Clarifying NEC Articles 725 and 800

Electrical engineers should understand the differences between NEC Article 725 and Article 800, and how to apply them in low-voltage telecom projects.

By Timothy E. Kuhlman, PE, RCDD, CH2M HILL, Portland, Ore. July 17, 2013

Learning objectives

  • Understand NEC Articles 725 and 800, and how they affect network cabling for electrical engineers.
  • Determine which Article to start from, depending on the circuit type being installed.
  • Identify how the circuits from the two Articles support each other in installation requirements, shared pathways, and cable type substitution. 

Why is there so much confusion among professionals in the telecom, networking, and electrical cabling industries on the application of National Electrical Code (NEC) Articles 725 and 800? The confusion is apparent when you ask the question, “What section of the electrical code is applicable to a telecom cable routed from a telecommunication room to a telecom outlet?” (see Figure 1). The segment of cable described makes up almost 100% of the cabling that is designed for a network connection to a desktop computer, workstation, and telephone. Figure 1 shows cabling connected to a network switch, which in turn is connected a file server not shown in the diagram. It also shows a cable connected to a 110 block, which is cross connected to a private branch exchange, typically known as a PBX.

I have asked this question on several occasions at conferences and seminars. From a rough survey, approximately 50% of the attendees believe this segment of cable is covered under Article 800, 40% believe it is covered under Article 725, and the other 10% don’t know. I also heard it dismissed as all being “low-voltage” cable, as if knowing the applicable code section doesn’t matter.

What is low-voltage and does it matter?

First, let’s address the 10% who believe this is a not an issue because this is only a “low-voltage” cable. Is the network or telephone cable in our example really low-voltage? Yes, it is low-voltage when compared to power and lighting circuits, but the text of the NEC does not describe networking or telephone cables as low-voltage.

Article 800 does not use the term “low-voltage” to describe circuits. In Article 725, there is only one reference to the term “low-voltage” as it relates to power limited tray cable, and that does not apply to our application. The NEC does not have a formal definition of the term in Article 100: Definitions. The term is defined in a few different articles in the description of clearances and of healthcare facilities, and in Chapter 6 for special equipment. None of these terms directly applies to a network cable.

Often the authority having jurisdiction (AHJ) over the installation of this cabling type requires some form of low-voltage or limited voltage license to be held by the contractor performing the installation work. I suspect this is where the casual use of the term “low-voltage” comes from, but it does not provide an excuse for being dismissive in understanding the correct classification for this circuit. There is a good reason why the NEC covers network cabling. Even though a network cable does not present harmful voltages, there are a variety of safety issues to be considered such as separation from power circuits, the correct classification of the cable for the environment, cable support, and routing, to name a few.

As more network cabling is installed, the more important it becomes to understand the proper classification of the circuit types. It is not hard to see in the telecommunications industry a steady pace of converging technologies. It started with data and telephone systems and has evolved to security, access control, closed-circuit TV, radio, and control systems.

If you have seen terms such as VoIP, RoIP, or SoIP (voice over Internet protocol, radio over Internet protocol, and storage over Internet protocol, respectively), then you are a witness to the convergence of other technologies with telecommunication systems. The continued convergence of systems using network cabling reinforces the need to properly understand the section of the code that applies to network cabling. As professionals in this industry, whether inspector, designer, or installer, we have an obligation to know how to properly classify the circuits in the NEC to ensure their safe design and installation.

Defining a communication circuit

Let’s consider the 50% of the folks who believe the circuit described in Figure 1 is covered under Article 800. They have a pretty good argument. The circuit described is commonly called a “communications circuit,” and aren’t the terms “telecommunications” and “communications” interchangeable? The cabling installed for a network connection in our example is an unshielded twisted-pair cable, which is very similar to the construction of cable used for legacy telephone cabling. To this day we commonly use the same cable and cabling components for both networks and telephones. The best argument supporting the application of Article 800 is that the cable installed for networking will be marked CM, CMR, or CMP per the NEC, and the NEC defines these in Article 800. If you were to take a moment to look at the network cable connected to your laptop or workstation, you would most likely see one of these cable markings.

