No disconnect at the meter

[Jeffrey L. Mathis]

I may be confused. I have always understood that in recent construction, there should be a main disconnect at the meter when the panel is anywhere but behind the meter. The house I did today is 5 years old and the main distribution panel is in the garage and looking fine, but the meter is on the opposite end of the house. Should there not be a disconnect at the meter?
JLMathis

[Jerry Peck]

I may be confused. I have always understood that in recent construction, there should be a main disconnect at the meter when the panel is anywhere but behind the meter. The house I did today is 5 years old and the main distribution panel is in the garage and looking fine, but the meter is on the opposite end of the house. Should there not be a disconnect at the meter?
JLMathis

First, the disconnect is not required at the meter.

Second, though, this sounds like it could be a problem: "the meter is on the opposite end of the house". I say "could be" because the conductors which run from the meter to the disconnect are the "service entrance conductors" and they are not allowed to enter the structure unless the disconnect is as close to the point of entrance as possible, i.e., typically that means through a wall with the meter on one side of the wall and the disconnect on the other side of that same wall ... and that's about it.

Now, you also need to consider where the "point of entrance" of the conductors is at. The conductors are considered as still being "outside the structure" if underneath the house and under a minimum of 2" of concrete, and most slabs are 4" nominal, so that is not a problem (but would be a problem if the conductors were run through a crawlspace, however, if the conductors were buried at least 18" below the crawlspace floor, the conductors would also be considered as outside the structure, the problem is getting from that 18" down to up into the structure).

The conductors would still be considered outside the structure if run through the structure in a raceway which is encased in concrete or brick at least 2" thick.

And if the conductors are installed in a vault meeting the requirements for it - unlikely to find in a house.

The question is now: How and where are the conductor run from the meter to the disconnect?

[Jeffrey L. Mathis]

The sheathed cable runs in the crawl space and is secured to the bottom of the joists. This is the way it's always done here. Always has been. Never gave it another thought. But I normally expect to see a disconnect at the meter.
jlm

[Jerry Peck]

The sheathed cable runs in the crawl space and is secured to the bottom of the joists. This is the way it's always done here. Always has been. Never gave it another thought. But I normally expect to see a disconnect at the meter.
jlm

With a disconnect at the meter, that cable becomes feeder conductors and that is allowed.

With no disconnect at the meter, well, that plain ain't allowed. :)

[Jeffrey L. Mathis]

I'm reading you clearer as I go. So adding a disconnect at the meter changes things? No difference in the composition of the cable, but just adding the disconnect changes it from a service entrance to a supply??
jlm

[Jeffrey L. Mathis]

Thanks.
30 years as a framer. Not so as an electrician.
jlm

[Jerry Peck]

I'm reading you clearer as I go. So adding a disconnect at the meter changes things? No difference in the composition of the cable, but just adding the disconnect changes it from a service entrance to a supply??
jlm

Yep.

That disconnect at the meter makes all the difference in the world for an installation like that.

Think of it this way:
- When there is a disconnect at the meter, that cable is protected with a properly sized overcurrent device, right?
- When there is no disconnect at the meter, that cable is protected by the overcurrent device back at the power company transformer - ever see a transformer explode and burn?

Utility pole transformer explodes & catches fire - YouTube (http://www.youtube.com/watch?v=vOfcntat7cE)

Exploding Power Pole - YouTube (http://www.youtube.com/watch?v=VJn1-bpYi0Y&feature=related)

Utility Pole Burns, Explodes in Wildwood Crest NJ - YouTube (http://www.youtube.com/watch?v=-oZQ8d9BWOI&feature=related)

[cuba_pete]

Yep.

That disconnect at the meter makes all the difference in the world for an installation like that.

I've got to say that I haven't seen such a good post about what could be a very subjective topic in a long while...on any forum. Good info Jerry.

[Bill Kriegh]

An over current device device at the meter is always a disconnect, but a disconnect at a meter isn't always an over current device.

[Jerry Peck]

An over current device device at the meter is always a disconnect, but a disconnect at a meter isn't always an over current device.

True, I should have said disconnect and overcurrent device at the meter.

One could make an overcurrent device which is not a disconnect, however, I do not recall having seen one. The overcurrent device is allowed by code to be part of the disconnect or to be immediately adjacent thereto. Thus one could have an unfused knife switch and an unswitched fuse holder (i.e., not a pullout either) - but I've never seen one.

[Frank Bombardiere]

So if there was an un-fused knife switch at the meter, would the downstream panel have to have the neutrals isolated or is that only if there is an overcurrent device upstream of the panel?

