I need to make sure the two ungrounded conductors are on separte phases so the neutral current will be cancelled. Can you give me examples where you would commonly use multiwire branch circuits in wiring a house? Like for instance the kitchen two small appliance branch circuits. Thanks!

I never understood the logic of trying to take this potentially dangerous shortcut. Code or no code, the neutral needs to handle the maximum potential unbalanced current. Period. In the case of two 20A circuits, that's 40A, in which case the neutral needs to be #8 and that's not what you get with 12/4.

Think about it... it's your license. And if something goes wrong, guess who the lawyers point the finger at? And they're not gonna want to hear you've done it that way for years and there never was a problem. It's the potenital they'll nail you on. And if you say there's no possibility of that happening, you'll then be labeled as incompetent.

What is so difficult about running dedicated neutrals, or a correctly sized neutral in the case of conduit? I understand copper is getting expensive, but since when is expense an excuse for shortcuts?

Run a dedicated neutral and this way you'll have nothing to worry about.

Joe

"I never understood the logic of trying to take this potentially dangerous shortcut. Code or no code, the neutral needs to handle the maximum potential unbalanced current. Period. In the case of two 20A circuits, that's 40A, in which case the neutral needs to be #8 and that's not what you get with 12/4."

If your (2)20A circuits are on opposite phases you would never have more than 20A on the neutral. The neutral will only see the difference between the (2) loads, not the combined total. Circuit A= 15A + circuit B=10A the neutral would carry 5A. Nothing dangerous about it.

It depends on the job.

The types of jobs I work at generally have homeruns in the hundreds of feet and many times 100s of circuits.

Using correctly wired multiwire branch circuits does the following for me.

1)Use less matrial, copper, cables and EMT.

2)Reduce the number of current carrying conductors in a raceway.

3)Significantly reduce voltage drop.

4)Reduce the number of terminations.

There is also plenty of office furniture (office cubes) that come pre-wired with multiwire branch circuits. The only way to code compliantly supply an existing multiwire branch circuit is with a multiwire branch circuit.

It is 'my license' and if I can not safely wire a multiwire branch circuit I have no business having a license.

Bob

For a visual explanation of a single phase Multi-Wire Branch Circuit see the follwing graphics.

A Neutral carrying the current of an imbalanced Multi-Wire Branch Circuit.



A true Neutral



Additive current from one leg



Roger

Of course there's no problem if they're on opposite phases. Question is, will it remain intact?

For example, NYC code requires #12 to be used on 15A lighting circuits. Does that make sense, when #14 is fine? The only reason I can see would be the possibility of Harry Homeowner or Harry Handyman (who it should be noted can't do any electrical work in NYC) by the substituting a 20A breaker into the lighting circuit, or a wiring change where part of the lighting circuit gets on a 20A feed.

I don't know if this is the actual reason, just my educated guess. It just doesn't make sense, if #14 can carry 15A, why is #12 mandated? (And it's not like #10 is mandated on 20A, 20A #12 is okay in NYC -- length of run aside.)

Same principle can be applied to multi-wire with shared neutral. Panel is tight on breakers, and the home owner decides to finish the basement, either himself or with a GC who does "everything". Hmmm, let me put some tandem breakers in and move this breaker (the one for the multi-wire circuit) over here (onto the other phase). Now if that multi-wire circuit is overloaded, it's possible the fire department will be stopping by.

Also, with a shared neutral, you now have 240V in a box somewhere. So let's say someone piggybacks a new 120V circuit on the two hot legs, then plugs something in this new circuit. I don't think those books at the orange box mention anything about multi-wire circuits.

I guess I should've clarified the screw-up factor. I know the folks here know all about multi-wire requirements, and would do it correctly. But how often does your original nice, neat work stay intact over the life of the installation? Something always seems to get hacked up.

Now I'm not one of those folks that believes we need to protect folks from themselves, but not sharing neutrals sure can eliminate a lot of potential future problems.