All of these arguments would lead you to the wrong conclusion. Unfortunately, I have seen manufacturers in our industry enforce this belief by mentioning only Article 800 in their design guides and other informative literature. I suspect some of this is due to the forefathers of our network cabling technologies coming out of the telephone industry where Article 800 is rightfully applied. To see why Article 800 does not apply to the segment of cable described in Figure 1, it is important to understand the scope and definition of Article 800.

Article 800 is titled “Communication Circuits.” Within the NEC it goes on to define the scope of the article in 800.1 as: “This article covers communication circuits and equipment.” Therefore, to understand the scope of this article, it is necessary to understand the definition of a communication circuit. Depending on the text, design guide, or industry standard you look at, you will probably see something that describes our circuit in Figure 1. In this case, these other sources that define this term are irrelevant. When interpreting the code, the perspective that matters is how the NEC defines a communication circuit. The NEC in 800.2 states:

Communications Circuit: The circuit that extends voice, audio, video, data, interactive services, telegraph (except radio), outside wiring for fire alarm, and burglar alarm from the communications utility to the customer’s communications equipment up to and including terminal equipment such as a telephone, fax machine, or answering machine.

In typical NEC language, this definition is not written in a form that would make it easy to understand. Without changing the meaning of the definition, consider the same sentence in a different form.

Communications Circuit:

The circuit that extends:

  • Voice
  • Audio
  • Video
  • Data
  • Interactive services
  • Telegraph (except radio)
  • Outside wiring for fire alarm
  • Burglar alarm, from the communications utility to the customer’s communications equipment up to and including terminal equipment, such as a:
    • Telephone
    • Fax machine
    • Answering machine. 

The first eight bullets are types of circuits being extended from a communication utility. The bottom three bullets describe the type of customer terminal equipment the utility is connecting to. Therefore, what the NEC is defining as a “communication circuit” is that which extends a communication utility circuit to a customer’s terminal equipment.

This is important to understanding the scope of Article 800. Article 800 applies only to those circuits being extended by the communication utility, with “communications utility” being the key term. In our example in Figure 1, the network cable is not extending a service from the communications utility. What is shown is a privately owned network circuit, and it is being extended in the building from a customer-owned network switch to a customer owned workstation.

Therefore, what the NEC is defining as a “communication circuit” is that which extends a communication utility circuit to a customer’s terminal equipment.

Within Article 800 there is no definition of a communications utility. There is, however, an informational note under section 800.1: Scope that refers you to section 90.2(B)(4) for the installation of circuits that are not covered by Article 800. Paragraph 90.2(B)(4) says that the “installations of communications equipment under the exclusive control of communications utilities located outdoors or in building spaces used exclusively for such installations” is not covered by the NEC. The informational notes at the end of 90.2(B)(4) define a “utility” as an entity that is typically recognized by public regulation such as in the case of electrical supply generation, telephone, and cable television services. Therefore, when Article 800 is referring to a communication utility in the communication circuit definition, it is referring to a circuit from a utility company.

From this we can tell the NEC does not apply to utility company buildings or the services outside that are controlled by a utility company. There are times when the telephone utility service needs to extend services (or circuits) inside a privately owned building—for example, in the case where T1 lines, digital subscriber lines (DSL), or plain old telephone system (POTS) lines need to be extended into a building to a customer’s telephone switch or network equipment. Article 800 provides the code requirements for this cabling. In modern telecommunication networks, the cabling of this type is a fraction of the cabling covered by NEC 800, but this wasn’t always the case. Since the privatization of telephone systems and the emergence of network computing, the services of the telephone companies’ equipment and cabling in the private enterprise have been diminishing.

If not in Article 800, then where?

The section of the NEC that applies to the cable illustrated in Figure 1 is covered by Article 725. This may not be immediately obvious by reading the first couple of pages of this Article. It is easier to come to this determination by first reading Article 645: Information Technology Equipment, which describes the code requirements for the cabling of computer systems (defined by the NFPA as information technology equipment). The cabling for our illustration in Figure 1 is for the data transport of the computing system. It is within section 645.3, Other Articles, in paragraph (D) that the NEC defines the electrical classification of a data circuit. Figure 2 shows the linkage between the sections within Article 645 to the circuit types defined in 725.

