By Patrick McLaughlin
As we have published in two articles over the past several months, the National Fire Protection Association’s 2017 edition of the National Electrical Code could-and many believe, likely will-include language addressing the copper media that supports the delivery of power to networked devices. (“PoE in the NEC? Signs point to yes,” September 2015; “Proposed revisions to 2017 National Electrical Code would impact PoE deployment,” January 2016). As of the time this article is being written, we do not have substantive updates to report concerning the revision process for the 2017 NEC, so the article does not include specific detail about that progress. Rather, this article will provide information on developments within the industry that have taken place since the previous articles were published.
In September 2015, Underwriters Laboratories (UL) made public a report on its fact-finding investigation on the delivery of power over four-pair local area network (LAN) cable. The full report is posted on the website of the Society of the Plastics Industry (SPI; plasticsindustry.org), which commissioned the UL study. The report is 164 pages in length; pages 44 through 164 are appendices populated with charts, tables, and similar informational graphics from UL’s investigation.
The report explains the approach UL took to its investigation, as well as the justification for that approach. Specifically, “Several possible investigation plans were considered,” the report says. “The first involved selecting a number of installation configurations considered to be representative of both typical installations and worst-case heating. This would be based on data already circulating on this subject. Different cable types would be tested in the various configuration.
“The second approach would be concentrated on installation configurations, testing a wider variety of configurations using a single cable type,” the report continues. “The result of these tests could then be used to determine a much narrower set of useful test configurations that could be used to test other cable types. Since this second approach would yield hard data that not only better supports the final test configurations but would also provide valuable data to support the NEC proposals related to powering over LAN cable, this is the approach that was chosen.”
Six specific cable types were tested: Category 5e, 24-AWG UTP; Category 5e, 26-AWG stranded FTP shielded patch cable; Category 6, 23-AWG UTP riser-rated CMR; Category 6, 23-AWG UTP plenum-rated CMP; Category 6, 22-AWG UTP plenum-rated CMP; and Category 6A, 23-AWG shielded plenum. UL said these cables were selected to include the lightest and heaviest wire gauges, and to include different diameters, which will affect thermal characteristics.
Individuals who enjoy-or can tolerate-reading industry standards may also enjoy or tolerate the fact-finding report and its appendices. For those without such an appetite, the report includes the following in its summary: “The investigation resulted in the development of data in support of comments/proposals related to including more extensive ampacity tables in the NEC to better manage powering over LAN cable systems … The investigation has identified that for certain powering over LAN cable installations with power levels exceeding 60 watts, overheating of the cables will occur. Cable heating can be managed via increased AWG size; cable design; material selection; installation practices. The investigation produced data leading to the development of testing requirements for ‘-LP’ cables.”
Talk of the town
The “more extensive ampacity tables” mentioned in UL’s report were a significant point of focus in a presentation delivered to the main assembly hall at the BICSI Winter Conference held in early February. Power over Ethernet was a common theme among several of the conference’s technical presentations, including one titled “Will High-Power Ethernet (HPoE) Change Installation Practices for New and Existing Cable Plants?” The presentation was delivered by Rick Foster, director of IT infrastructure and commissioning services for Innovative Engineering Services; and by Andrew Jimenez, vice president of technology for enterprise cabling and security solutions with Anixter.
Foster and Jimenez spent a considerable amount of time on the ampacity table that is included in this article (and which we also published in January). The duo emphasized that the table addresses four-pair, and therefore eight-conductor, cables, with conductor sizes ranging from 26 to 22 AWG, arranged in bundles of varying sizes up to 192 cables, and with three different cable temperature ratings (60, 75 and 90 degrees Celsius).
Additionally, they pointed out “Note 2” in the table, which reads, “Where only half of the conductors in each cable are carrying current, the values in the table shall be permitted to be increased by a factor of 1.4.” Going back to UL’s investigation report, it states, “Some implementations of remote powering over LAN cable utilize two-pair powering schemes … These powering methods are still widely used … There is less heat generated in each cable for the same amount of current per conductor. This would imply that the four conductors could each carry more current than each of the eight conductors in the four-pair power configuration to get the same effect. However, half the number of conductors carrying the current does not translate to twice the current, since the heating effect is related to the current squared … Data indicates that a factor of approximately 1.4 would be valid to estimate how much additional current could be carried by a four-pair cable when only half of the conductors are powered.”
LP in focus
Christopher DiMinico authored an article in our January issue (“Proposed revisions to 2017 National Electrical Code would impact PoE deployment,”) in which he quoted the following from the second draft revision of the NEC articles addressing power over LAN cables: “Cables with the ‘-LP’ suffix shall be permitted to be installed in bundles, raceways, cable trays, communications raceways, and cable routing assemblies. Cables with the suffix ‘-LP’ and a marked ampere level shall follow the substitution hierarchy of Table 725.144 … for the cable type without the suffix ‘-LP.’”
UL’s investigation report notes, “A need was identified to develop requirements for special-use cable that could be used as an alternative to traditional cables. Where the ampacity tables were sufficient for existing installations and new installations where the use of traditional cables was desirable, it was recognized that cables could be designed specifically to handle powering over communications cables without all of the limitations necessary for cables of unknown heating and heat dissipating characteristics. The objective, then, was to develop requirements that would permit a cable to be identified specifically for this type of installation and use.
“It has been well established that cable heating can be managed via: increased AWG size; cable design; material selection; installation practices. It was less clear how the interaction of these elements might affect a cable’s ability to handle increasing current levels. For example, there was little data available showing how changing a cable’s construction without changing the AWG size might affect temperatures, or how variations in the installation might affect the temperatures for the same cable.
“Obtaining data on these variations was necessary to determine how an ‘LP’ cable might be evaluated.” The report said that several tests were identified as being critical in this effort; those tests were different installations (testing the same cable in various configurations), orientation (horizontal versus vertical), cable comparison (different cable types and constructions), and test variations (effects of test variations on temperature rise under various conditions).
Anthony Tassone, principal engineer for wire and cable with UL, has authored a four-page document encapsulating some of the fact-finding study’s high points. The document, titled “Fact-Finding Study on Powering Over Data Cables,” is available on UL’s website. In it, he states, “UL is confident that by taking advantage of advances in cable design, LP cables will provide an attractive alternative to ampacity tables and bundle-size limitations. They provide an uncomplicated way to prepare installations for increasing power levels, and they are not as susceptible to the issues caused by excessive heat generation. UL continues to do research as part of an ongoing process to support growth and innovation in the cabling industry.”
In a separate document on its website, UL says, “UL has introduced a Limited Power (LP) Certification to simplify the cable choice and installation considerations. The ‘-LP’ cable designation indicates that the cable has been evaluated to carry the marked current under installation scenarios without exceeding the temperature rating of the cable.”
While that is quite a bit to digest-and it is only excerpts from lengthy documents-it is apparent that the use of LP cable is being positioned as a simpler alternative to abiding by a complex table when planning power-over-LAN cabling installations.
In his article published in January, DiMinico contended, “Consideration of the changes should be deferred to the 2020 Code revision cycle, permitting sufficient time for all interested and affected parties to review and comment” on matters including the ampacity table discussed in this article, as well as the research and investigation conducted in the development of the TIA’s TSB-184-A.
Our next report will update the progress of the 2017 NEC revisions, exploring the extent to which these PoE-related changes will be incorporated.
Patrick McLaughlin is our chief editor.