In view of the foregoing, I still say ideally it's best to run a dedicated neutral.

Just my 2 cents worth...

Joe

I like using multi-wire branch circuits because it can save time and money during the installation. The fella's on this board sold me on the idea a long time ago. Once you understand about the unbalanced load on a nuetral on a 3-wire circuit -- including proper installation of the OCPD's -- IMO you are good to go.

In my view it is entirely depends on the job a hand.

I am willing to bet you would look at this differently if your job was pipe and wire with multiple homeruns that are a couple hundred feet long.

All these big box stores springing up do not have panels in every aisle.

You are adding significant costs to a job like this using dedicated neutrals.

JMO, Bob

Although I have no intentions of doing this, when this subject came up before I posted this, taking Bob's Handbook chart a level further.





Can anyone come up with a Code reference to "shoot it down"?

Okay, Bob you got me, LOL! It would be ridiculous to use dedicated neutrals in pipe; one properly sized neutral would be the way to go. What a PITA it would be trying to track down which neutral goes to where.

Electure, that second illustration from Bob's Handbook with the incremental increase in neutral sizing works for me. The wire is properly sized for any potential unbalanced current throughout -- though you might have trouble selling that to a "difficult"
inspector, who might insist on #4 straight through.

Of course, there's still the GFCI and finicky computer equipment issues with shared neutrals -- which is one of the reasons I prefer dedicated neutrals -- but that's another issue, and I don't want to go back and beat that dead horse.

I gotta say, you guys ARE good!

Joe

Scott, no code referance to shoot it down, but the T&M factor could very well work against you.....

I ain't gonna say nothin' !!

Bob (Iwire) knows how I feel about this topic.

Tony, there's no need to be bashful, please, voice your opinion.

Roger

O.k, but remember you asked for it!! LOL....

First off, may I direct everyone's attention here:

https://www.electrical-contractor.net/ubb/Forum2/HTML/002445-2.html

(Third post down on that page)

'nuff said there, pretty much sums up my feelings about this whole thing.

As for the original post on this thread, I quote:

Bold emphasis mine...




As I read this, if this is a single phase residential panel, it would be impossible for all three hots to be on opposite legs (or phases if you prefer) in this application. Thus the certain possibility or neutral overload exists!!!

And isn't 12/4 NM configured as two hots (red/black) two neutrals (white/white with tracer) and ground?

Finally, we have the original poster, an experienced electrician as near as I can tell, about to set himself up with a possible Code violation(if the cable/conductors are not approved to be re-identified as in this application) at least and a fire hazard at worst? (Not trying to put down anyone here, either.)

As in the rant I referenced in the link, even those experienced in the field can be tripped up by the use or mis-use of multi-wire or shared neutrals.

And while individual experiences vary, most of mine with multi-wire have been to fix trouble caused in one way or another by the lack of dedicated neutrals. (I won't enumerate them here, but I have made several posts in different threads re this subject.)

In fact, the original wiring to my apt. was only TWO circuits, one 15 amp lighting circuit and one 20 amp circuit for ALL of the non-lighting loads!! It was served by an very old style of romex with (I'm not making this up) one #12 hot, one #14 hot, one #12 neutral and one bare #16 ground.

That got replaced with all new wiring after the original neutral connection at the split to the respective loads (in a 4s box in a 2nd bedroom receptacle) began to fail with my $5,000 worth of A/V gear energized. I caught it first with the voltmeter in the rack, then my Fluke 36 (after things had cooled down a bit) showed over 158 volts to the recept.

Now all circuits have their own neutral all the way back to my nice new SqD panel. And not a lick of trouble since.

You did ask...

edited to fix html...

[This message has been edited by mxslick (edited 05-10-2006).]

That's all well and good until the service neutral opens up.

Peter

Seems to me that every argument against multiwire branch circuits stems from either the improper installation or improper modification of the circuit.