[Benjamin Thompson]

Check with the local power company, some require an exterior disconnect and won't turn on the power if its not there.

[Jerry Peck]

So if there was an un-fused knife switch at the meter, would the downstream panel have to have the neutrals isolated or is that only if there is an overcurrent device upstream of the panel?

From the 2008 NEC:
- 230.91 Location.
- - The service overcurrent device shall be an integral part of the service disconnecting means or shall be located immediately adjacent thereto.

"So if there was an un-fused knife switch at the meter" and no overcurrent device until you get to a downstream panel ... and if that downstream panel is not "located immediately adjacent" to the un-fused knife switch ... then there is (theoretically speaking) no main service overcurrent device, the main in the downstream panel would be a 'panel main' device but would not be considered to be the 'main service overcurrent' device.

- VI. Service Equipment — Disconnecting Means
- - 230.70 General.
- - - Means shall be provided to disconnect all conductors in a building or other structure from the service-entrance conductors.
- - - (A) Location. The service disconnecting means shall be installed in accordance with 230.70(A)(1), (A)(2), and (A)(3).
- - - - (1) Readily Accessible Location. The service disconnecting means shall be installed at a readily accessible location either outside of a building or structure or inside nearest the point of entrance of the service conductors.
- - - - (2) Bathrooms. Service disconnecting means shall not be installed in bathrooms.
- - - - (3) Remote Control. Where a remote control device(s) is used to actuate the service disconnecting means, the service disconnecting means shall be located in accordance with 230.70(A)(1).

[Jerry Peck]

Jerry,
Why would the panel need to located adjacent to the knife switch?

Robert,

You did not read that well ... *the panel* is not required to be adjacent to the un-fused knife switch ... *the main service overcurrent* device is required to be adjacent to the un-fused knife switch.


Are you assuming that the conductors between the knife switch and the panel are not on the outside of the building?

I'm not presuming anything of the like.

I am only presuming that the overcurrent device is not located at the un-fused knife switch (obviously, the knife switch is "un-fused") and that the panel with the 'panel main' in it is not "located immediately adjacent thereto".

If the panel with the 'panel main' in it was "located immediately adjacent thereto", whether inside or outside, then the two could be considered to make up the service equipment.

[Jerry Peck]

Maybe you could explain what this means and why the panel main with OCP is not the service disconnect.

Robert,

Did you read the two NEC code sections I posted? One addresses the main service OVERCURRENT device and its location. The other addresses the main service DISCONNECT device and its location.

The main service OVERCURRENT device and the main service DISCONNECT are not required to be one and the same or even in the same enclosure.

However, if the main service OVERCURRENT device is not located in, and is not part of, the main service DISCONNECT device, then the main service OVERCURRENT device is required to be "located immediately adjacent thereto" in relation to the main service DISCONNECT device.

If the panel with the first overcurrent device is not "located immediately adjacent thereto" in relation to the main service disconnect, then the overcurrent device in that panel does not meet the requirements of the main service disconnect and is, therefore, NOT the main service overcurrent device.

Do you agree?

If not, explain why you disagree.

[Rollie Meyers]

If a non-fusible disco was used, the conductors would still have no overcurrent protection, & the install will still be non-compliant. overcurrent protection needs to be provided nearby where they enter the building, some have defined that as 5 feet, but the 5 foot limitation is not in the NEC.

In PG&E territory, SE cable is not listed in PG&E's "Greenbook" so SE cable has not been used for many years, so it is a strange beast to see for me.

[Jerry Peck]

If a non-fusible disco was used, the conductors would still have no overcurrent protection, & the install will still be non-compliant. overcurrent protection needs to be provided nearby where they enter the building, some have defined that as 5 feet, but the 5 foot limitation is not in the NEC.

In PG&E territory, SE cable is not listed in PG&E's "Greenbook" so SE cable has not been used for many years, so it is a strange beast to see for me.

Rollie,

That is a different issue, with the 5 feet from the point of entrance into the building.

If the main service disconnect is un-fused, then the overcurrent device is required to be "located immediately adjacent thereto" - there is not 5 foot allowance there, the distance is "immediately adjacent thereto", which would indicate a chase nipple with two lock nuts between to create just enough space to allow the two covers to be opened and closed without interfering with each other, but no more space than that.

[Jerry Peck]

Uh oh, Jerry just gave us the NEC definition of adjacent to. :)

Robert,

It's not the "adjacent to" part ... ;) ... it's the "located immediately" ... "adjacent thereto" part.