From the references in Figure 2, you would be correct in describing the network cable in our example as a “signaling circuit,” a “Class 2 or Class 3 circuit,” or a “limited-power circuit.” (Please note that nowhere is this defined as a low-voltage circuit. The NEC Handbook uses this description, but it is not in the text of the NEC itself.) When working with the NEC, my preference is to refer to a network cable as a limited-power circuit or just as a Class 2 or 3 circuit. The NEC recognizes a difference between a Class 2 circuit and a Class 3 circuit as a Class 2 circuit being intrinsically safe. A Class 3 circuit has a protection device incorporated into the power supply to make it intrinsically safe. However, for Class 2 and Class 3 circuits, the same sections of Article 725 apply. In the small power supplies I have seen for POE devices, media converters, powered telephone sets, and other small network devices, they have all been listed as Class 2 power supplies. The larger network equipment, such as multi-card chassis, have much larger power supplies and probably have some form of internal protection circuiting to make them intrinsically safe, although I have never seen them listed as a Class 3 device. Therefore, I often refer these as just limited-power or Class 2 circuits.

There are provisions in sections 645.3(D) and 725.139(D) for a Class 2 circuit to be considered a communication circuit and therefore be installed per Article 800. This is in the case where the Class 2 (network) circuit shares the same cable with the communications circuit. This is often misinterpreted to mean that when Class 2 cables and communication cables are combined, such as in a tray or conduit, all the cables can be considered communications cables and would be installed per Article 800. This is not what the code says. It states in both of the sections referenced that when the two circuit types share the same cable, they can be combined. Or in other words, both circuits have to be within the same cable jacket. The sections are written to say the two different types of circuits have to be within the same cable.

During the early days of 10Base-T Ethernet, you would often see where two pairs of a four-pair Category 3 cable were used for the Ethernet circuit and one pair of the same cable was used for telephone. The Ethernet circuit would be covered under Article 725, and if the telephone phone circuit was an extension of a telephone line from the telephone company, it would be covered under Article 800. Sections 645.3(D) and 725.139(D) allowed this convergence of technologies to be covered under Article 800. Nowadays, a circuit of this configuration is rare.

More on Article 725  

Additional information on the intent of Article 725 can be found in the NEC Handbook in the editors’ text that describes the scope of Article 725. The NEC Handbook states:

The installation requirements for the low-voltage wiring of information technology equipment (electronic data processing and computer equipment) located within the confines of a room that is constructed according to the requirements of NFPA 75, Standard for the Protection of Information Technology Equipment, are not covered by Article 725. Low-voltage wiring within these specially constructed rooms is covered in Article 645.

Also, if listed computer equipment is interconnected and all the interconnected equipment is in close proximity, the wiring is considered an integral part of the equipment and therefore not subject to the requirements of Article 725. If the wiring leaves the group of equipment to connect to other devices in the same room or elsewhere in the building, the wiring is considered ‘‘wiring within buildings’’ and is subject to the requirements of Article 725.

The last sentence of this quote from the Handbook addresses the cabling in our example. The editors’ text from Article 800 of the NEC Handbook complements this. It states:

Section 90.3, covering the structure of the NEC, specifies that Chapter 8 (comprised of Articles 800, 810, 820, and 830) covers communications systems and is not subject to the requirements of Chapters 1 through 7, other than where a requirement from these chapters is specifically cited by a Chapter 8 requirement. As an example, 800.24 references 300.4(D) and 300.11, 800.44(A)(3) references 225.14(D), and 800.90(C) references 500.5.

Although information technology equipment systems are often used for or with communications systems, Article 800 does not cover wiring of this equipment. Instead, Article 645 provides requirements for wiring contained solely within an information technology equipment (computer) room. (See 645.4 for a description of the type of information technology equipment room to which Article 645 applies.) Article 725 provides requirements for wiring that extends beyond a computer room and also covers wiring of local area networks within buildings. Article 760 covers wiring requirements for fire alarm systems.