I troubleshoot lots of branch circuit problems that are neutral related. They are usually caused by some numbskull stuffing 6-#12 into a red wirenut or something like that.
I have NEVER run into a case of a problem caused because there wasn't a dedicated neutral run for each circuit.

This stuff of installing the circuit so a DIY can work on it is not shared by any other trades or manufacturers. (Look under the hood of your truck, for instance). A professional should be called out if the DIY lacks the competence to work on the system.




[This message has been edited by electure (edited 05-14-2006).]

Lol, Peter. that's a good point. Especially in my case, as the service drop is the original one serving a total of 5 apartment units. (With the original "A" base style meters.) I'll try to get some pictures of this mess to put up here, I can call it "Little Service of Horrors."

Here's an interesting hypothetical thought: If the service neutral (ahead of all 5 services) were to open, would the load diversity of all the units minimize the possible imbalance?


and:


Bingo! That seems to reflect most folks' negative experiences, but:


In my business, that is the primary cause of trouble for both audio and control systems as I have mentioned in other threads. And in the case of multi-channel (circuit or load if you prefer) lighting systems like Panja, Vantage and Crestron as used in McMansions, the use of shared neutrals causes a lot of problems and in fact most of the manufacturers of that equipment will void warranty on equipment installed with multiwire feeds. As I do on any audio equipment not wired to the specifications I provide. Again, it all depends on the intended application.

I will probably shock a few by saying that I can see some applications where I personally would install multiwire, for example, circuits serving motor loads without VFD's; a lighting load consisting primarily of incandescent lighting; dedicated circuits for seasonal lighting......I'll probably think of a few more later.

As it stands, it really boils down to a cost vs. performance stance. In my case, the extra cost of the extra #12 white THHN (Around $50.00 for a 20-screen movie theatre) is more that repaid by the time and trouble saved in chasing out audio noise and erratic automation operation which I have encountered in installations without dedicated neutrals. Since a service call to a cinema is worth $65-$95 per hour with a four hour minimum, I would rather spend the extra up front (which by the way is factored into my installation bid)than have to do a gratis warranty service call.

If I can add a point to the case for multiwire circuits.

Recently one of our electricans wired a large clothing store which had tons of incedescent recessed lighting. Upon completion we received a call that a panel was overheating. Our service man reported back that everything seems kosher with the conductor sizing and connections, yet he said the panel was extreamly hot. The panel controled all the stores lighting and was a 200 amp 3 phase, 42 circuit panelboard . All 42 circuits were fully loaded, yet with in allowed tolerances and all the conductors were very neatly tie wraped.
I was called in at this point to attend a meeting with the local inspector, engineer, and all others involved. While we were looking at this panel I noticed what seemed like an excessive amount of netural conductors. 42 to be exact.
I asked our electrican why seperate neturals ?
His response was because the engineer called for it on plans. Well common sense here, if all the circuits are loaded to 16 amps, then all the neturals are loaded to 16 amps. Which seemed like alot of fully loaded conductors tiewrapped together in the wireways of the panelboard.
I explained my theroy to the project engineer and inspector and they both agreed the amount of fully loaded conductors must be causing the excessive heat and that shared neturals would reduce the loaded conductors thereby reduce the heat build up.
Before the day was out, I had a signed change order to remove 2/3's of the neturals in the panelboard. Once the was done, no more heat. Problem solved.
So Seperate neturals do not always mean a better instalation.

mxslick,
How does a correctly installed mulitwire branch circuit cause a problem? And if it does, why isn't the service which is always a multiwire circuit a problem?
Don

Hi Don, since you asked, I'll first link a couple of threads where I had given my answers to that question:
https://www.electrical-contractor.net/ubb/Forum2/HTML/002445.html
https://www.electrical-contractor.net/ubb/Forum2/HTML/001963.html


And now for some cut'n'paste from those threads to summarize:

