During final inspection for a new 500kVA UPS installation we discovered the installer had pulled all the cable through 4" EMT but had used screw-set couplings instead of the compression couplings we consider standard practice. We recalled seeing that in writing somewhere, couldn't recall in which mil-hdbk it was specified. Regardless, the mil-hdbks are guidelines only, which gives us the final call and we couldn't think of anything that justified taking all the wiring out to replace the fittings. All the cables were run with a ground wire, and the cable running from the bypass switch to the main switchgear was in an unprotected cable tray, so grounding and EMI are out as concerns, which were the only reasons we could think of that compression couplings would be better than screw set. This is in an earthquake zone, but I can't see how that would make a difference.

Do any of you know why compression couplings would be favored/required over screw set? Is it just for grounding which is moot with a decidated ground wire?

Sometimes folks just like the look- or because there's no screw poking out, to get caught on things.

Other times, folks are under the mis-impression that the compression types "ground" better. My expereince is the opposite; the nuts are sometimes foung loose; I have never had a set-screw loosen up.

As for project specs...well, I can't help you there. Sometimes customers want compression, sometimes they want steel, etc. If it was spec'd, you are stuck with replacing them. Sorry.

[This message has been edited by renosteinke (edited 11-25-2005).]

Take a look at this thread. http://www.eng-tips.com/viewthread.cfm?qid=140080&page=1

Iv'e seen compression steel fittings on alot of print spects for schools.
I really don't see any benefit in compression over set screw indoors.
On large conduit it's almost impossible to get the couplings so tight that they don't turn when tighting the next coupling.
They also have a tendancy to strip out the threads of the fitting on smaller conduit when over tightened.

Interesting, I'd not suspected the arguments were for longevity! Sounds like properly installed compression fittings are the way to go, but difficult to achieve- I'll have to remember to closely inspect them next time around; I'm the customer as well as the engineer. This installation is expected to last the lifetime of the UPS- 15 years. They used compression fittings on all the small conduit, but not the big ones, so cost and installation difficulty might have both come into play. I'm at least glad to hear there really isn't all that much difference between the two; if compression fittings really are that hard to install and always get done sloppily, we might just go back to screw-set, I'll research the specs some more and see exactly how they word it.

Thanks guys

we've done installs with set screws in production factories that have lasted much longer than the 15 years needed for the UPS.

On EMT runs, one of the only benefits to compression is if you need to vacuum a pull line through for wire installation. The suction from the vacuum tends to leak out with set screws.

Outside of that, you'll never notice the difference. (unless you water cool the conductors, then you'll make a mess;-))

I think set screw fittings are fine 2" and smaller. Most of the time here in Dallas, TX compression type are spec for indoor use. I don't know why.

Steve, in my experience it is not that it is "easy" to make "sloppy" connections with compression type connectors; nor is it a matter of "difficulty" very often.

What seems to hapen is that the connection is tightened, then made loose by the pipe, as the pipe gets moved about during installation. Then, after assembly, it's not so easy to get a wrench on all the fittings.

A set-screw, in contrast, is not affected by these rotational forces, and makes the pipe slightly out-of-round, which further resists twisting.

I would say that the biggest contributing factor was probably cost.
In bigger sizes like 4 inch, compression couplings can run as much as 3 times more expensive than set-screw couplings.
If the installation is indoors and has no need for weather-proofing, I can't say that I see a reason for compression couplings except asthetic value or ease of sucking a pull string through as was mentioned earlier.
If it's above the cieling or otherwise hidden, a wrap of "jap-wrap" at each end of the set-screw connector will greatly improve the runs jet-lining ability.
As for getting large compression couplings to tighten properly, the trick I've found is to use two strap or pipe wrenches that you think are completely too big for the job.
In regards to the grounding aspect, I've never had a problem with set-screw connectors as long as you make sure to tighten the screws all the way.

I wonder if anyone has applied a meter to a long run of EMT to determine the impedance when set screws are used, and compression.
I've heard third hand that a company did a test and found compression to offer a lower impedance at fault current levels.
Of course I always require a dedicated EGC to be installed within the EMT anyway, but the redundancy of the raceway is helpful IMHO.

I'll give UL a call next week seeing how they list both as acceptable for grounding.
Maybe they have data on it.

"As for getting large compression couplings to tighten properly, the trick I've found is to use two strap or pipe wrenches that you think are completely too big for the job."

Used two pair of 480s, two men, and still had conduits turning in couplings behind us.
The couplings did'nt seam loose but the conduit would turn unless you put a wrench on it.
This is over 2000 foot of 3 1/2 in a little over a year.
I've had alot of practice.

Edit to add, that all of these runs had redundant equipment grounds. Grounding must not be the reasoning for these specs.
Maybe it's, MUST BE BETTER it's used out side, and their harder to use.


[This message has been edited by russ m (edited 11-25-2005).]

As an engineer who's looking in from the outside, rather than installing conduit day-in and day-out, here's my own opinion on the matter.

I've seen way, way too many pieces of EMT hanging by the wires after they've come loose from set-screw fittings. I've especially seen that a lot around my kids' schools. (I suppose an argument could be made that conduit within reach of the kids at a school is "subject to damage," but that's another discussion...)

Because of this, my feeling, coming from real-life observations, is that set-screw fittings are unreliable, to the point of being dangerous.

My gut feeling, before I starting reading this thread, was that compression couplings are more reliable. Now, this discussion suggests that my gut feeling may have been wrong. However, I wouldn't be at all surprised if the engineer who wrote the specs for compression fittings was coming from the same place I am.