1 im·me·di·ate·ly adverb \i-ˈmē-dē-ət-lē also -ˈmē-dit-, British often -ˈmē-jit-\

Definition of IMMEDIATELY
1
: in direct connection or relation : directly <the parties immediately involved in the case> <the house immediately beyond this one>
2
: without interval of time : straightway <I'll make that call immediately>
See immediately defined for English-language learners »
See immediately defined for kids »
Examples of IMMEDIATELY
: the person immediately to my left
: Dinner was served immediately after the ceremony.
: The new law will become effective immediately.

[Garry Blankenship]

Interesting subject. I spent many years assuming the NEC allowed a maximum of 15' of unfused conductors inside a building. Recently challenged to find that, I discovered it is a Washington State requirement, ( my home turf ), not NEC. I was unable to find any NEC specific limit to how much unfused wire was allowed inside a building - - - disconcerting. "inside nearest the point of entrance" is subject to inturpretation. Is nearest 5', 15', 50' ? Is under the framed floor considered inside the structure ? In the ground or covered by 2" of code approved material, ( concrete basically ), is not considered "inside the building". Bill Kreigh is spot on regarding disconnect verses overcurrent device. I once installed a 1,600 amp enclosed circuit breaker w/o overcurrent protection, ( factory slugged basically ), to comply with a requirement for a land side disconnect. I do not even know, if 1,600 amp knife switches are made, but the size would be mamouth. This saved the added cost of carrying ground wires in the roughly 200' x 4 of parallel feeders from that disconnect to the service location. I believe the goal of the code is to provide reliable access for emergencies / emergency personell, ( fire fighters ), and the NEC, as written, falls short of that.

[Jerry Peck]

I was commenting because your definition makes an assumption as to what the CMP was thinking when they used the words immediately adjacent to. I see your point but I disagree with this assertion that an offset nipple would not be code compliant.

I'm not saying that an offset nipple would be non-compliant as that may be required to align the knockouts in the two enclosures, but I am saying that a chase nipple would be considered as being compliant by everyone. The offset nipple would probably also be considered as compliant by everyone, or nearly everyone.

An un-fused knife switch on the exterior and the overcurrent device on the interior would not, I believe, be considered as compliant by anyone. (At least not anyone with any sense and reading comprehension. ;) )

[Jerry Peck]

Interesting subject. I spent many years assuming the NEC allowed a maximum of 15' of unfused conductors inside a building. Recently challenged to find that, I discovered it is a Washington State requirement, ( my home turf ), not NEC. I was unable to find any NEC specific limit to how much unfused wire was allowed inside a building - - - disconcerting. "inside nearest the point of entrance" is subject to inturpretation. Is nearest 5', 15', 50' ? Is under the framed floor considered inside the structure ? In the ground or covered by 2" of code approved material, ( concrete basically ), is not considered "inside the building". Bill Kreigh is spot on regarding disconnect verses overcurrent device. I once installed a 1,600 amp enclosed circuit breaker w/o overcurrent protection, ( factory slugged basically ), to comply with a requirement for a land side disconnect. I do not even know, if 1,600 amp knife switches are made, but the size would be mamouth. This saved the added cost of carrying ground wires in the roughly 200' x 4 of parallel feeders from that disconnect to the service location. I believe the goal of the code is to provide reliable access for emergencies / emergency personell, ( fire fighters ), and the NEC, as written, falls short of that.

Many things in the above need commenting on, but I will only select a few.

1) "I believe the goal of the code is to provide reliable access for emergencies / emergency personell, ( fire fighters ), and the NEC, as written, falls short of that."
- That is not the goal, if it was the service equipment would be required to be outside, but it is not, the service equipment is allowed to be inside ... inside a locked house ... even down in the basement.
- The intent of the service disconnect part of the service equipment is to allow the occupants (or persons acting on their behalf) to disconnect power to the structure without having to call the power company to pull the meter or otherwise shut power off to the structure.

2) "I was unable to find any NEC specific limit to how much unfused wire was allowed inside a building - - - disconcerting. "inside nearest the point of entrance" is subject to inturpretation. Is nearest 5', 15', 50' ?"
- Typically, I have found that the length allowed is between 5 feet and 8 feet. The 5 feet comes from, according to many, that if you installed the meter and service equipment back-to-back on a wall, that the length required to go through the wall (2x6 framed or concrete block) is basically 1 foot (rounded off), plus another 3 feet to 4 feet of conductor length to go into the service equipment and turn up and make a bend down into terminals at the top of the panelboard bus bars (granted, one could say that every installation had to have the terminals at the bottom for less length, but the NEC is not a design manual, so the panelboard can be installed either way).
- The 8 feet comes from being under the slab and coming up through the concrete floor to the bottom of the enclosure and then going up and bending back down to top terminals, that would take approximately 8 feet of conductor length with the panel installed where the uppermost overcurrent device (the main breaker) is installed with its handle at 6 feet 7 inches above the floor.