1: The impedance of the neutrals at feeders and panel buss are (electrically) significantly lower than the #12 or #10 branch circuit wiring thus allowing any noise or harmonics to be "absorbed" (Not exactly the right word I'm wanting, but you get the idea);
2: ANY electronic load of ANY kind is far more sensitive to power quality issues than incandescant lighting or pure resistive loads like heating devices;
3: Electronic dimming controls of any make or style are VERY noisy (just take a portable AM radio near any of those things and you'll see what I mean.);
4: Most of these systems are extremely expensive and sophisticated in their control schemes and do not take kindly to noise issues on incoming power (not to mention the risks of a lifted neutral frying everything)

Finally, if the manufacturer says don't do it, then don't do it! There are reasons for such instructions.

(The above was in reference to Crestron multi-zone lighting dimmers)

and:

....but a few years back I did a refit on a screening room where the audio racks were serviced by 5 20amp breakers (on a 3-phase panel) with ONE #10 neutral for all 5 circuits!!! The client was wondering why the sound system had a massive hum and why he'd been replacing burned-out amplifiers every few months.
There were NO loose/bad connections, which was my first thought.

After I had the EC pull in the proper number of neutrals and wire the rack recepts correctly, the system was quiet and no more burning amps.

It has been proven many times in my line of work that a shared neutral causes noise issues. Many installs I've had to rework have been cured of noise by pulling dedicated neutrals for each circuit. I don't always understand why that's so, but it has proven itself many times.

(edit) To (re)quote iwire in one thread:

quote:Electrically sharing a neutral on a branch circuit is not any different than what happens at the panel where all the branch circuit neutrals share a single conductor back to the source.

My response was:

For non-power-sensitive loads, I agree. For audio gear, no. It's the same idea as using isolated grounds, IMHO.

But in summary, I can state that first off, I'm not an EE, and thus don't have the affirmative background in electrical theory to back up my position, BUT, I do have the 20+ years of experience in cinema and high-end audio where my position has been proven correct repeatedly. So my best guess at why the elimination of shared neutrals has cleared up many problems I have encountered is probably best explained by Item #1 in my cut'n'paste.

And as I had mentioned in one of the posts in those threads, statistically the percentage of trouble calls I have attended caused by shared neutral issues is quite low, well under 10%.

I think we can all agree that electricity doesn't always behave in the manner we wnat or expect.

Sorry for the long post....

see edit above for quote credit...


[This message has been edited by mxslick (edited 06-06-2006).]

And isolated grounds have three possible outcomes that occurr about equally...they help, they do nothing or the make the problem worse.
Don

mxslick,
Actually I don't get the idea. Harmonics are not absorbed by any conductor or the system.
That is a design problem of the equipment, not a problem caused by the multiwire branch circuit.
And how does the use of multiple neutrals help a RF problem where the noise is transmitted through the air?
Again, other than the open neutral issue, that is an equipment design problem, not a neutral problem. Most of the power quality problems that effect this type of equipment are generated by this type of equipment. Why should the electrical system be changed to try to prevent problems that are caused by poor equipment design?

Do you have any technical documentation that shows problems for "sensitive" electronic equipment are caused by multiwire branch circuits?
Don

[This message has been edited by resqcapt19 (edited 06-07-2006).]

Yup, it happened again...

Yesterday, doing an institutional re-model, I killed all three circuits in the box. I then opened the neutral- and saw a spark.

The neutral measured 50 volts to ground. It was, on the "load" side, definitely hot. We later found a number of fried surge protectors. The explanation: somehow another two circuits had been tied into that neutral. When I opened the connection, the voltage on the circuits became 208.

Properly done, and maintained, sharing a neutral can be a great technique. Give time, and Mr. Murphy, a chance however, and you've got troubles.

As I've said before, we have got to start identifying our neutrals. We also have to be wary of partly trained personnel trying to take this nifty "shortcut."

Finally, contrary to popular belief, the "white" wire is quite often "hot," and you might never find the way to shut it off completely.

John,
A change requiring that was approved at the proposal stage for the 2008 code.
Don