Q: I have a quick question. Could I get the exact wording of what the National Electrical Code 2002 states concerning abandoned cables in the plenum?
Brent Clements
Rice University
Houston, TX
A: The question may be quick, but the answer certainly is not. Let's begin with the definitions of abandoned cable. In NEC 2002, there are seven definitions for abandoned cable. Ironically, none are in Article 100, the section on definitions (yet)
- Article 640 Audio Signal Processing, Amplification, and Reproduction Equipment, Section 640.2 defines "Abandoned Audio dis- tribution Cable" as "installed audio distribution cable that is not terminated at equipment and not identified for future use with a tag."
- Article 725 Class 1, Class 2, and Class 3 Remote-Control, Signaling, and Power-Limited Circuits, Section 725.2 defines "Abandoned Class 2, Class 3, and PLTC Cable" as "installed Class 2, Class 3, and PLTC cable that is not terminated at equipment and not identified for future use with a tag."
- Article 760 Fire Alarm Systems, Section 760.2 defines "Abandoned Fire Alarm Cable" as "Installed fire alarm cable that is not terminated at equipment other than a connector and not identified for future use with a tag."
- Article 770 Optical Fiber Cables and Raceways, Section 770.2 defines "Abandoned Optical Fiber Cable" as "Installed optical fiber cable that is not terminated at equipment other than a connector and not identified for future use with a tag."
- Article 800 Communications Circuits, Section 800.2 defines "Abandoned Communications Cable" as "Installed communications cable that is not terminated at both ends at a connector or other equipment and not identified for future use with a tag."
- Article 820 Community Antenna Television and Radio Distribution Systems, Section 820.2 defines "Abandoned Coaxial Cable" as "Installed coaxial cable that is not terminated at equipment other than a coaxial connector and not identified for future use with a tag."
- Article 830 Network-Powered Broadband Communications Systems, Section 830.2 defines "Abandoned Network-Powered Broadband Communications Cable" as "Installed network-powered broadband communications cable that is not terminated at equipment other than a connector and not identified for future use with a tag."
So, what is this telling us? Notice the common thread, "and not identified for future use with a tag." Any cable addressed in these specific articles that is "tagged for future use" is not abandoned cable.
In some cases, having a connector terminated at one end (audio distribution cable; Class 2, Class 3, and PLTC Cable) or both ends (communications cable) of the cable means that it is not abandoned cable.
While in other cases, having the cable connected to equipment that is not a connector (fire alarm cable; optical-fiber cable; coaxial cable; network-powered broadband communications cable) means that it is not an abandoned cable. In NEC speak, a "connector" is "equipment," and that is why the "equipment other than a connector" verbiage is used.
Most Sections [640.3(A); 725.3(B); 760.3(A); 770.3(A); 800.52(B); 820.3; 830(A)] only address the "accessible portion" of an abandoned cable as "not be permitted to remain." Article 100 defines "Accessible (as applied to wiring methods)" as "Capable of being removed or exposed without damaging the building structure or finish or not permanently closed in by the structure or finish of the building." USo, this is not a surgical removal in a finished space. But if the entire area of the building is undergoing demolition for remodeling, then all of the cable should be accessible and removed.
And then there is Article 645—the one that did not bother to uniquely define "abandoned cable," which is only concerned with abandoned cable that is not contained in metal raceways. So, metal raceways full of dead cable under raised floors in data centers are not a concern?
You asked specifically about plenum spaces, but the same text applies to plenum, riser, and hollow spaces in the building. If it is "abandoned" and "accessible," take it out.
Seven definitions for almost the same term is a lot, even for the NFPA. So, for NEC 2005, the International Brotherhood of Electrical Workers Codes and Standards Committee is proposing to delete the current seven to add a new definition to Article 100, which would read: "Abandoned Cable. Installed cable that is not terminated at both ends at equipment and not identified for future use with a tag."
But so far, the other "stakeholders," who carefully crafted the various definitions in their respective Articles, are not supportive of this "one-size-fits-all" approach. It is going to be an interesting couple of years.
Q: I was reviewing the discussion on the BICSI Public Forum about laying cables parallel to each other in a fixed physical relationship. I noticed that one comment said that the European norm addresses it to some degree. Do you know if any of the U.S. standards have intentions to address it? I have a client in New Jersey, with a large data center, that is asking me for input. At present, they are laying large bundles of cables into cable trays, and they want the cables combed as they are installed. I told them I would check and let them know.
Bobby Ashton, Jr.
RCDD/LAN Specialist
South Windsor, CT
A: Intentions, yes. At the TIA TR-42.1 February 2002 meeting, a presentation (TR 42.1-2002-013) was made that referenced a list of contributions, previously submitted to either the TR-42.7 Telecommunications Copper Cabling Subcommittee or the Cable Working Group, dating from 1997 through 2002. Each of these contributions had addressed some negative effect on the transmission performance of category UTP cabling caused by bundling or other similar cabling installation practices.
TR-42.1-2002-013 suggested that, based on these contributions, a Study Group be formed to develop information on the expected variance in transmission performance of cables that are installed in close proximity.
TR-42.1-2002-013 further proposed that once the work was completed and validated, it would be drafted as a matrix, and published as an informative Annex to TIA/EIA-568B.1. The matrix would include alien NEXT and FEXT for Category 5e and Category 6 cables that are tie-wrapped at specified intervals, cables installed in a tray with 50% fill, cables installed in a conduit with 40% fill, and cables suspended on J-hooks.
But so far, no draft matrix has been distributed from the Study Group. Meanwhile, we will have to continue to scare our clients suffering from "overneatness tendencies," using threats of the dreaded "alien crosstalk." You are welcome to use the following. It has worked for me in the past:
Cabling systems and network hardware are designed to handle predictable noise, like NEXT and FEXT, which comes from within the same cable. This is why when you remove the sheath from a twisted-pair cable, each of the pairs is twisted at a different rate.
But when UTP cables of the same construction are neatly dressed and packed, you are inviting trouble. You are creating a quasi field-manufactured hybrid cable, because the twists are the same in each of the neighboring cables. This creates unpredictable noise between neighboring cables; the noise is called alien crosstalk. And the neater the cables are laid, and the closer you pack them together, the more likely you are to experience the problem.
For example, consider a 1000Base-T scenario using seven UTP cables neatly laid and packed. You will have all four pairs in each of the six disturbers, and one victim cable simultaneously, bidirectionally active.
That means that there are now 24 active pairs producing alien crosstalk, six of which are exact matches to each of the four pairs inside the victim cable. No wonder they call it "alien"—looks like an all-out invasion.
Now the network hardware must not only contend with the predictable noise—NEXT and FEXT from within the cable—but also with unpredictable noise from its six sets of disturber neighbors as well. Does neatness really count this much?
If the UTP cables are randomly laid and the pathway is not overfilled, then "alien crosstalk" should be one less network anomaly you have to address.
And what about testing? While your network can certainly suffer from the effects, measuring alien crosstalk is not something you want to attempt in the field. Most measurement models that I have seen use seven cables—six disturbers and one victim cable. That would mean seven sets of field test instruments in use at the same time to test one cable link. And there are currently no pass/fail limits proposed or set.
My advice: loosen up and avoid the problem.
The BICSI Public Forum is a great tool where you can post messages or documents on telecommunications premises cabling issues. If you would like to participate in the BICSI Public Forum online, see www.bicsi.org and click "Forums."
Donna Ballast is BICSI's standards representative, and a BICSI registered communications distribution designer (RCDD). Send your questions to Donna via e-mail: [email protected]