The NEC Handbook goes on to say that when telephone system wiring is also used to transmit data, it is covered by Article 800. Examples of this would be a fax or modem line extending into a building by a telephone service provider. The Handbook also states “communications equipment (see Article 100 for definition), such as the private automatic branch exchange shown in Exhibit 800.1, and all of the premises wiring for a communications circuit (see 800.2 for definition), are subject to the requirements of Article 800.” One may conclude from this sentence in the Handbook, that “all the premise wiring” from a telephone switch is covered under Article 800. However, keep in mind it is still referring to the definitions for a “communications circuit,” and that is a circuit that is being extended by a communication utility.

For a privately owned telephone switch, the cabling extending from the switch to the telephone outlets may be covered under Article 725. Article 725.121(A)(4) identifies the power sources for Class 2 and Class 3 circuits as “listed information technology (computer) equipment limited-power circuits.” The code then provides an informational note referring to UL 60950-1-2003, Standard for Safety of Information Technology Equipment. This UL document describes the requirements for a piece of equipment to be considered “information technology equipment.” With UL 60950-1-2003 there is a variety of equipment listed that includes data switches, routers, and telephone equipment such as a private branch exchange or telephone switch.

Therefore, if a privately owned telephone switch is UL 60950-1-2003 compliant, the cabling from it can be covered under Article 725. What is important to note here is that NEC 725.121(A)(4) provides a reference to the UL standard identifying information technology equipment and the UL standard goes on to identify a PBX as a type of information technology equipment.

Don’t be fooled by the cable markings on commonly seen on network and telephone cabling. Although we see CM, CMR, and CMP listed cables used for network cables, that doesn’t mean these are Article 800 circuits. Article 725 allows for the substation of CM, CMR, and CMP cables for the Class 2 and Class 3 listed cables.

A little grey in the interpretation

The code is a complex document that has been revised repeatedly. Some interpretations seem crystal clear while other interpretations are still in the grey. From the code interpretations listed above, I would argue that clearly Ethernet and similar network copper cabling is covered by Article 725. Communication circuits, as defined by the NEC, are circuits that extend from a utility (telephone company) to some type of terminal equipment (telephone, fax, or answering machine). I believe these two interpretations are very clear.

The grey area is the cabling from the PBX to the telephones. The NEC Handbook states, “communication equipment, such as a private automatic branch exchange shown in Exhibit 800.1, and all of the premises wiring for a communications circuit (see 800.2 for definition), are subject to the requirements of Article 800”—even though the UL Standard 60950-1 referenced from Article 725 lists a PBX as information technology equipment. For telephone cables from a PBX, it appears either 800 or 725 could be applied. It turns out that the code requirements in applying 725 or 800 are so similar that installing a cable compliant to Article 725 is also compliant for Article 800.

Telephone circuits that are VoIP circuits are no different than other local area network (LAN) circuits and would be covered under Article 725. With the technology shift to more companies installing VoIP systems for telephony applications, the role of Article 800 inside a building is diminishing. As more and more services are transported over Ethernet copper circuits, it is important for us to know that we need to apply Article 725 to their installation.

Timothy Kuhlman is an electrical engineer with CH2M Hill, where he is a technologist in the industrial and advanced technologies business unit. He has 23 years of experience in the field of telecommunication system design and is registered as professional engineer in 15 states. He is a member emeritus of the Consulting-Specifying Engineer editorial advisory board.


CM, CMR, and CMP: NEC cable listing designation for communication cables per article 800.110. CM: general purpose communications; CMR: riser communications; CMP: plenum communications. 

General purpose cable listings: CL2 or CL3. Per substitution table 725.154(G) cable types CM allowed. 

Plenum cable listings: CL2P or CL3P. Per substitution table 725.154(G) cable types CMP allowed. 

Riser cable listings: CL2R or CL3R. Per substitution table 725.154(G) cable types CMR allowed.

RoIP: Radio over Internet protocol (IP). Often a traditional FM channelized or trunked system integrated using Internet protocols.

SoIP: Security over Internet protocol (IP). A security system using Ethernet or IP as its transport protocol. 

VoIP: Voice over Internet protocol (IP).