PVC Expansion Couplings and Siding Replacement Problems

[mkgrady]

What code cycle made expansion couplings a requirement?

I have to resolve a problem where a condo complex was built around 1982. There are 24 buildings with meter centers containing 4-6 meters. All service laterals are in 4" PVC and each unit is fed underground in 1 1/2" PVC. Some of the conduits have pulled out of the male adapters and some conduits are pushing up on the enclosure making the enclosure cock eyed.

The reason I was called by the owners is because a GC was hired to put new siding on a six of the buildings. When he disconnected one of the meter center from the wall it blew up. The fasteners were external so he may not have opened the enclosure to unscrew it from the wall. He did not have the same problem at the other five buildings but he should have had an electrician work on the meter center not a carpenter. It may be that the original installation has screws through the back of the enclosure in addition to the external ones and when they pried it off the wall a screw fell in and shorted the bus.

I have been asked to evaluate the condition of the 18 other buildings, to make recomendations for any fixes, and to tell the owner if the GC is responsible for the cost of the repairs. I am told the GC invoiced the condo for $10K to replace the meter center and to go to five of the other buildings he has re-sided to clamp the PVC conduits to the wall under the meter center.

My first impressions are that an electrician would have recognized if a meter center was cockeyed (nobody knows if it was) it may be a disaster waiting to happen and it would not have been moved without a thorough inspectuion for internal problems, and that going to the other five buildings to clamp the conduits to the wall is a way for the GC to save face and get paid for the extra work

 

[iwire]

I don't think you are looking for expansion fitting requirements, I think you are looking for the requirements for ground movement. You can choose to use an expnasion fitting to address the issue but that is just one way.

Currently the rule is 300.5(J)

(J) Earth Movement. Where direct-buried conductors,
raceways, or cables are subject to movement by settlement
or frost, direct-buried conductors, raceways, or cables shall
be arranged so as to prevent damage to the enclosed conductors
or to equipment connected to the raceways.

Informational Note: This section recognizes ?S? loops in
underground direct burial to raceway transitions, expansion
fittings in raceway risers to fixed equipment, and, generally,
the provision of flexible connections to equipment subject
to settlement or frost heaves.

I have no idea at all if this section was anywhere in the NEC in 1982.

 

[dicklaxt]

Normally,riser pipes of exposed 4 to 5 feet is not a problem but horizontal pipe in lengths greater than even 15 to 20 feet can give you problems with thermal expansion.

Land subsidance is certainly another issue to be dealt with as noted by iwire.It is noticebly evident in coastal areas.

dick

 

[mkgrady]

I don't think you are looking for expansion fitting requirements, I think you are looking for the requirements for ground movement. You can choose to use an expnasion fitting to address the issue but that is just one way.

Currently the rule is 300.5(J)
I have no idea at all if this section was anywhere in the NEC in 1982.

I removed one of the covers of the incoming section. The service lateral wires ran straight up from the conduit into the main lugs. Seems like an S curve in those wires would have been better.

 

[mkgrady]

Normally,riser pipes of exposed 4 to 5 feet is not a problem but horizontal pipe in lengths greater than even 15 to 20 feet can give you problems with thermal expansion.

Land subsidance is certainly another issue to be dealt with as noted by iwire.It is noticebly evident in coastal areas.

dick

These conduits run only 2-3 feet from the surface of the ground to the bottom of the enclosure. Does that say expansion couplings are not necessary? Is the ground considered the conduit support making the added clips unnecessary?

The contractor added unistrut and pipe clamps under the enclosure to prevent future problems. I'm thinking this is a waste of time and a red herring to justify the extra charges.

Since Iwire mentioned the S curve in the wires I'm thinking that is the right fix. Only the incoming wires are run straight and tight as a banjo string. The rest have pleny of slack and of course are much smaller and flexible.

 

[steve66]

Is the conduit just a sleeve for the service wires between the meter and the ground, or does the conduit continue underground back to the source of the service?

If the conduit continues under ground, the problem that Iwire mentioned is that the conduit will move as the ground freezes and thaws (at least it would in this area). So the conduit has to move, but the meter box can't since it is fastened to the wall.

If the conduit is just a sleeve, and it ends shortly after entering the soil, an S-loop would probably let the conduit just move with the meter. An S-loop won't do anything for a complete conduit run.

 

[dicklaxt]

These conduits run only 2-3 feet from the surface of the ground to the bottom of the enclosure. Does that say expansion couplings are not necessary?

No not by anymeans,what I said was it is normally not a problem with the expansion/contraction of PVC. If you look at Table 352.44 NEC 2008 you will see the movement across 100 degree change/100 feet is 4.06 inches,now if you expand those numbers.

4.06 divided by 100 feet ='s .0406 inches /foot and you are dealing with 3 feet then the movement is .0404 times 3 feet ='s .1218 inches (barely over 1/8th inch.This is based on a temp differential of 100 degrees difference from installation ambient to maximum extreme above or below.Lets split the difference and say the difference is 50 degrees now we are dealing with only 1/16th inch.It certainly will cause movement but in the case of the example I don't believe to be a problem.

Now if ground upheaval in this geographic area is 1 inch plus or what ever then I see a problem and an expansion fitting would be the solution but a slack loop is still needed too IMO.

Going a step further the thermal expansion of copper is only 1.1 inch/100 feet so when the PVC moves 4 inches the copper is going to be put under a tug of war if a slack loop is not provided.

I don't have a clue as to when the Code required EXP. Fittings.

dick

 

[mkgrady]

I don't know if the incoming feed is in conduit the whole way or just sleeved beneath the meter center.

I spoke with the inspector this morning. He answered the question about what caused the blow up. A carpenter unscrewed the meter center from the wall in order to replace the wall surface behind it. Some time before the enclosure was moved the incoming wires had put such a strain on the bus that one or more of the lugs/bus had broken loose from its insulator. When the carpenter disconnected the enclosure from the wall those legs made contact with each other. I don't know if it was hot to hot or hot to ground but there was plenty of melted metal when he was done.

This tells me that the carpenter (GC) is responsible for making repairs because they are required to pull a permit and use an electrician to unscrew electrical equipment from a wall. If they had hired an electrician he would likely have seen a problem with the cockeyed equipment.

The inspector believes that if the conduits had been secured to the wall beneath the meter center it would have prevented the problem. The theory being that as the conduit settled in the trench the wires pulled on the lugs/bus and enclosure and if they were clipped to the wall the wires would not have put a strain on the equipment. I'm not sure I am convinced of that theory but I did discuss with him adding enough wire to form an S curve to alleviate the problem.

He did not think the S curve would help and suggested that the wires be lengthened so that a U shape could be put in the wire. There may be enough space in the enclosure to run the wires past the lugs and make a U turn and then make the connection. I think that might work. He also said that expansion couplings would not have helped