[Garry Blankenship]

Many things in the above need commenting on, but I will only select a few.

1) "I believe the goal of the code is to provide reliable access for emergencies / emergency personell, ( fire fighters ), and the NEC, as written, falls short of that."
- That is not the goal, if it was the service equipment would be required to be outside, but it is not, the service equipment is allowed to be inside ... inside a locked house ... even down in the basement.
- The intent of the service disconnect part of the service equipment is to allow the occupants (or persons acting on their behalf) to disconnect power to the structure without having to call the power company to pull the meter or otherwise shut power off to the structure.

2) "I was unable to find any NEC specific limit to how much unfused wire was allowed inside a building - - - disconcerting. "inside nearest the point of entrance" is subject to inturpretation. Is nearest 5', 15', 50' ?"
- Typically, I have found that the length allowed is between 5 feet and 8 feet. The 5 feet comes from, according to many, that if you installed the meter and service equipment back-to-back on a wall, that the length required to go through the wall (2x6 framed or concrete block) is basically 1 foot (rounded off), plus another 3 feet to 4 feet of conductor length to go into the service equipment and turn up and make a bend down into terminals at the top of the panelboard bus bars (granted, one could say that every installation had to have the terminals at the bottom for less length, but the NEC is not a design manual, so the panelboard can be installed either way).
- The 8 feet comes from being under the slab and coming up through the concrete floor to the bottom of the enclosure and then going up and bending back down to top terminals, that would take approximately 8 feet of conductor length with the panel installed where the uppermost overcurrent device (the main breaker) is installed with its handle at 6 feet 7 inches above the floor.

OK, I'll bite. I said the amount of unfused wire allowed inside a structure is subject to inturpretation. You do not agree and go on with speculation and inturpretation. To back your position you say "typically", "I have found", "according to many", "the length allowed is between" X & X, and then go on with a maybe formula based upon one specific set of construction materials and circumstances. If that is not inturpreting the distance allowed and speculating as well, I do not know what is.

On the code intent, ( goal of the code ), you say reliable access for emergencies is not it, but do not say what is. Exterior accessed services are a security problem. The code exists for our safety and emergency disconnection of a service is critically important. Fire Marshalls are probably the single largest influence on electrical code. Utility company participation or convenience has nothing to do with it. All services can be turned off by non-utility people. Many services are CT metered and pulling those meters would accomplish nothing other than allowing some free power use. The lack of specificity in the code on the length of unfused wire allowed inside a building is a fire fighters bain. The code needs to do a better job of defining this. My inturpretation of the OP is it's not legal because the panel is not "nearest" the point of entry. If an AHJ considers the crawl space outside the building line, it can be legal. So you need to disconnect the power in an unoccupied building that is on fire - - - where do you go ?

[Jerry Peck]

OK, I'll bite. I said the amount of unfused wire allowed inside a structure is subject to inturpretation. You do not agree and go on with speculation and inturpretation.

I guess my post was not as clear as my thinking of what I was posting - I did not disagree with that it was subject to interpretation, I was disagreeing with this:

I spent many years assuming the NEC allowed a maximum of 15' of unfused conductors inside a building. Recently challenged to find that, I discovered it is a Washington State requirement, ( my home turf ), not NEC. I was unable to find any NEC specific limit to how much unfused wire was allowed inside a building - - - disconcerting. "inside nearest the point of entrance" is subject to inturpretation. Is nearest 5', 15', 50' ?

I was disagreeing with all the dimensions except for the 5' one, which I stated in my post and explained why many use that, and then provided 8' and explained when some use that.


On the code intent, ( goal of the code ), you say reliable access for emergencies is not it, but do not say what is.

I did say what it was - see 1) in my post and the second item under 1).


So you need to disconnect the power in an unoccupied building that is on fire - - - where do you go ?

Read the first item under my 1). I agree with you that, in my opinion, the service disconnect 'should' be outside, but acknowledge that 'should' is not stated in the code, in fact, the code states that it is allowed inside OR outside.

Many fire departments will not put water on a burning structure which has electrical power to it. Most will call the power company to disconnect the power - and sometimes that can take a l-o-n-g time.

However, the primary purpose of all codes is to get people out of the building safely, the codes are not there to 'protect the building or lessen the damage to the building', so, with that said, going back to your example of an unoccupied building ... by definition of 'unoccupied' there is no one in the building, so ... with the primary objective being to get people out of the building ... that was already accomplished by the fact that the building was unoccupied to start with.