I further note that my gut feeling is biased by the fact that there are far more set-screw connections around where I've been than compression connections. Therefore, I'm a lot more likely to see set-screw fittings that have come loose, just because of the numbers. So, if there were a lot of compression fittings around, perhaps I would be seeing those conduits hanging by the wires, too.

All in all, it still seems like a compression connection would be more secure...




[This message has been edited by SolarPowered (edited 11-28-2005).]

Solar,
The cheap (low quality) compression couplings often strip out when they are installed leaving a very poor connection. I don't see this same problem with the low quality set screw couplings. As far as damage, the die cast fittings of both types break easy when they are subjected to force.
As far as the conduits in schools that are subject to the kids, the only occupancy that I use more supports than in schools is prisons. In both cases the code minimum won't do.
Don

In Florida the prison spec was compression fittings, made up wrench tight.

I think that assumption is right on the head... When it comes to conduit coming apart, ground path and survivablity, there is only one that does better than set screw, and compression, RMC threaded connection.

Spare going to RMC, survivablity with EMT is going to depend mostly on support, and the 1 hole strap is the culprit there. A mineralac, or 2 hole strap will help if not on strut. When it comes to ground path and survivability (set screw or compression) steel is better IMO. You can torque it much more than die cast, have a better connection, and be less prone to breakage down the line.

I would think any exposed conduit in a prison would need to be RMC just for security/safety reasons, if not vandal resistance.

Set screw fitting have an advantage in that they have a snaller outside diameter. This is advantageous when drilling through 6" thick concrete walls. Once firmly tightened, the screws can be cut off and the area filed smooth. [Now I am going to run, duck and cover.]

Note: I have seen compression connections pulled apart but never screw couplings. But then I haven't seen too many old work situations with screw connections.

~Peter

Something I noticed with a few cast compression fittings in the 1-2" range is the inside ring is a hair too big. You can crank it as much as you can but it wont hold the pipe very tight. Take it apart & cut a hair off the inside ring. Then just twisting the nut by hand locks the pipe in tight. More of a factory defect. The point is sometimes you could clamp the pipe and then tighten the fitting thinking it's all good but it may not be.

In a set screw if the pipes in all the way you should fell the set screw clamping on the pipe. The pipe wont jump out.

If pipe is not secured well enough for the situation then that's the cause of the failure not the fitting.

If you move the pipe back and forth in a set screw the end of the pipe tends to colaps where the screw is and causes the problems.

Steel compresion is the best short of RMC or maybe IMC.

Also some may cheat om compression because the can't get a wrench on it very well. In a tight place or the middle knock out on a box between 2 pipes. I wish they made some special wrench for this.

When you tighten a connector you have 4 things that want to move on you. The lock nut, the connector, the connector nut, & the pipe. Sometimes it feels like you need 3 wrenches. Simmilar with couplings holding the pipe, coupling, & turning the nut.

Tom

The key to having a set screw fitting not fail is as Tom said, not setting the screw, backing it out, shift it a little, set it again, etc. And bury that screw. Spinning it until it meets the pipe is not enough, you need to bury the screw head into the shoulder. Think of it as a one time shot, position it, bury it.

I have always preferred set screw fitting because I have found so many loose compression fittings. I figured it was the result of expanding and contracting over the years. Inherant flaw in design and the fact that rather than having the tightness determined by something measurable (screw buried into shoulder) one must work from "ahh, at oughta do 'er".

[This message has been edited by Jps1006 (edited 11-28-2005).]

I found the clause this morning in MIL-HDBK-411B "POWER AND THE ENVIRONMENT FOR SENSITIVE DOD ELECTRONIC EQUIPMENT" which discourages set-screw type couplers for EMI reasons. As luck would have it, this MIL-HDBK is approved for public release and distribution unlimited, so I'll copy/paste the section: (It's OCR from a scanned-in pdf, so excuse the typographical errors)

MIL-HDBK-411B: 5.5.6.5.1 Perferred conduit types. Shielding of communications and data circuits from ac-generated noise is improved by enclosing all power conductors in metallic conduit or continuous metallic duct. Ac fault protection and ac power conductors will not be enclosed in the same conduit with communications and data cables. There are three general types of metallic conduit acceptable for power distribution for critical DoD communications and computer facilities: rigid, flexible, and thin wall. Poly-vinyl -chloride (PVC), or similar nonmetallic construction, is not acceptable for power distribution where high reliability is a requirement. Metal conduit with set-screw type connectors should also be avoided. For maximum shielding effectiveness, metal conduit with threaded connectors should be used. To maintain this effective shield, all joints between sections of conduit should be treated with a conductive lubricant or caulk. All power conduits must be sized to accommodate the phase conductors, neutral conductor, and a fault protection conductor (green wire).

Interesting, another MIL-HDBK prohibits use of EMT for any conductor over 1/0 size... Either way, it appears our policy was for EMI reasons, vice safety grounding or mechanical reasons.

[This message has been edited by SteveFehr (edited 11-29-2005).]

Has anybody else seen damaged insulation on wire from set screw connectors? If those set screws are tightened too much there is a definite point for damaging insulation to the point of having an unwanted ground.

I have never seen that and don't think that it is even possible to cause damage to the conductors in the conduit unless you replace the supplied screw with a longer one. Yes, the screw does put a "dimple" in the conduit when fully tightened, but that will not cause conductor damage.
Don

IMHO either compression or set screw is fine when you PROPERLY SUPPORT YOUR PIPE RUNS!!!!!Most of the problems seem to occur because there is to much pressure on the coupling (racks not level and square to oneanother or to much distance between them)